USF Libraries
USF Digital Collections

Translation, adaptation, and validation of an instrument to evaluate HIV/AIDS knowledge and attitudes for use with Salva...

MISSING IMAGE

Material Information

Title:
Translation, adaptation, and validation of an instrument to evaluate HIV/AIDS knowledge and attitudes for use with Salvadorian high school students
Physical Description:
Book
Language:
English
Creator:
Zometa, Carlos Salvador
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla.
Publication Date:

Subjects

Subjects / Keywords:
confirmatory factor analysis
El Salvador
validation
cross-cultural adaptation
HIV/AIDS knowledge and attitudes
Dissertations, Academic -- Interdisciplinary Education -- Doctoral -- USF   ( lcsh )
Genre:
government publication (state, provincial, terriorial, dependent)   ( marcgt )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Summary:
ABSTRACT: This study translated, cross-culturally adapted and validated an instrument's scores for use in public high schools in San Salvador, El Salvador. The original instrument consisted of items developed by the Centers for Disease Control and Prevention (CDC) to assess HIV/AIDS knowledge and five dimensions of attitudes (Abstinence, Peer-pressure, Condom use, Drug use, and Threat of HIV) in grades 7 to 12 in the United States. Items were translated into Spanish using the back-translation method. The instrument was cross-culturally adapted using guidelines proposed by Guillemin, Bombardier, and Beaton (1993). A cross-culturally equivalent version of the original instrument was obtained using three different Salvadorian review panels and two pretests with Salvadorian high school students. An expert panel of HIV Salvadorian professionals validated the content and established its cultural acceptability for public school use.A total of 483 students from 30 randomly selected public high schools in El Salvador participated in a series of validation studies. Confirmatory factor analysis of the translated instrument was used to evaluate the factorial validity of the five-factor attitudinal model. As part of the validation process, the translated Abstinence and Condom use subscales from the CDC were correlated with similar translated subscales from Basen-Engquist et al.'s (1999) published study as a measure of concurrent validity. Finally, internal consistency reliability (Cronbach's alpha) was determined with 483 students and test-retest reliability was obtained with a subsample of 39 students. Six major conclusions were: (1) The methodology used was successful in cross-culturally adapting the instrument. (2) HIV/AIDS content was rated as culturally acceptable and valid for use in public high schools of El Salvador.(3) The reliability of the scores from the knowledge section was moderate (test-retest reliability coefficient = .49 and coefficient alpha = .57). (4) Reliability (coefficient alpha) of the five attitudinal subscales was inconsistent: .55 (Peer-pressure), .58 (Abstinence), 0 (Condom use), .24 (Drugs), and .30 (Threat of HIV). (5) Confirmatory factor analysis provided support for a 4-factor attitudinal model (Peer-pressure, Abstinence, Drug use, and Threat of HIV). (6) Concurrent validity of the translated CDC Abstinence subscale was strong. The results provided support for the methodology to cross-culturally adapt an instrument. The psychometric properties from the knowledge section and the attitudinal component related to abstinence were acceptable but additional research is needed before the Spanish instrument can be used with confidence in El Salvador.
Thesis:
Thesis (Ph.D.)--University of South Florida, 2004.
Bibliography:
Includes bibliographical references.
System Details:
System requirements: World Wide Web browser and PDF reader.
System Details:
Mode of access: World Wide Web.
Statement of Responsibility:
by Carlos Salvador Zometa, III.
General Note:
Includes vita.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 269 pages.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001498168
oclc - 57714944
notis - AJU6771
usfldc doi - E14-SFE0000507
usfldc handle - e14.507
System ID:
SFS0025198:00001


This item is only available as the following downloads:


Full Text

PAGE 1

Translation, Adaptation and Validation of an Instrument to Evaluate HIV/AIDS Knowledge and Attitudes for use with Salvadorian High School Students by Carlos Salvador Zometa, III A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Interdisciplinary College of Education University of South Florida Major Professor: Robert Dedrick, Ph.D. Date of Approval: August 1, 2004 Keywords: HIV/AIDS Knowledge and attitude s, confirmatory factor analysis, crosscultural adaptation, validation, El Salvador Copyright 2004, Carlos S. Zometa

PAGE 2

i TABLE OF CONTENTS List of Tables v List of Figures viii Abstract ix Chapter 1. Introduction 1 Purpose of the Study 9 Specific Purposes 9 Limitations 11 Summary of the Introduction 11 Chapter 2. Literature Review 13 Purpose of the Study 13 El Salvador 14 HIV epidemic 15 HIV education 16 Adolescent Risk Factors 19 HIV/AIDS Education 21 Effective Components 22 Role of Behavioral Theories 23 Knowledge of HIV/AIDS 26 Modification of Attitudes 28 Development of the CDC HIV/AIDS Instrument 32 Content and Construct Validity 34 Content Validity 34 Construct Validity 35 Reliability 37 Cross-Cultural Adaptation 39 The different guidelines 39 Guidelines from the IQOLA 43 Synthesis of Cross-Cultural Adaptation Methods 44 Spanish Language Adaptation and Factorial Validation 47 Adaptation of an HIV/AIDS Instrume nt in Hong Kong 50 Summary of Literature Review 51 Chapter 3. Methods 53 Purpose of the Study 53 Specific Purposes 55 Construction of the English Language Version of the CDC Instrument 55

PAGE 3

ii HIV Knowledge 56 HIV/AIDS Attitudes 58 Study 1. Forward-translation, Synthe sis and Back-translation of Instrument 60 Forward Translation 60 Synthesis 62 Back-Translation 63 Study 2. Conceptual Equivalence by Salv adorian Expert Panel 63 Study 3. Cultural Adaptation, Content Validity and Readability Review by Salvadorian Expert Panel 66 Cultural Adaptation 66 Content Validity 67 Role of Salvadorian Lay Panel 68 Reading Level of the Instrument 68 Study 4. Pretesting with Salvadorian Adolescents 69 Study 5. Confirmatory Factor Analysis and Concurrent Validity 71 Confirmatory Factor Analysis 71 Concurrent Validity 73 Data Collection 74 Research Protocol 74 Data Analysis of Specific Purposes 76 Summary of the Methods Section 79 Chapter 4. Results 80 Purpose of the Study 80 Study 1. Forward-translation, Synthesis and Back Translation of Instrument 81 Forward-translations 81 Synthesis 81 Back-translation 89 Study 2. Conceptual Equivalence by Salvadorian Panel 90 Results from the Salvadorian Expert Panel 90 Review by Lay Panel 93 Study 3. Cultural Acceptability, Content Va lidation and Readability 95 Cultural Acceptability 95 Content Validation 97 Readability 100 Study 4. Pretest with Salvadorian High School Students 101 Study 5. Reliability Study, Confirmatory Factor Analysis and Concurrent Validity Reliability Study 105 Reliability of Knowledge Scores 105 Participants 105

PAGE 4

iii Internal Consistency 107 Test-retest reliability 109 Reliability of Attitudinal Scores 111 Descriptive Statistics of Attitudinal Items 111 Descriptive Statistics of Subscales 112 Internal Consistency 114 Test-retest 116 Reliability of the Attitude s Scores from Basen-Engquist et al. (1999) 117 Descriptive Statistics 117 Internal Consistency 119 Test-retest 119 Study Five. Confirmatory Factor Analysis 120 Model 1 120 Fit Indices 120 Standardized Factor Loadings 120 Correlation of the Factors 124 Modification of the Model 125 Model 2 and Model 3 127 Model 4 Confirmatory Factor Analysis of Each Subscale 129 Model 5 and 6 Confirmatory Factor Analysis 131 Model 5 131 Model 6 133 Correlated Errors 133 Standardized Factor Loadings 134 Study Five. Concurrent Validity 135 Summary of the Results 136 Chapter 5. Discussion 137 Purpose of the Study 137 Purpose One 138 Differences in the Knowledge Section 139 Differences in the Attitude Section 140 Purpose Two 143 Purpose Three 144 Readability 144 Content Validity 144 Purpose Four 146 Reliability of the Knowledge Scores 146 Reliability of the Attitude Scores 148 Purpose Five 151 Confirmatory Factor Analysis of Model 6 152 Purpose Six 158

PAGE 5

iv Significance and Implications of the Results 159 Significance 159 Implications 166 Limitations and Directions for Future Research 170 References 172 Appendices 182 Appendix A. The English Knowledge and Attitude Survey 183 Appendix B. Sample Form for the Forward-translation into Spanish 188 Appendix C. Sample Form Sent to Bili ngual Reviewers for the Synthesis 189 Appendix D. Sample of Back-translation Form 190 Appendix E. Form for Semantic Equiva lence (condensed version) 191 Appendix F. Cultural and Conten t Validity Form 193 Appendix G. Letter of permission from the government 194 Appendix H. Informed Consent for Students in the Pretest 195 Appendix I. General Informed Consent Form 197 Appendix J. Informed Consent for Test-retest 199 Appendix K. Results of the Forward-transl ation of the English Instrument 201 Appendix L. Results of Synthesis of Translations by the Bilingual Reviewers 206 Appendix M. Results of the Back-translation 213 Appendix N. First Spanish Version 220 Appendix P. Summary of Results from the Conceptual Equivalence 222 Appendix P. Individual Results fr om Cultural Acceptability and Content Validity of Knowledge and Attitudinal Items 224 Appendix Q. Final Spanish Instrument 231 Appendix R. Final Spanish Instrument 254 About the Author End Page

PAGE 6

v List of Tables Table 1. Items Assessing HIV Knowledge 57 Table 2. Source of Attitudinal Items 60 Table 3. Qualifications of Translators 61 Table 4. The Definition of Semantic, Idio matic, and Conceptual Equivalence 65 Table 5. Qualifications of Panelists 66 Table 6. Characteristics of Salvadorian Lay Panel 68 Table 7. Hypothesized Five-factor Model of Attitudes From the CDCs items 72 Table 8. Basen-Engquist et al.s (1999) Condom and Abstinence Model 74 Table 9. Analysis of the Six Specific Purposes 77 Table 10. Problematic Items and Comment s Made by Bilingual Reviewer 1 82 Table 11. Problematic Items and Comment s Made by Bilingual Reviewer 2 83 Table 12. Problematic Items and Comment s Made by Bilingual Reviewer 4 85 Table 13. Problematic Items Discussed by Bilingual Reviewer 1 and 4 87 Table 14. Bilingual Reviewer 3s Comment s About the First Spanish Version 88 Table 15. Items that Differed in the Origin al Version and the Back Translation 90 Table 16. Suggestions and Changes to the Knowledge Items 91 Table 17. Suggestions and Changes to the Attitudinal Items 93 Table 18. Recommendations Made by the Lay Panel 94 Table 19. Average Rating of Appropriat eness for Salvadorian Culture by Seven Panel Members 96 Table 20. Mean Scores for Content Va lidation by Seven Panel Members 98

PAGE 7

vi Table 21. Changes to Knowledge Items as Suggested by the Salvadorian Panel During the Cultural Review 99 Table 22. Changes to Attitudinal Items as Suggested by the Salvador Panel During the Cultural Review 100 Table 23. Pretest Results of the Knowledge Section 102 Table 24. Pretest Results of the Attitudinal Section 103 Table 25. Characteristics of the 10 Public High Schools 106 Table 26. Score Distribution of the 15 Knowledge Items 107 Table 27. Mean Score and Item-to-total Co rrelation of the 15 True/False Items 108 Table 28. Characteristics of the 15 True /False Knowledge Scores at Time 1 and Time 2 110 Table 29. Descriptive St atistics for the 23 CDC Translated Items 112 Table 30. Descriptive Statistics of Attitudinal Subscales 113 Table 31. Pearson Correlation Matrix of th e Five Attitudinal Subscales (n = 393) 114 Table 32. Cronbachs Alpha and Item-tototal Correlation of the 23 Translated CDC Attitudinal Items 115 Table 33. Cronbachs Alpha and Item-to-total Correlation of the Subscale 116 Table 34. Characteristics of Attitudina l Items at Different Times 117 Table 35. Descriptive Sta tistics for Basen-Engquist et al.s (1999) Items 118 Table 36. Descriptive Statistics of Basen-Engquist et al.s (1999) Attitudinal Subscales 119 Table 37. Cronbachs Alpha and Item-to-total Correlation of Basen-Engquist et als (1999) Subscales 119 Table 38. Comparison of Basen-Engqui st et al.s (1999) Condom and Abstinence Subscales at Different Times 120 Table 39. Goodness of Fit Indices to Examine Model Fit 121

PAGE 8

vii Table 40. Standardized Factor Loading, Error Variance, and Squared Multiple Correlation Coefficient of th e Attitudinal Variables from Model 1 124 Table 41. Correlation of the Five factors 125 Table 42. New Paths Suggested by the Lagrange Test 126 Table 43. Goodness of Fit Indices of Model 1 and Model 2 127 Table 44. Standardized Factor Lo adings for Model 1 and Model 3 128 Table 45. Summary of the Goodness of Fit Indices from the CFA of the Five Subscales 130 Table 46. Summary of the Standard ized Factor Loadings from the Separate CFAs 131 Table 47. Description of New Pa ths Among the Errors in Model 6 134 Table 48. Standardized Factor Loadings for Model 6 135 Table 49. Correlation Between Ba sen-Engquists Subscales and the CDCs Items 136 Table 50. Test-retest and Internal C onsistency of the Attitudinal Subscales 149 Table 51. Items with an Apparent Association to More than One Subscale 155 Table 52. Paths in Model 6 from an Error of an Item to a Correlated Error 157 Table 53. Suggested Changes to the Attitudinal Section 167

PAGE 9

viii List of Figures Figure 1. Five Dimensional Attitudinal Model. 8 Figure 2. Steps Outlining the Research Methodol ogy of the Multi-phase study 54 Figure 3. Standardized Factor Loadings of the Five-dimensional At titudinal Model 122 Figure 4. Proposed Four-dimensional Atti tudinal Structure of Model 5 132 Figure 5. Standardized Factor Loadings of th e Four-dimensional Attit udinal Model 153

PAGE 10

ix Translation, Adaptation, and Validation of an Instrument to Evaluate HIV/AIDS Knowledge and Attitudes for use with Salvadorian High School Students Carlos Salvador Zometa, III ABSTRACT This study translated, crossculturally adapted and validated an instrument's scores for use in public high schools in Sa n Salvador, El Salvador. The original instrument consisted of items developed by the Centers for Disease Control and Prevention (CDC) to assess HIV/AIDS knowle dge and five dimensions of attitudes (Abstinence, Peer-pressure, Condom use, Drug use, and Threat of HIV) in grades 7 to 12 in the United States. Items were translat ed into Spanish using the back-translation method. The instrument was cross-culturally adapted using guidelines proposed by Guillemin, Bombardier, and Beaton (1993). A cross-culturally equivalent version of the original instrument was obtaine d using three different Salvad orian review panels and two pretests with Salvadorian high school student s. An expert panel of HIV Salvadorian professionals validated the content and established its cultural acceptability for public school use. A total of 483 students from 30 randomly selected public high schools in El Salvador participated in a seri es of validation studies. Confir matory factor analysis of the translated instrument was used to evaluate the factorial validit y of the five-factor attitudinal model. As part of the valida tion process, the translated Abstinence and Condom use subscales from the CDC were corr elated with similar translated subscales from Basen-Engquist et al.'s (1999) published study as a measur e of concurrent validity.

PAGE 11

x Finally, internal consistency reliability (Cronbach's alpha) was determined with 483 students and test-retest reliability was obt ained with a subsample of 39 students. Six major conclusions were: (1) The me thodology used was successful in crossculturally adapting the instrument. (2) HI V/AIDS content was rated as culturally acceptable and valid for use in public high schools of El Salva dor. (3) The reliability of the scores from the knowledge section was modera te (test-retest reli ability coefficient = .49 and coefficient alpha = .57). (4) Reliability (coefficient alpha) of the five attitudinal subscales was inconsistent: .55 (Peer-pressure), .58 (Abstinence), 0 (Condom use), .24 (Drugs), and .30 (Threat of HIV). (5) Confirma tory factor analysis provided support for a 4-factor attitudinal model (Peer-pressure, Ab stinence, Drug use, and Threat of HIV). (6) Concurrent validity of the translated CDC Abstinence subscale was strong. The results provided support for the met hodology to cross-culturally adapt an instrument. The psychometric properties from the knowledge section and the attitudinal component related to abstinence were acceptabl e but additional research is needed before the Spanish instrument can be used with confidence in El Salvador.

PAGE 12

CHAPTER 1 INTRODUCTION According to the Joint United Nations Program on HIV/AIDS (UNAIDS, 2002), Acquired Immunodeficiency Syndrome (AIDS) has affected more than 40 million people worldwide. Currently, leaders from both the developed and underdeveloped regions of the world are working together to fight the epidemic. Despite these efforts, an estimated 5 million new cases of infection with the Human Immunodeficiency Virus (HIV) occur yearly (UNAIDS, 2004). Of particular relevance to this study is the number of cases of AIDS and HIV in Latin America. These cases have multiplied recently, estimated at 1.2 million and projected to increase (CDC, 2002a; UNAIDS, 2002). The trend in Latin America foreshadows the epidemic currently ravaging countries in Africa and underscores the urgency to protect the population in Latin America. Since a cure has not been found and vaccines continue to prove to be unsuccessful, expanding health promotion programs and educating people about HIV transmission are the best alternatives. Adolescents and young adults represent a high-risk segment of the population that can greatly benefit from educational programs. This idea is being promoted by a plan from the United Nations that uses a 10-step strategy to prevent the transmission of HIV among young persons. The plan calls for an increase of school-based HIV prevention programs in areas with a high prevalence of AIDS (United Nations Children Fund [UNICEF], UNAIDS, 2002, & WHO, 2002). The effect of such programs is that youths with a better education acquire knowledge, 1

PAGE 13

confidence and social skills to protect themselves against the virus (UNICEF et al., 2002, p. 27). A wealth of information from the evaluations of HIV and teen-age pregnancy prevention programs has been generated in the United States (Kirby, 1999, 2001). This information is vital to the development, implementation, and evaluation of new programs. Curriculum planners of school-based programs who need guidance and information can rapidly access valid information from the Centers for Disease Control and Preventions report, Programs That Work (CDC, 1999), and the report, Emerging Answers authored by D. Kirby (2001). Both sources published a list of evaluated programs in the United States that have successfully reduced the risk of HIV infection among adolescents. The major findings were that the curriculum of the most effective programs contained a knowledge component that included: how HIV is transmitted, abstinence as the best way to avoid pregnancy, the risk associated with having multiple partners for sexual intercourse, and condom efficacy for those who decide to have sexual intercourse (CDC, 1999; Kirby, 2001). The curriculum of successful programs was also complemented with material from behavioral theories that targeted psychosocial factors and skills. Psychosocial factors are norms, attitudes, self-efficacy, and beliefs (Basen-Engquist et al., 1999); these factors can directly affect antecedents of sexual behaviors. As this information is disseminated and integrated into the curriculum of prevention programs in Latin America, it is anticipated that positive results will be achieved. However, the results of state-of-the art programs must be documented with psychometrically sound instruments written in Spanish. Unfortunately, the lack of rigorously validated instruments to assess needs and the effectiveness of HIV education 2

PAGE 14

in Spanish-speaking countries are major obstacles facing evaluators today in Latin America. Researchers in Latin America have two options. The first option is to develop a new instrument, and the second option is to adapt an existing one from another study (Guillemin, Bombardier, & Beaton, 1993). The former option is problematic because the methodology requires lengthy developmental lead time and ample resources (Guillemin et al., 1993). The latter option is the best option for countries with few resources because a cross-cultural adaptation of an instrument written in English is less demanding and rapidly produces an instrument that can be used to conduct needs assessments and program evaluation. The cross-cultural adaptation of a pre-existing instrument is a process designed to ensure that the same construct is measured in two different populations. An adaptation is necessary to limit the impact of culture and language on the measurement properties of an instrument. Individuals culture and language mediate their perception of the construct (Brislin, Lonner, & Thorndike, 1973). When a construct has a different interpretation in two populations, the results from the instrument are not valid. Therefore, methodology to establish a cross-cultural adaptation is needed to minimize the effect of culture and language on the interpretation of an instrument (Brislin et al., 1973; Guillemin et al., 1993). Several guidelines and methods have been proposed to adapt instruments cross-culturally to different populations (Bullinger et al., 1998; Guillemin et al., 1993; Herdman, Fox-Rushby, & Badia, 1998; Skevington, 2002). Researchers philosophical views of how to proceed with the cultural adaptation affect their preferred method and therefore these methods vary (Vijver & Leung, 1997). The effectiveness of different 3

PAGE 15

methods has been supported through a series of validation studies in which instruments written in English have been adapted to other cultures (Bullinger et al., 1998; Skevington, 2002). A comparison of the methods revealed several common strategies that have been applied to different situations. These strategies include three phases. The first phase is the translation; the second is the adaptation of the instrument to a different culture; and the last phase is an examination of the psychometric properties. The first strategy in the process of cross-culturally adapting an instrument consists of a series of forward-translations and subsequent back-translations to the original language (Brislin et al., 1973). The forward-translation process consists of changing text from one language to another so as to retain the same meaning. A committee and experienced translators work together to review and synthesize the forward-translation. Afterwards, the best translation is examined by individuals unfamiliar with the goals of the project; these individuals back-translate the content to the original language. The original instrument is then compared to the back-translation to confirm the original meaning of the instrument. The second step is to obtain a linguistic adaptation of the items to meet the cultural characteristics of the target population (Guillemin et al., 1993). Bullinger et al. (1998), Herdman et al. (1998), and Skevington (2002) have offered a range of complex methods to adapt instruments. However, Guillemin et al.s (1993) guidelines are less complicated, and are easily implemented in an environment with scarce resources. Guillemin et al.s guidelines consist of a three-step procedure that relies on a panel formed of subject matter experts, bilingual reviewers, and lay persons who compare the original item with its translation to determine the equivalence of the instruments 4

PAGE 16

concepts. Semantic equivalence, the first step, involves a review focusing on grammatical structure and equal meaning of the translated items. The second step, idiomatic equivalence, involves translating idioms and colloquialisms into expressions understood by the target culture. Conceptual equivalence, the last step, determines the use of words in the same social setting as the target population (Guillemin et al., 1993). The third and final stage makes use of statistical procedures to verify and compare the measurement characteristics of the instrument (Bullinger et al., 1998; Gandek & Ware, 1998). Gandek and Ware (1998) state that a cross-cultural adaptation relies on validation strategies to compare the equivalence of a construct in different countries. They also report that determining content and construct validity provides evidence about the instruments appropriateness and serves as a measure of equivalence of interpretation across different populations. Additionally, Gall, Borg, and Gall (1996) recommend establishing the reliability and validity of the scores to increase the confidence of judgments derived from instruments. Therefore, the reliability and validity of the instruments scores are vital for an instrument that is adapted cross-culturally. This multi-strategy approach for a cross-cultural adaptation makes use of a step-wise sequence that gradually transforms the instrument for use in another culture. There are several conditions that facilitate the adaptation process. First, if the constructs being measured are well defined and have similar interpretations in the two cultures, the process is facilitated. Another facilitating condition is if the variables have universal recognition (Hui & Triandis, 1985). An example of a universal concept is the knowledge component of HIV in educational programs. The risk factors and behaviors associated with transmission of HIV have been clearly defined. Because they are taught in many 5

PAGE 17

countries, they are also universally recognized, which facilitates adaptation. In contrast, attitudes related to HIV are less universal and, therefore, their definition across cultures has to be analyzed. The analysis of the attitudinal construct involves verification of the importance of the items in the target culture and validation of the scores from the instrument. When languages and cultures are similar, the adaptation from the original culture to a different culture is straightforward (Hui & Triandis, 1985). Conversely, when language and cultural variations are great, adaptations become increasingly complex. The adaptation of an instrument developed in the United States is expected to progressively increase in difficulty when adapted for use in England, Spain, and China, respectively. Based on the previous information, an instrument, published by the Centers for Disease Control and Prevention (2002b) and used to evaluate HIV education in grades 7 to 12 in the United States, was adapted for use in El Salvador. The instrument is conveniently available at the CDCs website (accessed May 2004), and serves as an item bank that can be customized to evaluate HIV curricula. The instrument was developed between 1989 and 1992 by an independent company (IOX Assessment Associates) and was designed to assess constructs in seven broad areas: Knowledge, Beliefs, Attitudes, Student Confidence, Friends Views, Intentions, and Behavior. The validity of the seven constructs was established by a series of reviews by panels of HIV experts in the United States. Further, to ensure that content, vocabulary, and directions were appropriate for students in grades 7 to 12, the instrument was continuously field-tested with small groups of students (CDC, 2002b). After determining 6

PAGE 18

that the instrument was appropriate for use with adolescents, the final version was reviewed, edited, and approved by a national panel of HIV experts. The instrument for this particular study in El Salvador was constructed by combining the CDC's items designed to assess HIV/AIDS knowledge and attitudes of students in high school. To assess HIV/AIDS functional knowledge, 15 knowledge items from the CDCs item bank were selected. The Attitude section was constructed by selecting 23 items from the CDCs item bank that represented five different attitudinal dimensions (see Figure 1). An additional six items from Basen-Engquist et al.'s (1999) published study were included to evaluate the concurrent validity of the CDC's Abstinence and Condom use subscales. Demographic information to determine gender, age, grade, and previous HIV education was obtained through additional items. All items were translated into Spanish, adapted for use in El Salvador, and tested with students in public high schools in El Salvador. The translation and adaptation of an instrument for use in another cultural group poses a number of methodological and measurement challenges. In this study, the methodological challenge was to obtain a cross-cultural equivalent, adjust the instruments content to ensure it was relevant to the target population, and ensure that the material was not culturally offensive to Salvadorian students. Further, the instruments readability had to be at the appropriate Spanish level to assure the content was understood by the students. The measurement challenges included obtaining acceptable measures of reliability and validity within the target population. 7

PAGE 19

Threat of HIV Drug use Condom use Abstinence Pee r p ressure 1 6 11 15 19 2 7 12 20 16 21 23 4 9 13 17 24 5 10 8 14 18 Figure 1. Five-dimensional attitudinal model. 25 E1 E6 E11 E15 E19 E2 E7 E12 E20 E8 E16 E21 E23 E4 E9 E13 E17 E24 E5 E10 E14 E18 E25 8

PAGE 20

Purpose of the Study The purpose of this study was to take an instrument developed by the CDC to assess HIV/AIDS knowledge and attitudes of students in grades 7 to 12 in the United States and adapt it for use in El Salvador. The first phase utilized the back-translation method to obtain the first Spanish version of the instrument. In the next phase, an expert panel of reviewers from El Salvador established the conceptual equivalence of the Spanish instrument. The third phase evaluated the instruments content validity and cultural acceptability through a panel composed of HIV experts, health professionals, schoolteachers, and students from El Salvador. The last phase involved collecting evidence of the validity of the scores from the Spanish version of the CDC instrument; 483 public high school students in the metropolitan area of San Salvador participated in the validation phase. Reliability of the scores was measured by the test-retest method and coefficient alpha. Confirmatory factor analysis was used to evaluate if the attitudinal construct had five different dimensions (Peer-pressure, Condom use, Abstinence, Drug use, and Threat of HIV). As part of the concurrent validation process, the Condom and Abstinence subscales from the translated CDC instrument were correlated with a set of existing attitudinal items from a published study by Basen-Engquist et al. (1999). Specific Purposes 1. Determine the success of the cross-cultural adaptation by contrasting the first translated version of the CDC instrument with the final Spanish version of the translated CDC instrument. 9

PAGE 21

2. Determine, using a Salvadorian panel, the level of cultural appropriateness of the Spanish version of the translated CDC HIV/AIDS knowledge and attitudinal items for use in El Salvador. 3. Ascertain the level of readability of the final Spanish version of the translated CDC instrument and evaluate content validity. 4. Evaluate the reliability of the scores from the Spanish version of the translated CDC instrument using a sample of Salvadorian students in grades 10, 11, and 12. 5. Determine if the Spanish version of the translated CDC instrument contains five attitudinal dimensions (Peer-pressure, Condom use, Abstinence, Drug use, and Threat of HIV) when analyzed using confirmatory factor analysis. 6. Evaluate the concurrent validity of the Abstinence and Condom use subscales from the Spanish version of the translated CDC instrument. Limitations The following limitations were encountered: 1. Self-reported measures have an unknown effect on the validity and reliability of the scores from an instrument. 2. Different response rates among students in different schools may have introduced bias into the study. 3. Reliability and validity data for instruments used in Latin America are not available. 4. Reliability of the scores and construct validity of the CDCs items for use in the United States have not been determined. 10

PAGE 22

5. The sample of Salvadorian adolescents used in the study (grades 10, 11, and 12) may not be representative of all adolescents in Latin America and thus limits generalizability. 6. The selection of schools in El Salvador was limited to the metropolitan area of the capital San Salvador. This limits the sample to urban students and excludes students from other cities and rural areas of El Salvador. 7. The sample excludes private schools. Private schools have a larger percentage of the population from the higher socio-economic group than students in public schools. 8. The sample consists of adolescents in school and excludes students who dropped out of school or those who did not have the resources to go to school. Summary of the Introduction The economic and social structure of many countries in Latin America has been affected by the transmission of HIV/AIDS among their populations. Adolescents are particularly susceptible to infection with HIV. Therefore, to reduce the incidence of HIV in Latin America, officials have relied on schooland community-based educational programs that prevent the transmission of HIV. To create effective programs, tools that provide reliable and valid scores must be used to evaluate these programs. The availability of research instruments can be increased in Latin America by translating and adapting existing instruments from other cultures. Because language and culture can have various effects on how people interpret an instrument, careful strategies must be followed to ensure an appropriate adaptation. In this study, an existing instrument developed by the CDC to evaluate HIV/AIDS knowledge and attitudes in 11

PAGE 23

grades 7 to 12 was translated into Spanish and cross-culturally adapted; the scores were then validated for use in El Salvador. 12

PAGE 24

CHAPTER 2 LITERATURE REVIEW Purpose of the Study The purpose of this multi-phased study was to translate, cross-culturally adapt, and validate the scores from an instrument that assessed the HIV/AIDS knowledge and attitudes of high school students in El Salvador. The first phase utilized the back-translation method to obtain the first Spanish version of the instrument, which was developed by the CDC to evaluate HIV/AIDS knowledge and attitudes of students in grades 7 to 12. In the next phase, an expert panel of reviewers from El Salvador established the conceptual equivalence of the Spanish instrument. The third phase evaluated the instruments content validity and cultural acceptability through a panel composed of HIV experts, health professionals, schoolteachers, and students from El Salvador. The last phase involved collecting evidence of the validity of the scores from the Spanish version of the CDC instrument; 483 public high school students in the metropolitan area of San Salvador participated in the validation phase. Reliability of the scores was measured by the test-retest method and coefficient alpha. Confirmatory factor analysis was used to evaluate if the attitudinal construct had five different dimensions (Peer-pressure, Condom use, Abstinence, Drug use, and Threat of HIV). As part of the concurrent validation process, the Condom and Abstinence subscales from the translated CDC instrument were correlated with a set of existing attitudinal items from a published study by Basen-Engquist et al. (1999). 13

PAGE 25

The literature review provides an overview of the epidemiology of HIV in El Salvador and describes specific aspects of the HIV curriculum in the public high schools in San Salvador. As the HIV epidemic spreads around the world, adolescents are at an increased risk of infection due to a variety of risk factors. The review emphasizes that the most sensible way to reach adolescents is by increasing school-based prevention programs. Additionally, the effectiveness of curricula that integrate components of behavioral theories is discussed in view of the fact that these programs are effective in reducing the risks associated with HIV infection. Specifically, the review highlights the role of attitudes and knowledge as mediating variables in behavioral theories. Methodological issues related to an instruments translation and adaptation from one language to another are discussed. These issues include obtaining the correct literal translation by the back-translation method and establishing guidelines for cross-cultural adaptation. Next, measurement issues related to the scores from an adapted instrument are discussed. This includes measuring the reliability and validity of the scores in a new population. In particular, the review will discuss several examples of construct validation of an instrument as a method to ascertain a cross-cultural adaptation. El Salvador El Salvador is the smallest country in Central America with an area of 21,040 square kilometers, but is one of the most populated with 6.4 million people (Central Intelligence Agency, 2004). A birth rate of 28 per 1,000 contributed to an annual population growth of 1.8%. The impact of young persons on the composition of the population is evident by their large demographic distribution of 37% and 58% in the categories of 0 to 14, and 15 to 64 years of age, respectively (Central Intelligence 14

PAGE 26

Agency, 2004). As the younger generation enters the reproductive years and becomes sexually active, it will encounter the HIV epidemic. HIV epidemic. The Health Department of El Salvador estimated a prevalence of 5,876 cases of HIV and 5,793 of AIDS; together they represent an infection rate of 0.6% of the Salvadorian population (Ministerio de Salud Publica y Asistencia Social, 2003). However, the United States Agency for International Development (2004) and UNAIDS (2000) report an underestimation of HIV cases and project the actual number to be in the range of 20,000 to 30,000. The majority of cases in El Salvador was reported in individuals under the age of 39. The characteristics include a disproportionately high concentration in the urban areas of San Salvador and an infection ratio for male-to-female of 3:1 (USAID, 2004). The prevalence of HIV/AIDS in young persons was reported as 2,500 and 4,000 cases in age groups 15 to 24, and 25 to 34, respectively (Ministerio de Salud Publica y Asistencia Social, 2003). Further, heterosexual contact was the most frequently reported risk factor for transmission. The other risks in decreasing frequency were mother-to-child, homosexual and bisexual contact, and drug use (Ministerio de Salud Publica y Asistencia Social, 2003). Some researchers have questioned the accuracy of the reported risk factors and point out that the large number of infected men suggests an under-reporting of homosexual and bisexual behavior (USAID, 2004). The data indicated that sexual transmission was the most common route of HIV infection, and as such it should be a priority in prevention programs. Another index for sexual activity is the rate of sexually transmitted infections (STI). In El Salvador, the incidence of STIs in 2001 and 2002 was 52,705 and 54,197 cases, respectively (Ministerio de Salud Publica y Asistencia Social, 2003). Since STIs facilitate 15

PAGE 27

transmission of HIV, they are also important to control. The urgency to prevent transmission of HIV is based on the projection of an increase in the prevalence to 1.6% for the year 2010 (World Bank, 2003). To meet the enormous challenge, the government has therefore taken the leadership role and mobilized non-governmental and international organizations to participate in a national plan (Health Department Official, personal communication, September, 2003). These agencies have adopted a very comprehensive plan described in detail at the Health Departments website ( www.mspas.gob.sv accessed April 2004). Briefly, the plan includes prevention, capacity building, and expert consultation throughout the country. Other components include monitoring HIV infections by strengthening epidemiological surveillance and conducting research to determine the effectiveness of their programs (Ministerio de Salud Publica y Asistencia Social, 2003). The Department of Education of El Salvador is also involved, and it has a curriculum that targets the transmission of HIV among adolescents. HIV education. The school system in El Salvador consists of parvularia for pre-school children, basica for students in grades 1 to 9, and media for grades 10 to 12 (high school). San Salvador and its surrounding area are in the Department (State) of San Salvador. San Salvador had a total of 58,028 students in high school, with 30,054 in public and 27,974 in private high schools (Departamento de Educacion de El Salvador, 2002). The distribution of students in public high schools in San Salvador according to grade was as follows: 10 (13,776), 11 (9,315), and 12 (6,608); this sample consisted of 54% female (Departamento de Educacion de El Salvador, 2002). 16

PAGE 28

The official curriculum offered in high school is titled Psicologia de la Adolescencia, which is translated as Psychology of Adolescence, and was written by Lic. Daysi Miriam Figueroa Erazo. The curriculum is spiral and characteristically the concepts are progressively covered during one year and revisited during the latter years of school. The central theme revolved around the mental and physiological changes associated with adolescence. An examination of the curriculum by this researcher determined that students in the 10 th grade learn about personal identity, morality, problem resolution, and sexuality (Figueroa Erazo, 2001a). Sexuality included information and activities about sexual relations, HIV/AIDS, and sexually transmitted infections. In particular, the unit on sexuality was comprehensive as it included didactic information and case studies to review the concepts. Specifically, the topics were: the biological aspects of the male and female anatomy; issues regarding masturbation, prostitution, abstinence, and sexual abuse; and the consequences of sexual activity (Figueroa Erazo, 2001a). The latter section, taught in 11 th grade, informed students about STIs and HIV/AIDS and then discussed the symptoms and risk factors associated with transmission. The students were taught about the origin and proliferation of HIV around the world. The main focus of the HIV lessons was on the six risk factors of transmission that are also taught in the United States. The unit emphasized the symptoms associated with AIDS (Figueroa-Erazo, 2001a). Prevention of HIV centered on the avoidance of risky behavior, such as not having sexual relations with unknown persons, avoiding drug needles, not breast-feeding if infected, avoiding body secretions, and the use of condoms. 17

PAGE 29

Condoms were not vigorously promoted in this section; however, they were discussed in other sections of the curriculum. An advantage of a spiral curriculum is that it progressively introduces a deeper understanding of the content at each grade level. Figueroa Erazos (2001b) curriculum titled the Psychology of Adolescence does not isolate issues of sexuality. Instead its strength was a deeper exploration of sexuality through a social, religious, and personal perspective. Further, the psychology of sexuality was a central theme that included sexual responsibility and consequences of sexual activity. Additionally, schools are allowed to request charlas or mini-lessons from outside sources to complement the curriculum or to reinforce concepts (schoolteacher in El Salvador, personal communication, September, 2003). The non-profit organization Fundasida, an outside source, is very involved throughout the country. They provide educational assistance to public and private schools. An effective way to deliver the curriculum has been through charlas or discussion sessions for students. Their presence has gained the respect of many persons throughout the country, and they have established themselves as leaders in HIV/AIDS services. It is anticipated that the continued effort between the government and many other organizations will heighten awareness and stop the transmission of HIV in young persons. In conclusion, the government of El Salvador has a comprehensive plan to prevent the transmission of HIV. The plan has a specific component to target adolescents in school. Therefore, it is anticipated that a valid instrument will be useful to evaluate the effectiveness of their HIV prevention program. 18

PAGE 30

Adolescent Risk Factors Adolescents in many parts of the world are at increased risk of infection with HIV due to developmental, economic, and social factors. Developmental factors related to emotional, cognitive, and physiological maturity influence adolescents decisions about premature sexual relations that put them at risk of HIV infection, as well as other sexually transmitted diseases (Henderson, Champlin, & Evashwick, 1988). Other factors, such as economic reliance on adults for shelter, food, and protection, made adolescents vulnerable to sexual exploitation by others (Stephenson, 2003; UNESCO, 2003). Drug use and drug-related behaviors among adolescents also increase their risk of infection. While the percentage of drug users in this age group is small, transmission is possible from any of the following behaviors: injecting drugs, sharing needles, and having sex with a known drug user (CDC, 1988, 2003c). The rate of sexual activity among adolescents in the United States and Latin America represented the primary risk of infection with HIV and other sexually transmitted infections (CDC, 2003b, 2003c; Cleland, Ferry, & WHO, 1995). The CDC (1998) and Kann et al. (2000) reported that students in the United States are involved in risky sexual behaviors that include one or more of the following acts: unprotected sexual intercourse (anal and vaginal), contact of the mouth with a sexual organ, a large number of sexual partners, high frequency of sexual acts, and lack of condom usage. Additionally, more students have sexual relations at a younger age. It is estimated that more than half of 17 year olds have had sexual relations (Alan Guttmacher Institute, 1999; CDC, 1998). 19

PAGE 31

Likewise, sexual activity is a concern in Latin America and the Caribbean. The rate varied with some countries reporting much higher rates than others (Cleland et al., 1995). Rodriguez-Kelly (2001) reported that at least 25% of adolescents in Latin America will have sexual intercourse during their high school years, thus increasing their risk of HIV infection. The rate of sexual activity was also a concern in El Salvador, it was reported that 57% of women 15 to 49 years of age reported to be living with a male partner (Asociacion Demographica Salvadorena, 2003). Likewise, men in El Salvador reported their first sexual experience at 16 years of age (Asociacion Demographica Salvadorena, 2003). Regardless of the country, young persons around the world face similar risk factors for HIV. The risks include sexual relations with older partners; power differences that affect condom negotiation, and lack of education (Cleland et al., 1995; UNESCO, 2003). The negative consequences of sexual activity are evident in the rates of adolescent pregnancy, abortion, infection with HIV, and other STIs (Alan Guttmacher Institute, 1999; CDC 2002a; Choi & Coates, 1994; Cleland et al., 1995). A unique problem in underdeveloped countries is the large number of people between 25 and 30 years of age with AIDS. The epidemiological data support the conclusion that these infections were acquired during adolescence (UNAIDS, 2002). The morbidity associated with sexual activity is disturbing, and the human and monetary costs to society are high. Promoting healthy sexual behavior and educating adolescents are the best methods to prevent sexually transmitted infections and infection with HIV. 20

PAGE 32

HIV/AIDS Education In response to these concerns, policymakers have turned to educational programs to mitigate the negative consequences of sexual activity. The CDC (1995, 1999), Kirby (2001), and others (UNAIDS, 1997) have reported that harmful behaviors acquired in adolescence can be changed through educational interventions. In general, these programs have many shortand long-term benefits to society. An immediate effect is the reduction of negative consequences among adolescents who are currently sexually active. The long-term effect will be evident when adolescents enter adulthood and inevitably become sexually active. It is expected that they will then apply the knowledge and skills to prevent diseases. Adolescents in the United States receive formal HIV education either through schoolor community-based programs. The philosophical foundation has become a controversial topic yielding two polar sides. On one side, advocates of abstinence education support the premise that adolescents should wait to have sexual relations until marriage. They prefer knowledge, values, and attitudes that focus on the negative consequences of having sexual relations (Kirby, 2001). The other side favors a comprehensive approach that includes information about abstinence, contraception, and the use of condoms to prevent sexually transmitted diseases. Both sides claim their programs are superior and effective. The effectiveness of abstinence education has not been rigorously tested. But some studies indicate that adolescents did not delay the initiation of intercourse (Kirby, 2001, 2002), and other studies had negative or inconclusive results (Kirby, 2001). 21

PAGE 33

Effective components. The components of effective programs in the United States have been identified through rigorous evaluations of comprehensive HIV/AIDS, pregnancy prevention, and sexuality education programs (CDC, 1999; Kirby et al., 1994; Kirby, 2001). Effective programs contain a knowledge component that includes: how HIV is transmitted; abstinence as the best way to avoid pregnancy; the risk associated with having multiple partners; and to a lesser degree condom efficacy (CDC, 1999, 2004; Kirby, 2001). Additional components that strengthen a curriculum are psychosocial factors and skills. Psychosocial factors are norms, attitudes, self-efficacy, and beliefs (Basen-Engquist et al., 1999), and they can directly affect antecedents of sexual behaviors. Kirby (2001) reported that sexual antecedents, such as sexual beliefs, attitudes toward intercourse or condoms, sexual norms, refusal skills, and contraceptive behavior, be addressed in prevention programs. Unlike the reasons given by supporters of abstinence education, these antecedents are known to influence sexual behavior. Skills that help prevent the transmission of HIV, an additional component of the curriculum, have been taught in schools throughout the United States. The CDC (2004) reported that skills to resist peer pressure to engage in sexual behavior, goal setting, and communication skills related to sexual behaviors were taught in 68%, 62%, and 57% of the schools, respectively. Skills are common in the curriculum; however, their effect has not been widely reported, in part because they are difficult to assess. A paper and pencil test, the common method to assess, is not the most appropriate; instead, the correct procedure is to directly observe if the person has mastered the skill. This poses a challenge for those who want to evaluate sexuality-based skills since the behaviors take 22

PAGE 34

place in private. Nevertheless, an educational program that incorporates skills has been associated with effective outcomes (Kirby, 2001). St. Lawrence et al. (1995) evaluated a program that modified three types of skills related to: (1) discussing condoms before having sex; (2) refusing pressure to have sex; and (3) talking about HIV-risk behavior with peers. They reported a significant increase in condom use and lower frequency of unprotected intercourse (St. Lawrence et al., 1995). A meta-analysis by Johnson et al. (2003) supported the previous findings. Johnson et al. (2003) reported that interventions that reinforced interpersonal skills reduced the risks associated with HIV infection. However, a point that is overlooked is that a knowledge component is a prerequisite to learning a skill. A knowledge component is often complementary in many programs and is therefore a key element in prevention. Finally, the knowledge and attitudinal components are less difficult to evaluate than a skills component. Knowledge and attitudes are evaluated in less time and with fewer personnel than skills. The works of Basen-Engquist et al. (1999) and DiClemente, Zorn, and Temoshok et al. (1986) established the validity and reliability of the previous components and favored their inclusion in an instrument. Additionally, the role of knowledge and attitudes in behavioral theories--as mediating variables that affect sexual behavior--make them an important component to integrate in a curriculum. Role of behavioral theories. The nature of adolescent sexual behavior is complex as a vast number of antecedents combine with individual factors to produce a variety of outcomes (Kirby, 2001). The CDCs (1999) publication Programs That Work and Kirbys (2001) report Emerging Answers summarized the evaluation of pregnancy prevention, sexually transmitted infections, and HIV prevention programs. They 23

PAGE 35

examined the outcomes and confirmed the difficulty of altering sexual behavior in adolescents. They reported that a small percentage of school-based programs had a significant impact in reducing risk factors associated with adolescent activity. A consistent finding among theses two studies and others (Basen-Engquist et al., 1999; Fisher & Fisher, 2000; Kirby, 1999, 2001) has been the success of theory-based prevention programs. Initially, behavior-changing theories were essential in altering and explaining unhealthy behaviors in other areas of public health (Green & Kruter, 1999). Consequently, their usefulness has been transferred to adolescent sexual behavior (Green & Kruter, 1999; Kirby, 2001). The advantage of the theories is the ability to systematically target more than one psychosocial factor at once, inform about disease transmission, and teach skills (Fisher & Fisher, 2000; Kirby, 2001). The social cognitive theory, social influence theory, social inoculation theory, and the theory of reasoned action serve as the foundation of successful prevention curricula with adolescents (Fisher & Fisher, 2000; Kirby et al., 1994). A detailed description of each theory has been provided by Azjen and Fishbein (1980), Bandura (1986), and Fisher and Fisher, (2000). Kirby et al. (1994) described the advantages of using social learning theories in a curriculum to prevent HIV/AIDS and teen-ageage pregnancy among adolescents. They stated that the inclusion of a cognitive and social component facilitates and increases learning among this population. Protective behaviors are learned through both components, and the theories take advantage of the how learning occurs. For example, youths gain [their] understandings and beliefs directly through education and indirectly by observing the behavior of others (Kirby et al., 1994, pp. 353-354). A 24

PAGE 36

knowledge component may be effective when associated with a theory that presents the information in a didactic manner and is reinforced with interactive strategies. Further, they reported the advantage of social learning was related to an increase in learning that occurred when students observed teachers or by the student's interaction with other students. In addition, they cite the importance of these theories in helping adolescents understand and resist societal pressures that can influence behavior. Yet, Basen-Engquist and Parcel (1992) caution that theory alone is not sufficient to guarantee behavior changes among adolescents and that other factors related to the school affect the results. These findings support Kirbys (2002) observation that certain schools have a lower pregnancy rate, and the rate may be influenced by specific characteristics unique to the school. To maximize success, a theory-based program was included with a multi-component program that addressed the many factors that affect sexual behavior (Basen-Engquist et al., 1997). One of the objectives was to influence and thus change mediator variables by engaging students in direct instruction, role playing, or participating in hands-on activities. The other objective was to involve the entire school and thus raise awareness of issues related to sexuality (Basen-Engquist et al., 1997; Coyle et al., 2001). The combination of both objectives had positive results on risk factors. Regardless of the type of program, behavior theories were found to be successful in changing mediator variables that affect behavior (Basen-Engquist et al., 1997; Kirby, 2001). These variables include intentions, beliefs, attitudes, skills, self-efficacy, and AIDS knowledge (Basen-Engquist & Parcel, 1992; CDC, 1999; Jemmott & Jemmott, 2000; Kirby et al., 1994). The outcome of implementing different strategies has been 25

PAGE 37

positive changes in behaviors related to abstinence and sexual intercourse. Two important mediator variables common in many behavioral change theories are knowledge and attitudes. The discussion of these variables is important, as they are the targets for change by many HIV/AIDS interventions with adolescents. Knowledge of HIV/AIDS General knowledge and functional HIV knowledge are two distinct components found in prevention programs. General knowledge includes biological information about the human immune response, characteristics of the virus, and human development. This type of information is ineffective in changing behavior (CDC, 2002b). In contrast, functional knowledge is essential in preventing the transmission of HIV (CDC, 2002b; Main et al., 1994). Functional knowledge informs students about the consequences of touching semen, vaginal fluids or blood (CDC, 1988, 2002b, 2003b; Quackenbush, Clark, & Nelson, 1995). Additionally, functional knowledge includes information about transmission of HIV by a mother to her child, risks associated with drug use, and misconceptions or myths about HIV transmission (CDC, 1988, 2002b). Functional knowledge is an essential component of many programs (Kirby, 2001; Quackenbush et al., 1995) and has been associated with positive outcomes. A meta-analysis of school-based sexuality programs by Song, Pruitt, McNamara, and Colwell (2000) revealed that the curriculum increased students sexual knowledge. They subdivided knowledge into six categories: general sexual knowledge, pregnancy, family life, contraception, HIV/AIDS, and STDs. The HIV/AIDS, pregnancy, and contraception effect sizes were 0.39, 0.26, and 0.18, respectively (Song et al., 2000). The results 26

PAGE 38

suggest that the programs effectively increased functional knowledge and provide support for their inclusion in prevention programs. Choi and Coatess (1994) analysis support the inclusion of a knowledge component; they reported that school-based programs increased HIV/AIDS knowledge. Additional support was provided by Kim, Stanton, Li, Dickerson, and Galbraithss (1997) analysis of 40 AIDS prevention interventions. They reported that 88% of the studies targeting adolescents had a statistically significant change in knowledge. Other studies demonstrated that an increase in knowledge was also associated with unexpected positive effects. For example, an increase in AIDS and condom knowledge among students in the age group 14 to 17 was associated with an increase in discussions of sexual topics between adolescents and their parents (Miller, 2001). Likewise, Siegel, DiClemente, Durbin, Krasnovsky, and Saliba (1995) reported a significant (p < .0001) correlation between AIDS knowledge and tolerance of those who have AIDS. Additionally, the objectives in most programs had to be satisfied through a knowledge component. An objective in some programs has been to clarify misconceptions and myths about HIV transmission. Myths are promoted by friends, which confuse nave teen-agers and lead to unsafe behavior. Other objectives were to stop teen-age students from having intercourse, to delay sexual activity among virgins, and to use effective contraception if involved in sexual intercourse. A knowledge component served to inform and thus protect students by equipping them with the correct and necessary information. In summary, increasing functional knowledge related to HIV/AIDS has been an objective in prevention programs, and the evidence indicates a positive association with 27

PAGE 39

successful outcomes. The role of knowledge is important because of its associated benefits. However, since the benefits did not always translate to changes in protective behaviors, other factors were also identified as important mediators (Kirby, Korpi, Adivi, & Weissman, 1997). Modification of Attitudes Since the demarcation of the attitude construct in the 1930s, there have been many explanations and experiments testing its psychological properties (Ajzen & Fishbein, 1980; Mueller, 1986). Louis Thurstone defined attitude as the affect for or against a psychological object (cited in Muller, 1986, p. 3). Ever since, investigations to determine the relationship between attitude and behavior have advanced the understanding of the attitude construct. A predominant assumption was that attitude and behavior were directly related. However, an important finding was that changing an attitude was not always associated with a change in behavior, and this highlighted the difficulties inherent in predicting behavior (Greenwald, 1989). But the benefits of modifying attitudes, according to Beckler and Wiggins (1989), were their effect on four basic human functions. Accordingly, these principles can be useful in the prevention of HIV/AIDS. The first function was to guide behavior toward valued goals and away from aversive events (Beckler & Wiggins, 1989). The disease AIDS is an aversive event whose negative consequences can be the focal point needed to modify attitudes toward the threat of HIV/AIDS. The second function included a cognitive component that allowed an individual to manage and simplify informational processing tasks. This feature allows persons to process relevant HIV/AIDS information in a manner that allows 28

PAGE 40

them to improve a particular attitude. The third function was related to how individuals communicate information about their personality and values. The fourth function protected people from unacceptable or threatening thoughts, urges, and impulses. The previous two functions were seen in many HIV prevention programs. Therefore, based on these four functions, attitudes related to sexuality, abstinence, and HIV infection can be modified and should be the focus of well-designed interventions to alter behavior. The common denominator of behavioral theories is their attempt to change behaviors by modifying attitudes (Ajzen & Fishbein, 1980; Kirby et al., 1991). As emphasized in the Theory of Reasoned Action, attitudes influence intentions that then affect behavior (Ajzen & Fishbein, 1980). This logic has been applied in studies that modified attitudes to change HIV/AIDS risk behaviors and increase condom use (Albarracin, Johnson, Fishbein, & Muellerleile, 2001; Kirby, 2001; Kirby et al., 1991; Main et al., 1994). The behaviors that need to be modified include delaying sexual intercourse, reducing the number of sexual partners, increasing the use of condoms, and eliminating the use of drug needles. These behaviors were commonly addressed by theory-based programs that targeted adolescents attitudes toward abstinence, condom use, peers, and HIV infection. The modification of attitudes has been associated with positive preventive behaviors among adolescents (Kirby, 2001; Main et al., 1994). For example, Main et al. (1994) made use of components of the Theory of Reasoned Action to change the determinants of risky behavior and found significant changes in adolescents intentions to engage in safer sexual practices. Attitudes, an important element of the program, were changed in a positive direction. Of particular importance, attitudes about the vulnerability 29

PAGE 41

to HIV infection (p < .04) reached significance. In contrast, attitudes toward condoms and talking to a partner about sex were altered and almost reached significance (Main et al., 1994). Students in the intervention group reported an increased use of condoms and fewer sexual partners than the comparison group (Main et al., 1994). The causality of attitudes was not established, but their role in changing behavior and intentions cannot be ruled out. Likewise, Jemmott, Jemmott, and Fongs (1992) study of inner city African-American adolescents utilized videos, games, and exercises to change attitudes about risky sexual behavior. Analyses of covariance determined that attitudes to engage in healthy sexual behaviors were significantly different (p < .007) than the control. They determined that intentions and attitudes (r = .70 and r = .68, respectively) toward risky sexual behaviors may serve as predictors of future behavior in African-American adolescents (Jemmott et al., 1992). Thus, an intervention that changed attitudes and increased knowledge was responsible for the increase in use of condoms and a decrease in the frequency of sexual intercourse. Kirbys (1991) research with adolescents provided evidence of the role of attitudes and norms in their decision to delay intercourse. He found a significant difference in the intervention group with regard to teen-agers perceptions of their peers involvement in intercourse and norms about unprotected sex (p < .001 and p < .05, respectively). Although the findings were not statistically significant, the advantage of changing the previously mentioned attitudes was important because their protective effect was evident after 18 months. But initiation of sexual intercourse, one of the primary objectives of the program, was statistically significant after the 18-month follow-up. 30

PAGE 42

Kirbys (1991) findings support the contention that adolescents are inclined to engage in sexual relations when they perceive their peers to be sexually active. ODonnell, Myint-U, ODonnell, and Stueve (2003) provided evidence to support Kirbys (1991) observations about the effect of peer-pressure. They reported that higher scores on peer norms and outcomes about sexual relations were associated with a delay of sexual intercourse among students that were between 12 and 16 years of age. Basen-Engquist et al. (2001) examined the effect of a multi-component HIV, STD and pregnancy prevention program on attitudinal and psychosocial variables in high school students. They reported that the program decreased the frequency of sexual intercourse without a condom and affected psychosocial variables in a positive direction. Specifically, attitudes toward condoms changed and were nearly statistically significant (p = .09). The attitudinal changes were associated with more condom use and a reduction in the number of sexual partners. However, a notable finding was that the number of students having sexual intercourse was not affected by the intervention. This highlights the challenges faced by all program planners and underscores the need to delay sexual involvement among younger adolescents. In the present study, attitudes and knowledge about HIV/AIDS were included in the instrument because of their role in the prevention of HIV among adolescents. These variables were viewed as important elements that mediated behaviors (CDC, 1988; Jemmott & Jemmott, 2000). The objectives of the Knowledge component included information about the behaviors that place students at risk for HIV, the importance of blood and body fluids as infection sources, the correct preventive behaviors, and the identification of misconceptions about risks. The objectives of the attitudinal component 31

PAGE 43

included attitudes toward peers, abstinence, drug use, threat of infection, and condom use. The information derived from the evaluation of educational programs can be useful to program planners and has implications for educational programs. The results can be used to write specific objectives, select the correct pedagogical strategies, and establish the outcomes for an evaluation. More importantly, the results of an evaluation can determine if the HIV/AIDS content has to be adjusted to the target population. Implications for educational programs include selecting personnel, preparing teaching materials, ordering classroom supplies, and training teachers. Additionally, the information was also essential to a process evaluation that determines how the program is being implemented. In summary, the inclusion of knowledge and attitudes into behavioral change theories has been shown to increase the effectiveness of interventions that attempt to reduce risky behaviors. Because of the potential benefits, the transfer of this information to other countries is appropriate and should be integrated into the curriculum of schools. Therefore, the design of state-of-the-art programs with ambitious objectives must be followed by an evaluation. Unfortunately, the lack of rigorously validated instruments to assess HIV education in Spanish-speaking countries is a major obstacle facing evaluators today in Latin America. Development of the CDC HIV/AIDS Instrument The CDCs leadership role in preventing the transmission of HIV is evident by the quality of programs, research, and publications that are disseminated to the American public. They have published guidelines for HIV school-based education (CDC, 1988, 32

PAGE 44

2003b), compiled a list of effective programs for adolescents (CDC, 1999), and developed an instrument to evaluate HIV education for all grade levels (CDC, 2002b). The instrument to evaluate school-based programs can be obtained through a published handbook (CDC, 2002b), as well as from the CDCs website. The developers of the instrument, IOX Assessment Associates and personnel from the CDC, designed it to assess seven different constructs related to HIV/AIDS education in schools. The constructs included: Knowledge, Beliefs, Attitudes, Student Confidence, Friends Views, Intentions, and Behavior (CDC, 2002b). These areas are common in most curricula taught in middle and high schools, and they are theoretically important in preventing the transmission of HIV. Additionally, the instrument is designed to assess HIV education for students in grades 7 to 12. They report that content validity was established with panels of experts and by personnel in the CDCs Division of Adolescent and School Health (CDC, 2002b). A national developmental review panel and a national advisory panel of experts, who advise the CDC on matters related to AIDS, were responsible for the appraisal and revision of the content. The methodology is not uncommon. Aday (1996) reports that such a review by an expert panel is a valid method to ascertain the content of an instrument. Basen-Engquist et al. (1999) and Kirby et al. (1994) developed an instrument to evaluate an HIV intervention by relying on qualitative reviews by a panel of HIV professionals and reported the method yielded valid and reliable results. The CDC (2002b) reported that adjustments in the HIV/AIDS instrument were made for reading level, understanding of items, and directions until they were appropriate for adolescents in the United States. The instrument was field-tested with students from 33

PAGE 45

the target population and subsequently edited. This process was repeated until the content was found to be adequate for adolescents. Next, a panel of HIV/AIDS experts conducted a series of reviews and revised the instrument to ensure that the content was valid. The CDCs instrument is a useful resource for evaluators as it serves as an item bank. The item bank provided the flexibility to design an evaluation to meet the specific objectives of a particular curriculum. Selecting appropriate questions to meet school board approval is another advantage, especially in light of frequent refusals to approve questions about sexual behaviors. While the instrument created for this study is derived from the CDCs items, it is considered a new instrument because the items will be translated and cross-culturally adapted for use in El Salvador. Therefore, the reliability and validity of the scores are important psychometric measures to be evaluated for this population in El Salvador. Content and Construct Validity If the interpretation of the scores from the Salvadorian high school students is to be made with confidence, it is necessary to establish the content validity of the items and the construct validity of the attitudinal subscales. Validity is defined by Gall et al. (1996) as the appropriateness, meaningfulness, and usefulness made from test scores (p. 773). The intent of the HIV/AIDS Knowledge component was to quantify students knowledge regarding transmission of HIV, risk behaviors, and misconceptions. The intent of the attitudinal component was to assess adolescents favorable or unfavorable evaluations toward five different attitudinal constructs. Content validity. Content validity is defined as the degree to which the scores yielded by a test adequately represent the content, or conceptual domain, that these scores 34

PAGE 46

purport to measure (Gall et al., 1996, p. 250). One method to establish content validity is to ask professionals who work in a particular area to generate items that are important or representative of a domain (Aday, 1988). Another method, recommended by Crocker and Algina (1986), is to define the domain, select a panel, and obtain items by providing specific instructions of the process and the domain. Afterwards, all the items are collected, sorted, and analyzed for their relevancy toward the domain. Mueller (1986) cautioned that content validity cannot be measured with a statistical index, and, therefore, the process requires documentation. Documentation for the present study was provided by panelists who were familiar with both the HIV/AIDS domain and the characteristics of the target population. Construct validity. Construct validity is defined as the extent to which a particular test can be shown to assess the construct that it purports to measure (Gall et al., 1996, p. 249). Construct validity is more difficult to establish than content validity. There are several conditions that facilitate the construct validation process. The first condition is to operationally define the construct. For example, if the attitudinal domain is to be validated, it would have to be defined. As a psychological construct, it was defined by Mueller (1986) [as] an idealized abstraction that is subject to scientific study and can be measured through inferences about peoples beliefs and behaviors. Items are then written to represent the definition of the construct that is to be assessed. Another condition is to establish a specific relationship to other constructs. Hypotheses about the construct are then formulated that specify the nature of the association between other psychological constructs (Mueller, 1986). Predictions about the variables role within the construct are formulated, and afterwards the results provide the 35

PAGE 47

necessary information about the similarity or dissimilarity between constructs. Quantifying the relationships among constructs is essential to this type of validation. In particular, the process relies on previous studies that provided evidence to support or renounce the construct. Included among the procedures to establish construct validity are the differentiation between groups, factor analysis, and the multitrait-multimethod matrix method (Crocker & Algina, 1986). Factor analysis has a long history as an acceptable method to establish construct validity (Thompson & Daniel, 1996). Confirmatory factor analysis (CFA). Stevens (1996) and Thompson and Daniel (1996) have noted that confirmatory factor analysis (CFA) is currently being used extensively to establish construct validity. Factor analysis is a method that determines the internal structure of an instrument. Stevens (1996) has stated that if the theoretical evidence underlying the construct is strong or if factors are specified a priori, CFA is advantageous. The HIV/AIDS instrument designed by the CDC, as an example, was purported to represent five attitudinal factors. Additionally, theoretical evidence indicating the usefulness of the attitudinal construct and other psychosocial factors in prevention programs have been established in other studies (Basen-Engquist et al., 1999; Main et al., 1994). These two considerations favored assessing the factorial validity of the CDC's attitudinal section of the instrument with CFA. The CFA is a powerful method that is used to test and confirm hypotheses regarding a theoretical construct; it provides statistical information through an analysis of the observed variables (Byrne, 1994). The researcher restricts the instruments observed variables or items to specifically load on only one latent or unobserved factor (Byrne, 1994). The relationship of the items to the assigned latent variable is then evaluated. The 36

PAGE 48

analysis determines if a model is acceptable by determining the adequacy of its goodness of fit to the sample data (Byrne, 1994, p.6). Programs that perform a CFA are LISREL, EQS, and Proc Calis (Byrne, 1994; Hatcher, 1994; SAS Institute, 2004). The Proc Calis program by the SAS Institute (2004) requires equations similar to the ones written for the LISREL program for the analysis of data with latent variables (Hatcher, 1994). The predetermined factors serve as the basis for the structural equations that account for the association of the variables to the factors. The covariance matrix of the items and the maximum likelihood method are two important properties of the Proc Calis program for CFA (Hatcher, 1994). Further, the model is specified to include the observed variables as the dependent variable. The latent variables and the errors in the model are the independent variables. The independent and dependent variables are placed in the equations according to a predetermined model; afterwards the model is tested. To determine if the model fits the data, the program makes use of fit indices, such as the 2 root mean square error of approximation (RMSEA), Bentler and Bonetts Non-normed Index, and Bentlers comparative fit index. These indices serve to determine if the model fits the data (Hatcher, 1994). The CFA also reveals the correlation among the latent variables, provides factor loadings, and determines the magnitude of the errors associated with the items. The CFA is a powerful tool to establish construct validity. Reliability A common definition for reliability is that the instrument consistently measures the intended construct. Reliability is viewed differently by the proponents of Classical Test Theory and Modern Test Theory. The former theory is relevant to this study and is 37

PAGE 49

comprehensively explained by Crocker and Algina (1986). Briefly, the classical model assumes that the scores from an instrument are represented by the equation X = T + E (Crocker & Algina, 1986, p. 107), where X is the score from the instrument, T is the true score, and E is a random error. The Classical theory asserts that a reliability coefficient can be calculated from either a single administration of an instrument or from two different administrations. When a single administration is given, the internal consistency of the scores is measured with Cronbachs alpha (Crocker & Algina, 1986). Crocker and Alginas (1986) interpretation of alpha is that it is the percentage of the total score variance that is due to true score variance (p. 139). Another common interpretation of coefficient alpha was that it describes how the items are functioning together as a unit. The second method for estimating reliability is the test-retest method. The same instrument is given to the same group of persons at two different times. A Pearson product-moment correlation coefficient (coefficient of stability) is calculated between the two different scores. The interpretation of this coefficient is different from that of coefficient alpha. The test-retest provides information of the measurement errors that result from the fluctuations of an examinees observed scores around the true score because of temporary changes in the examinees state (Crocker & Algina, 1986, p. 133). The source of error is usually from either the administration of the instrument, scoring, guessing, incorrectly marking the instrument, and temporary changes in attitude (Crocker & Algina, 1986; Mueller, 1986). Thompson and Vacha-Haase (2000) noted that reliability is derived from the scores of a specific population and is not a property of the instrument. They stated that reliability should be reported as a measure associated with the scores. Their observation 38

PAGE 50

was important because it placed the emphasis on the scores and the target population. In other words, the fact that scores are reliable for one group does not guarantee that the scores will be reliable for other groups. Cross-Cultural Adaptation The different guidelines. Researchers who need an instrument to evaluate programs typically have to develop a new instrument or adapt a pre-existing one (Guillemin et al., 1993). The first option is challenging as the methodological procedure and needed resources increase the development time and cost (Guillemin et al., 1993). The second option is to adapt an existing survey to a new population; under the right economic conditions and scientific evidence, this alternative is favorable over developing a new instrument. The United States and other Western nations dominate the scientific literature, which facilitates the adaptation of instruments to countries with limited resources. An important issue has been the elimination of cultural bias inherent in the construction of the original instrument as it affects the interpretation of the items in a new population (Vijver & Leung, 1997). In particular, the challenge is to obtain two instruments that are proven to be equivalent at the conceptual, operational, item, and scalar level (Hui & Triandis, 1985). However, the methodology of a cross-cultural adaptation is dependent on the purpose of the instrument. When the purpose is complex, such as comparing the intelligence or scholastic aptitude of individuals from two different cultures, the process according to Herdman et al. (1998) requires a complete and complex investigation of how the two cultures perceive the construct. 39

PAGE 51

In contrast, when the purpose is less complex and the construct has universal qualities, as in the case of providing an instrument to assess HIV/AIDS knowledge and attitudes in a different country, the cross-cultural adaptation is less demanding. Favorable conditions must also exist for this type of adaptation. One condition is the recognition within both cultures of a construct with a universal aspect (Hui & Triandis, 1985). For example, the risk factors responsible for the transmission of HIV have been established worldwide and have a universal quality. Additionally, adaptations are advantageous when the construct is concrete (Hui & Triandis, 1985), and the origin of the language is similar. As an example, an instrument in English will theoretically be uncomplicated when it is adapted to Spanish, and more complicated if adapted to Chinese or Swahili. The methodological and measurement characteristics of cross-cultural adaptations have been established through a variety of studies including those that have attempted to compare the health status and quality of life among different countries (Schmidt & Bullinger, 2003). In these studies three types of methods have been used. The first type of method is the simplest and is illustrated in studies by Daza, Novy, Stanley, and Averill (2002), Ruiz, Berrocal, Lopez, and Rivas (2002), and Davis, Tang, Chang, and Noel (1999). The second type is represented by the guidelines proposed by Guillemin et al. (1993). Finally the third type is reflected in the guidelines proposed by the IQOLA group (Bullinger et al., 1998), Herdman et al. (1998), and the WHOQOL group (Power, Harper, & Bullinger, 1999; Skevington, 2002). This third type is the most complex and combines qualitative and quantitative techniques to cross-culturally adapt and obtain an equivalent instrument. 40

PAGE 52

A comparison of the different methods indicates that certain elements are essential to the process. In particular, the guidelines recommended by Guillemin et al. (1993) and the IQOLA group (Bullinger et al., 1998) differ, but they contain elements that complement each other. These elements include the back-translation method, analysis of the translations by more than one person, synthesis of several translations, cultural review of items, pretesting, and validation of the scores. Since the purpose of this study was to obtain a new instrument for use in El Salvador, the method used in this study will rely mostly on Guillemins guidelines to adapt an instrument, and it will be complemented with aspects recommended by Bullinger et al. (1998) and Gandek and Ware (1998). Guillemins guidelines. According to Guillemin et al. (1993) a cross-cultural adaptation is divided into two components: translation and adaptation. The first involves a change from one language to another to obtain a literal meaning. The instrument must be forward-translated by at least two different translators. The forward-translations should retain the same meaning. Afterwards, the translations should be reviewed by a committee. The use of qualified individuals at all stages is essential and directly affects the quality of the final product. The back-translation involves comparing the instrument to the original version. This key step should involve individuals who are not aware of the original instrument. According to Guillemin et al. (1993), their unfamiliarity will reduce bias, and a review by different committees will improve the quality of the cross-cultural adaptation. The adaptation phase is the process in which the words from the first language have to match the semantic, idiomatic, cultural context, and lifestyle of the target 41

PAGE 53

population. In parallel with the first phase, Guillemin et al. (1993) recommend an evaluation by a multi-disciplinary committee and emphasized the use of bilingual reviewers who are familiar with the culture of the target population. The first step is to establish semantic equivalence. This involves establishing the same meaning as the original items and matching the composition of the sentences in both instruments (Guillemin et al., 1993). To ensure the same meaning, the translated item is written with the same grammatical structure and it should have a similar vocabulary. In addition to semantic equivalence, the adaptation attempts to achieve idiomatic equivalence. Idiomatic equivalence refers to the translation of idioms and colloquialisms into expressions understood by the target culture. The interpretation of colloquialisms may affect how students respond to the items. Finally, the adaptation phase attempts to achieve conceptual equivalence. Conceptual equivalence relates to the legitimacy of the concept in the target population. The concept must be relevant and understood according to the life experiences of the persons in the new culture (Guillemin et al., 1993). For example the word cousin can refer to a firstor a second-generation family member in some cultures. The interpretation of the word has to be the same for every person--either a first or a second-generation member, but it cannot have both. Conceptual equivalence assures that the meaning and interpretation of the items are the same for all of the persons in the sample. Guillemin et al. (1993) noted that the process yielded a cross-cultural equivalent between the source and the adapted version. The final step in the cross-cultural adaptation is to pretest the instrument with the target population and to identify problematic items using the probing technique. The advantage of the probing technique is that it uncovers items that are not culturally 42

PAGE 54

relevant. The guidelines recommended by Guillemin et al. (1993) provide a sequential process that aims to produce a quality instrument. However, a drawback of their method is the reliance on qualitative methods and the lack of quantitative strategies to confirm the adaptation. Guidelines from the IQOLA. The International Quality of Life Assessment (IQOLA) is a complex, multi-national project that created a cross-cultural equivalent of an instrument to be used in more than 10 different counties (Bullinger et al., 1998; Gandek & Ware, 1998). The method used to create this cross-culturally equivalent instrument was comprehensive, involving many professionals from different nations and large samples from different countries. The process was implemented over a 10-year period, and the method consisted of three steps. Briefly, the first step involved numerous forwardand back-translations with qualitative and quantitative measures (Bullinger et al., 1998). This initial phase culturally adapted the content and used complex procedures to establish the conceptual equivalence of the instrument. This phase identified difficulties related to the translation, eliminated misunderstood items, and revised or eliminated problematic items. The second step established content validity and construct validity of the scores from different countries. Content validity determines how adequate the sample of items in the instrument reflects the larger domain. In the IQOLA project, content validity was established by comparing the instrument to a known standard (Gandek & Ware, 1998). Further, construct validity relied on evidence related to the variables in the construct, the internal structure, and the analysis of the theoretical relationship between scale scores and external criteria (Gandek & Ware, 1998, pp. 953-954). The internal structure was 43

PAGE 55

established by principal components factor analysis and structural equation modeling. The results provided evidence of the theorized structure and served as a comparison across countries (Keller et al., 1998; Ware et al., 1998). The goal of the IQOLA project was to reproduce the conceptual model of the health survey across different populations, and their results showed that the dimensions were consistent with those of the original instrument (Keller et al., 1998). The discovery of having the same construct among the different populations supported their claim that the translation and adaptation methods were appropriate. Additionally, they noted that the interpretations of the scores were the same in each country. The success of the adaptation and ensuing method to establish a cross-cultural equivalent of the IQOLA to more than 10 countries in Europe proved the robustness of their method. The third step in the IQOLA process was to obtain normative data with representative national samples (Bullinger et al., 1998). Gandek and Ware (1998) stated that normative data make it possible to interpret the scale score for an individual respondent or a group of respondents in comparison to the distribution of scores for the norming sample (p. 995). A drawback of the methodology developed by the IQOLA group was the amount of resources, length of time, and qualified personnel involved at each stage. The benefits exceed the shortcomings, and the methodologys success has been demonstrated through many studies (Bullinger et al., 1998; Gandek et al., 1998; Keller et al., 1998). Synthesis of Cross-Cultural Adaptation Methods A synthesis of the common components to adapt an instrument is an alternative to following one single method. The components include the back-translation method, 44

PAGE 56

analysis of the translations by more than one person, synthesis of several translations, cultural review of items, pretesting, and validation of the scores. Additionally, the use of expert and lay panels, bilingual reviewers, and expert translators is essential for the process. The findings reported by Guillemin et al. (1993) and Bullinger et al. (1998) are in agreement with Brislin et al. (1970), who found back-translation to be a valid method. The procedure requires several forward-translations of the instrument by experienced translators and several reviews of the translations by others. A back-translation to the original language then serves as a comparison between the meanings of the two languages. A literal meaning should be retained for the translation to be successful. However, the number of forward-translations to obtain a quality translation varied from five to one translation. The availability of qualified translators, expert reviewers, and funds were resources that affected the number of translations. Studies by Daza et al. (2002) and Ruiz et al. (2002) used one or two back-translations followed by a review of the instrument by two or three individuals familiar with the research. Both reported successful results. Based on the previous results, one or two back-translations can be acceptable for an adaptation. The role of language usage is an essential step in a cross-cultural adaptation. The method has to systematically eliminate problematic items that are inappropriate or culturally irrelevant. Guillemin et al. (1993) qualitatively evaluated and adjusted items for semantic, idiomatic, and conceptual equivalence. In contrast, Bullinger et al. (1998) quantitatively evaluated the items according to translation difficulty, clarity, common language, and conceptual equivalence. Since language is the primary form of 45

PAGE 57

communication, this step is crucial. The simplicity of Guillemins analysis of language usage is more feasible than the others. To determine if the items are understood as intended, a pretest with a sample from the target population is necessary. The instrument is tested with a small group and then analyzed for understanding. Bullinger et al. (1998) recommended the cognitive debriefing approach where participants were asked about the items difficulty and had to identify items that upset them or that were confusing, whereas Guillemin et al. (1993) advocated the probing technique. The straightforwardness of the probing technique makes better use of time and is appropriate for high school students. The intent is to solicit an open-ended response to compare the true meaning of the problematic item with what was implied or not understood. The protocol recommended by Guillemin et al. (1993) is to ask the question: What does this item mean? The students responses are then compared to the meaning of the item and the process is useful in that it identifies inappropriate or problematic items. A review by an expert and lay panel allows the introduction of culturally relevant material that is unfamiliar to researchers. Therefore, the addition of panels improves the results of the adaptation by filtering out culturally irrelevant material. The guidelines from the IQOLA recommended calculating the reliability of the scores and establishing content and construct validity. Reliability can be determined by the test-retest method and by calculating the internal consistency of the instrument. Content validity can be determined by the use of an expert panel that compares the items in the instrument against those from an established domain. Establishing content validity 46

PAGE 58

is essential because it provides assurance that the items represent the domain being assessed. Factor analysis has a dual role in a cross-cultural adaptation. The first is that it can be used for evaluating construct validity. The second is that it serves to verify the success of the adaptation. A successful adaptation will produce conceptually equivalent instruments, and theoretically they will yield the same factor structure. Keller et al. (1998) explained that they obtained the same factor structure across 10 different countries because the constructs were equally interpreted in each country. Factor analysis is therefore a very useful tool for ascertaining a cross-cultural adaptation. Spanish Language Adaptation and Factorial Validation Hui and Triandis (1985) assert that cross-cultural adaptations are straightforward when languages are similar and the construct is clearly defined. Studies by Daza et al. (2002) and Ruiz et al. (2002) provide an example that supports their opinion in regard to the use of an instrument written in English and adapted for use in Spain. Daza et al. (2000) adapted and validated an instrument with less difficulty when Hui and Triandiss principles were satisfied. In contrast, Ruiz et al. (2002) had a more difficult and less clearly defined construct that yielded an adapted instrument that did not retain the factorial dimensions found in the original. These findings raise questions about the translation, previous validation, and the difficult nature of a cross-cultural adaptation. Daza et al. (2002) compared the English and Spanish translations of the Depression Anxiety Stress Scale-21 and established the Spanish instruments validity with confirmatory factor analysis (CFA). Daza et al. (2002) examined the factorial dimension and internal consistency of the two instruments. In particular, if the adaptation 47

PAGE 59

was successful, they expected the internal consistency, convergent validity, and factor structure of the English and Spanish instrument to be the same. They found the convergent and discriminant validity of the translated instrument to be statistically significant and similar to the English version. Daza et al. (2002) reported that the internal consistency was adequate for the original and the adapted version. Using CFA, they determined the adapted version had the same factor structure as the original, and an important finding was that the factor loadings were significant for a three-factor model (p < .01). They concluded that the three-factor structure consisting of depression, anxiety, and stress was similar to the original constructs. Based on this result, they were able to ascertain the construct validity of the adapted instrument. The replication of the same factor structure provided evidence that supported the success of the methodology used for translation and adaptation. The use of a previously validated instrument that had many items that were tested in different populations appeared to facilitate Daza et al.'s (2002) adaptation. Daza et al. (2002) utilized the back-translation method that yielded a quality translation. Their use of a panel of bilingual Hispanics to establish cultural relevance was advantageous because the panels culture was similar to the sample population, and they were able to make relevant contributions that increased the quality of the items. A limitation of the study was the small sample size (n= 98) that was used in CFA; a much larger size has been recommended for CFA. In contrast, a different study by Ruiz et al. (2002) translated the English version of the Weight Efficacy Life-Style questionnaire into Spanish for use in Spain. Ruiz et al.'s (2002) procedure included two back-translations and a review by two professors. Unlike 48

PAGE 60

Daza et al. (2002), they did not report the use of a panel or a pretest, and it was likely that these components may have improved their results. Ruiz et al. (2002) reported that both versions of the instrument had an adequate measure of reliability. However, the factorial dimension of the two instruments was not similar, thus questioning the effectiveness of the translation and the adaptation. In particular, the items interpretation was questioned because they were not tested with a sub sample. Ruiz et al. (2002) tested a five-factor model with CFA (n = 345), and they reported that the Spanish version did not retain the five original factors. Subsequently, items were deleted and retested, yet the analysis did not produce the original five factors. To determine the best model that fit the data, another CFA (three-oblique factor model) was performed, and this yielded a three-factor model with an acceptable goodness of fit index. The primary reason for not attaining the factorial structure of the self-efficacy construct in Ruiz et al.'s adapted version was the complexity of the construct, and the secondary was due to the similarity among the items. Ruiz et al. (2002) reported that the constructs were correlated and that the items loaded on several factors. Apparently, multi-dimensionality of the items affects the results. The overall results of the previous study suggested that the translation of certain items might have contributed to the three-factor structure. According to Ruiz et al. (2002), certain items lacked conceptual equivalence after the translation, and social differences among the cultures affected how some items were interpreted. Therefore, the items of an instrument must be conceptually clear to prevent an ambiguous interpretation, and all of the items must be translated properly. A shortcoming of the study was the lack of a pretest that could have identified the cultural acceptability of the items for use in 49

PAGE 61

Spain. Guillemin et al. (1993) recommended the use of a panel or a pretest to identify cultural bias and ambiguous items. In summary, the use of a condensed methodology by Ruiz et al. (2002) and the complexity of their construct affected the results of the adaptation. Adaptation of an HIV/AIDS Instrument in Hong Kong Davis et al. (1999) reported on a procedure for translating and adapting an English version of the International AIDS Questionnaire into Chinese for use with adolescents in Hong Kong. Unlike the methods recommended by Bullinger et al. (1988) and Guillemin et al. (1993), this procedure was brief. The authors stated that the original instrument was comprised of items from other instruments. One of the instruments was developed in a Western nation and the other was an instrument designed specifically for a Chinese population. The original instrument consisted of 35 items. The items were back-translated and reviewed by bilingual researchers for face validity (Davis et al., 1999). According to Davis et al. (1999), the instrument was field-ready after this short procedure. A total of 863 Chinese students answered the instrument and a CFA determined that the translated version had the same four-factor structure as the English version. They reported eliminating 17 items to obtain a final instrument with 18 items. The overall reliability (Cronbach alpha) of the scores was .73 and the following subscales--Transmission myths, Attitudes, Facts, and Personal Risk--were reported to have an alpha of .73, .66, .45, and .48, respectively. The elimination of items to raise the reliability of the scores and to obtain a four-factor model raised questions about the adaptation. This issue was not addressed by the authors. However, Skevington (2002) cautioned that the reliability of an adapted instrument is lower than that of the original instrument. This 50

PAGE 62

observation is similar to Gullemin et al. (1993) in that they believe that adapted instruments are similar to a new instrument, and the psychometric properties may not be the same as those of the original instrument. Summary of Literature Review The literature review discussed the epidemiology of HIV/AIDS in El Salvador and listed the pertinent aspects of the HIV curriculum taught in the public high schools of San Salvador. The risk factors that placed adolescents at risk of infection, with a particular emphasis on sexual behavior, were discussed because they increase the risk of transmission of HIV. In particular, the discussion focused on the effectiveness of schooland community-based educational programs with curricula that integrate components of behavioral theories to reduce the risk factors associated with the transmission of HIV. Two important components, HIV/AIDS knowledge and attitudes, were discussed because they are mediating variables in behavioral theories. Knowledge related to transmission, behavioral factors, and myths have been associated with effective programs. However, other components must also be present to change behavior. The attitudinal component, a key psychosocial factor, mediates behaviors that increase the risk of HIV transmission (Basen-Engquist et al., 1999). Specifically, attitudes related to abstinence, condom use, risk of HIV, drugs/steroids, and peer-pressure have been associated with behavioral change (Basen-Engquist & Parcel, 1992; Kirby et al., 1991). Therefore, educational programs that contain these components have to be evaluated with instruments that provide reliable and valid scores. The theoretical evidence indicates that the cross-cultural adaptation of an existing instrument is one way to provide new tools for researchers in other parts of the world. 51

PAGE 63

Methodological issues related to an instruments translation and adaptation from one language to another were discussed. These issues include obtaining the correct literal translation by the back-translation method and establishing guidelines for cross-cultural adaptation. Measurement issues related to an adapted instruments reliability and validity were outlined. In particular, the content and construct validation of scores from an instrument was reviewed because these validation procedures provide a method to ascertain the success of the cross-cultural adaptation. Given this background, an instrument developed by the CDC was translated, adapted, and the scores validated for use in El Salvador. 52

PAGE 64

CHAPTER 3 METHODS Purpose of the Study The purpose of this multi-phased study was to translate, cross-culturally adapt, and validate the scores from an instrument that assessed the HIV/AIDS knowledge and attitudes of high school students in El Salvador. The first phase utilized the back-translation method to obtain the first Spanish version of the instrument, which was developed by the CDC to evaluate HIV/AIDS knowledge and attitudes of students in grades 7 to 12. In the next phase, an expert panel of reviewers from El Salvador established the conceptual equivalence of the Spanish instrument. The third phase evaluated the instruments content validity and cultural acceptability through a panel composed of HIV experts, health professionals, schoolteachers, and students from El Salvador. The last phase involved collecting evidence of the validity of the scores from the Spanish version of the CDC instrument; 483 public high school students in the metropolitan area of San Salvador participated in the validation phase. Reliability of the scores was measured by the test-retest method and coefficient alpha. Confirmatory factor analysis was used to evaluate if the attitudinal construct had five different dimensions (Peer-pressure, Condom use, Abstinence, Drug use, and Threat of HIV). As part of the concurrent validation process, the Condom and Abstinence subscales from the translated 53

PAGE 65

CDC instrument were correlated with a set of existing attitudinal items from a published study by Basen-Engquist et al. (1999). An outline of the research plan is presented in Figure 2. Phase 1: Translation of English Instrument into Spanish. Location: United States. a. Two forward-translations by native Spanish-speaking translators. b. Synthesis of first Spanish version by four bilingual individuals. c. Back-translation by native English speaker. d. Comparison of original and back-translated version. e. Two individuals will resolve discrepancies. Phase 2: Establish the Conceptual Equivalence by Comparing the English and Spanish Version Location: El Salvador. a. Bilingual review with five professionals from El Salvador. Evaluate for semantic, idiomatic, and conceptual equivalence. b. Lay panel review to edit directions, items, and answer choices. c. Principal investigator and Lay Panel will edit and identify problematic items. Phase 3: Readability, Content & Cultural Acceptability Review by Salvadorian Panel Location: El Salvador. a. Seven-member Salvadorian expert panel will determine the content and cultural acceptability. 1. Problematic items will be revised. 2. Summarize findings and report level of agreement. b. Determine reading level with two Salvadorian experts. c. Revise instrument as necessary with Lay Panel. Phase 4: Pretest of Spanish Instrument in El Salvador a. Pretest instrument with Salvadorian adolescents. Form two groups of four individuals (two male and two female students). Use probing technique to check for understanding, interpretation, ambiguous items, and cultural relevance. b. Edit instrument and pretest if necessary. c. Edit final instrument with the assistance of the Lay Panel. Phase 5: Administration of Instrument in El Salvador a. Randomly select 10 public high schools in San Salvador. Randomly select classrooms to obtain a representative sample of students in grades 10, 11, and 12. b. Provide permission slips to 500 students to obtain a sample of at least 300 students. c. Randomly select 20-50 students for a one-week test-retest reliability study. d. Descriptive statistics. e. Univariate analysis and item analysis. f. Multivariate (CFA) Statistics. g. Determine concurrent validity with the embedded items. Figure 2. Steps Outlining the Research Methodology of the Multi-phase Study. 54

PAGE 66

Specific Purposes 1. Determine the success of the cross-cultural adaptation by contrasting the first translated version of the CDC instrument with the final Spanish version of the CDC instrument. 2. Determine using a Salvadorian panel, the level of cultural appropriateness of the translated CDC HIV/AIDS knowledge and attitudinal items for use in El Salvador. 3. Ascertain the level of readability of the final Spanish version of the CDC instrument and evaluate content validity. 4. Evaluate the reliability of the scores from the Spanish version of the CDC instrument using a sample of Salvadorian students in grades 10, 11, and 12. 5. Determine if the Spanish version of the CDC instrument contains five attitudinal dimensions (Peer-pressure, Condom use, Abstinence, Drug use, and Threat of HIV) when analyzed by confirmatory factor analysis. 6. Evaluate the concurrent validity of the Abstinence and Condom use subscales from the Spanish version of the CDC instrument. Construction of the English Language Version of the CDCs Instrument The Centers for Disease Controls Assessment Instruments for Measuring Student Outcomes Grades 7-12, Booklet 6 (CDC, 2002b) was ordered from the CDCs website in May 2002 and served as a source for the knowledge and attitude items. The booklet serves as a question bank that can be modified to assess seven constructs related to HIV/AIDS education: Knowledge, Beliefs, Attitudes, Student Confidence, Friends Views, Intentions, and Behavior (CDC, 2002b). 55

PAGE 67

The booklet described the process used by IOX Assessment Associates and the Centers for Disease Control to develop the items. They employed five different review boards to systematically construct, test, and revise the instrument. The procedure included field-testing with small groups of students to confirm that directions, items, and vocabulary were appropriate for this age group. After revising the items, they were field-tested again with adolescents, and the items were revised until they were suitable. Further, several panels of HIV evaluation experts met to review the items to ascertain content validity. Therefore, the content and face validity were acceptable. The instrument for use in El Salvador was constructed by taking 38 items from the CDCs booklet and 6 items from a study by Basen-Engquist et al. (1999). These items were translated and cross-culturally adapted to assess HIV/AIDS knowledge and attitudes of public high school students in El Salvador. As described in Appendix A, age, grade, gender, and previous HIV education were included to assess demographic characteristics. HIV Knowledge The instrument assessed HIV functional knowledge with 15 items. The CDC (2002b) defines functional knowledge as information about the transmission of HIV, testing, risk behaviors, and prevention. Specifically, the 15 items assessed knowledge related to: (1) sexual contact with an infected person; (2) sharing needles with an infected person; and (3) myths about transmission. The information in Table 1 displays the 15 knowledge items from the CDCs item bank and provides the correct answer for each item. 56

PAGE 68

Table 1 Items Assessing HIV Knowledge Items Assessing Knowledge Source 1. You can't get AIDS if you have sex only once or twice without a condom. (F) CDC 2. A person can "pass" an HIV antibody test (test negative) but still be infected with HIV. (T) CDC 3. Condoms are 100% effective in preventing HIV. (F) CDC 4. Males can pass HIV on to others through their semen. (T) CDC 5. You can get HIV by sitting on the seat of a toilet that a person with AIDS has used. (F) CDC 6. Abstinence from sex and drugs is the best way for teen-agers to avoid getting HIV. (T) CDC 7. You can get HIV from drinking from the same glass or water fountain that a person with AIDS drank from. (F) CDC 8. HIV can be found in semen, vaginal fluids, and blood. (T) CDC 9. A person can get HIV by sharing drug needles. (T) CDC 10. HIV can be found in breast milk. (T) CDC 11. Once you are infected with HIV, you are infected for life. (T) CDC 12. People infected with HIV are usually very thin and sickly. (F) CDC 13. Some people have gotten HIV by swimming in the same pool as someone with AIDS. (F) CDC 14. You can get HIV from a mosquito bite. (F) CDC 15. Latex condoms can be used to prevent HIV/AIDS. (T) CDC, Reworded Note Items: 2, 4, 6, 8, 9, 10, 11, and 15 were true; 1, 3, 5, 7, 12, 13, and 14 were false. This instrument made use of a true and false scale with five options: I am sure it is true, I think it is true, I dont know, I think it is false, and I know it is false. 57

PAGE 69

Each question was worth one point; however, two responses were possible per statement. For example, if the correct answer is true, the following scale was used: I am sure it is true 1 point, I think it is true 1 point, I dont know 0 points, I think its false 0 points, I am sure its false 0 points (CDC, 2002b, p. 6). False statements were scored in reverse. The eight true and seven false statements balanced the instrument and thus avoided a response set. The inclusion of do not know" to the scales was intended to eliminate guessing and thus increase the validity of the scores. The scores ranged from 0 (all incorrect) to 15 (all correct), and a percentage was calculated for each individual. This type of scale has been used in the United States and Latin America. In the United States, DiClemente et al. (1986) developed a modified True/False/Do Not Know scale with 30 items to assess AIDS knowledge (reliability was not reported). Jemmott et al. (1992) used the scale developed by DiClemente et al. (1986) to assess HIV/AIDS knowledge in a population of adolescents and reported a coefficient alpha of .73. Likewise, the feasibility of using a modified True/False scale written in Spanish was demonstrated with Honduran adolescents (Fundacin Fomento en Salud, 2001). Therefore, the data indicated that the scores from a true and false scale with five options were reliable. HIV/AIDS Attitudes Researchers have established the reliability and validity of Likert scales for measuring HIV and AIDS-related attitudes. For example, Coyle et al. (1999) used a four-point scale and reported an internal consistency of .78 and .87 for scores from a population of high school students attitudes toward sexual intercourse and condoms, 58

PAGE 70

respectively. The items in Coyles study assessed two of the same constructs that were important to this study. Also, with a similar rating scale, both Carvajal et al. (1999) and Silva and Ross (2003) measured adolescents HIV attitudes in Latin America and reported reliability of .70 and .88 for the scores, respectively. Based on the previous evidence, Likert scales are appropriate to assess attitudes. The instrument used in this study is presented in Appendix A. The information in Table 2 shows the number of items by source and describes the construct represented by each item. Twenty-three items from the CDC item bank were used to assess five dimensions of the attitudinal HIV/AIDS construct: (1) Peer-pressure; (2) Abstinence; (3) Condom use; (4) Drugs use; and (5) the Threat of HIV infection (CDC, 2002b). Additionally, six items from an instrument developed by Basen-Engquist et al. (1999) to measure attitudes of adolescents were added as part of the concurrent validity study. Reliability and validity of the six items were established by Basen-Engquist et al. (1999). The reliability of the scores from a four-point scale was reported to be .78 for the Abstinence subscale and .87 for the Condom use subscale. The 23 items in the attitudinal component of the CDCs instrument have two characteristics that increase internal consistency. The first is the balance of positively and negatively worded statements. As suggested by Muller (1986), it lowers the chance of answering in a set response. The second is that the instrument covers the full domain of the attitude construct by including affective, behavioral, and cognitive statements (Muller, 1986). 59

PAGE 71

Table 2 Source of Attitudinal Items Source Items Dimension Centers for Disease Control and Prevention 23 Peer-pressure, Condom use, Abstinence, Drug use, and Threat of HIV Basen-Engquist et al. (1999) 6 Condom and Abstinence The attitudes were measured with a five-point Likert scale that ranged from strongly agree, agree, not sure, disagree, to strongly disagree and the responses were assigned values of 1 to 5, respectively. Negative statements (items 1, 3, 4, 6, 7, 10, 13, 15, 16, 18, 22, 23, and 26) were scored so that strongly agree = 1, agree = 2, not sure = 3, disagree = 4, and strongly disagree = 5. The scores for positive statements (items: 2, 5, 8, 9, 11, 12, 14, 17, 19, 20, 21, 24, 25, 27, 28, and 29) were reversed so that strongly agree = 5, agree = 4, not sure = 3, disagree = 2, and strongly disagree = 1. Higher scores indicated more favorable attitudes and lower scores unfavorable attitudes. Study 1 Forward-translation, Synthesis, and Back-translation of the Instrument Forward-translation The back-translation method is the process of translating an instrument from one language into another language and then translating it back to the original language. This allows the comparison between the original and its back-translation. According to Brislin et al. (1970), the method was valid for translating an instrument written in English and 60

PAGE 72

translated into Spanish. The purpose is to compare the written text and the meaning of the original with the back-translation. A critical factor of the process is the selection of qualified individuals who are proficient in two languages and who have expertise in the subject. Therefore, a challenge of this process is identifying competent translators, bilingual reviewers, native language speakers, and content matter experts because their inclusion will increase the quality of the final translation and affect the adaptation (Brislin et al., 1973; Guillemin et al., 1993; Hui & Triandis, 1985). The number of forward-translations that yields a quality translation has yet to be determined. Previous studies have used from one to five. Bullinger et al. (1998) demonstrated that at least three different translations were necessary for an acceptable result. In contrast, some obtained successful results with one and two translations (Becher et al., 1999; Davis et al., 1999). The feasibility of two forward-translations was demonstrated by Alonso et al. (1998) who obtained two forward-translations of an instrument written in English and adapted in Spain. Table 3 Qualifications of Translators Variable Translator 1 Translator 2 Back-Translator 1 Bilingual 1 Bilingual 2 Bilingual 3 Bilingual 4 Birth Place El Salvador Nicaragua USA El Salvador Colombia Puerto Rico USA Mother Tongue Spanish Spanish English Spanish Spanish Spanish English Educational Level Bachelors Bachelors Masters Masters Masters Bachelors Masters Experience 20 years 30 years 2 years 7 years 10 years 2 years 5 years 61

PAGE 73

Therefore, this study made two different translations by requesting two bilingual individuals to carry out a forward-translation. Their qualifications include a college education, Spanish mother tongue, fluency in English and Spanish, and a combined experience of 50 years translating documents (see Table 3). As demonstrated in Appendix B, the instruments directions, items, and answer choices were sent electronically to the translators in a form derived from a study by Beaton, Bombardier, Guillemin, and Ferraz (2002). Subsequently, two versions of the original instrument were independently translated and returned to the primary investigator via email. The translators documented their difficulties on the form and were interviewed to discuss problematic items. Synthesis The two forward-translations were merged by a committee of four bilingual individuals who had relevant Masters degrees and/or Spanish as the mother tongue (see Table 3). A form (adapted from Beaton et al. 2002) was provided to each reviewer. The form (see Appendix C) was sent electronically to Bilingual Reviewers 1, 2 and 4. Reviewer 3 participated in the last step and edited the final version. Bilingual Reviewers 1, 2, and 4 read the two forward-translations and the English version. Afterwards, they compared the two Spanish translations or made their own translation if needed and selected the item that retained the idea or meaning found in the English version. This process produced three translated versions. Bilingual Reviewers 1 and 4 synthesized these translations by selecting the best translation for each item. This was achieved by comparing the items, selecting the best meaning, and then agreeing on the best translation. Next, Microsoft Word XPs (2003) Spanish dictionary (Salvadorian 62

PAGE 74

version) corrected the spelling and placement of Spanish accents on the words. Confirmation of the synthesis was then made with Bilingual Reviewer 3 who reviewed the English version, two forward-translations, and the first Spanish version. Back-translation The first Spanish version was sent via email (see Appendix D) and was back-translated by an individual unfamiliar with the English version. The method was recommended by Brislin et al. (1973) and Schmidt and Bullinger (2003) and yielded a quality translation. Back-translator 1 had a Masters degree in English-Spanish language translation and her mother tongue was English. Discrepancies were listed and sent to the primary investigator via email. An interview with the translator was scheduled to resolve problematic items. The back-translated English version and the original version were compared to verify if the meaning of each item was the same. The translated items that did not match the original structure were reevaluated and rewritten. Study 2 Conceptual Equivalence by Salvadorian Expert Panel The objective of the adaptation is to have two instruments that are conceptually equivalent. For instruments that are translated and adapted from one language to another it involved the replication of the construct and the elimination of cultural biases that influence the construct's interpretation (Hui & Triandis, 1985; Vijver & Leung, 1997). A vital aspect of the process of establishing conceptual equivalence is to adjust the words to the cultural level of the target population and then to determine if they are relevant and understood as intended. It was expected that establishing a conceptually equivalent instrument would yield acceptable psychometric measures. 63

PAGE 75

According to Guillemin et al. (1993), the process includes establishing the semantic, idiomatic, and conceptual equivalence of an instrument that was adapted from one culture to another. The process systematically replicates the construct from the first culture to the target culture. Guillemin et al. (1993) suggested that a review by a multi-disciplinary committee of content matter professionals, linguists, and individuals from the target population is necessary to assure that the cultural content will be adequate. Leplege, Ecosse, Verdier, and Perneger (1998) and Bullinger et al. (1998) made use of a committee composed of experts and reported that it was an integral part of establishing conceptual equivalence. This study identified an expert panel of Salvadorian professionals by soliciting names from the Salvadorian government, non-governmental organizations, and a search of HIV-related websites in El Salvador. After the initial recruitment, the professionals were informed of the study and asked to participate in the evaluation. Five Salvadorian experts agreed to review and edit the instrument to achieve a Spanish version that was a semantically, idiomatically, and conceptually equivalence for Salvadorian culture. The panel consisted of three medical doctors familiar with issues related to HIV/AIDS and two teachers experienced with the school system. The form in Appendix E was delivered to the panel members place of work or handed to them. Directions on the form explained how to proceed with the evaluation, and the checklist described in Table 4 served as a reference guide. 64

PAGE 76

Table 4 The Definition of Semantic, Idiomatic, and Conceptual Equivalence Variable Definition Semantic Words and the composition of the sentences have the same meaning as the original item. Idiomatic Translation of idioms and colloquialisms into expressions understood by the target culture. Conceptual Validity of the concept and experiences of the target population. The ideas/concepts are the same for each item and will be understood by students in El Salvador. The concepts are relevant to the experiences of Salvadorian adolescents. Note. The definitions were adapted from Guillemin et al. (1993). The semantic, idiomatic, and conceptual nature of the items was judged by each panelist on a scale that was pass (1) or fail (0). A frequency for each of the three categories was determined by adding the passing scores for each item. Categories with a frequency of three or more were considered to have reached equivalence. If an item received three negative ratings, it was rewritten and evaluated by the Lay Panel. It was noted that one of the reviewers did not evaluate the semantic or idiomatic equivalence of the items, and only the scores from the other panelist were included. A very useful aspect of the process were the panelists comments and suggestions which led to many revisions that improved the cross-cultural adaptation of the instrument for use with high school students. The corrected items were added to the instrument, edited by the Lay Panel, and included in the next review phase. 65

PAGE 77

Study 3 Cultural Adaptation, Content Validity, and Readability Review by Salvadorian Expert Panel Cultural Adaptation A seven-member panel evaluated the appropriateness of the instruments cultural content. The panel identified items that needed to be modified or edited to reflect Salvadorian culture. This step ensured the items were interpreted as intended by the students in El Salvador. Table 5 describes the characteristics of the multi-disciplinary expert panel that agreed to assist with the cultural adaptation and content validation. Table 5 Qualifications of Panelists Person Profession Place of Employment A Medical doctor Ministry of Health B Medical Doctor Ministry of Health C Medical Doctor Ministry of Health D Medical Doctor FUNDASIDA E Teacher Montessori Academy F Teacher/Principal Principal of Montessori Academy G Teacher Montessori Academy Each person was provided with the studys objectives, instructions, and characteristics of the attitudinal domain (see Appendix F). Comments and suggestions were solicited from the panel, and it was expected that they would improve the 66

PAGE 78

instruments use with adolescents in El Salvador. Cultural appropriateness was rated with a Likert scale that ranged from not at all appropriate, slightly appropriate, moderately appropriate, to very appropriate and was scored from 1 to 4, respectively. Cultural appropriateness was determined by calculating an average score for each item. A score of 2 indicated an item was slightly inappropriate, and, for that reason, items with an average of 2 or less were culturally inappropriate and were reevaluated. In contrast, a score of 3 or 4 indicated the item was culturally acceptable to Salvadorian adolescents. The results according to the panelists are reported in Chapter 4 (Results) and in the Appendix. Content Validity The instruments content validity was established through a two-step process. The first step was an evaluation of the items with a 4-point scale by an expert panel. Scores from the panel were summed and an average was calculated. Items with a mean score of 2 or less were considered as not appropriate for the domain. A similar rating procedure by Jemmott, Jemmott, and Fong (1998) indicated this strategy determined if the HIV content of an instrument was adequate. The second step followed Adays (1996) recommendation to solicit items from experts familiar with the domain. This study asked HIV experts to provide items that were missing from the Spanish instrument. The extra items would detect aspects of the domain that were not included in the instrument and identify items that were culturally relevant to students in El Salvador. The results from both steps were triangulated to establish content validity. The panel was asked to complete the evaluation within four days; however, four individuals were not able to meet the deadline. Because of a strict schedule with the 67

PAGE 79

upcoming pretest with the students, it was decided to proceed with the analysis as planned and to utilize only three completed reviews. However, the final analysis included all the forms and the previous decision was found not to have affected the results because the rating and comments were similar to those from the first group. Additionally, a review of the forms indicated that some of the categories were not completely filled out. Again, there were sufficient values to calculate an average score and the results were not affected. Role of Salvadorian Lay Panel Table 6 shows the individuals who formed a lay panel that reviewed the instrument throughout the adaptation. The instruments directions, items, and response categories were proofread by this panel to identify items that needed to be reworded. Other functions were to corroborate the other panel's recommendations, edit new material, and oversee the revisions that occurred after the pretest. In particular, their expertise with the Spanish language and awareness of the culture was not only helpful in correcting spelling errors and placing accents on the words, but an asset in culturally adapting the final version. Table 6 Characteristics of Salvadorian Lay Panel Person Training/Occupation H Psychologist/wife/mother of one child I Agronomy/Father of three children J Accounting/Housewife 68

PAGE 80

Reading Level of the Instrument Readability formulas, such as the SMOG, Lorge, Fox index, and the Cloze Procedure, were frequently used to determine the reading level of books and other reading materials (Contreras, Garcia-Alonso, Echenique, & Daye-Contreras, 1999). The formulas were adequate for determining the reading level of books and reading passages. However, valid readability formulas for instruments that consisted of only one or two sentences were not available. The Word XP (Microsoft, 2003) program has a built-in readability formula (Flesch-Kincaid), and an analysis determined the readability in English was at a grade level of 5.2. Educators in El Salvador were consulted about the most appropriate procedure to determine the instruments readability. They stated that experienced educators could analyze the instrument and determine its reading level. Two educators were selected for the readability analysis. The first reviewer was the principal of the Montessori Academy in San Salvador. The second reviewer was a psychologist/sexuality educator in one of the public high schools. They read the final instrument and determined the reading level. As an additional measure, the lay panel was asked to determine the reading level of the instrument. The reliability of the method in which persons estimate the reading level has not been investigated, and the results have to be interpreted with caution. Study 4 Pretesting with Salvadorian Adolescents The purpose of the pretest was to identify ambiguous, irrelevant, difficult to read, or misunderstood items. Also, this phase identified material that was not culturally relevant or inappropriate for use with students in El Salvador. The instrument was 69

PAGE 81

administered under normal testing conditions, and afterwards the probing technique described by Guillemin et al. (1993) was the method that determined the level of understanding and identified problematic items. The results of the probing technique served to corroborate the findings from Study 3. It also facilitated the preparation of the final instrument by including language that was relevant to high school students. Permission was obtained from the government of El Salvador to conduct the study in the public high schools in San Salvador and a permission letter was issued to all the participating schools (see Appendix G). The principal of the school was informed in person about the study and then asked to participate. Ten high schools were randomly selected from a larger pool of 30 high schools. Two schools from this group were randomly selected for a pretest. However, due to a conflict with the schedule only School 5 participated. One week before the pretest eight Salvadorian high school students (four male and four female) were asked to participate. The study was explained to them, and they were given a signed consent form as requested by the Institutional Review Board (IRB) at the University of South Florida (demonstrated in Appendix H). The day of the pretest, the onsite administrator from School #5 provided an empty classroom. Two groups of four, consisting of an equal number of male and female students, were scheduled for two different testing sessions. The nature of the study and the right not to participate was explained again to all the participants. A $10 incentive was offered to each student. The students were asked to circle items that were unclear, not understood, or culturally inappropriate. The primary investigator encouraged the use of written comments on the instrument to avert non-participation among shy students. At the end of the test session, the items were read aloud and the proctor asked: What does 70

PAGE 82

this mean? The students provided answers aloud, and if the correct meaning was not provided, students were then asked to explain their interpretation. Moreover, they were encouraged to discuss circled or problematic items and asked to explain their perception of these items. The findings are presented in the results section. The comments or suggestions made during the two pretests were compared to identify problematic items. The directions, items, and answer choices of the instrument were revised as needed and were included in the final instrument. As an additional measure, the lay panel edited items that were reworded or changed during this stage. To summarize, the pretest served to corroborate the results of Study 3 and helped to prepare the final instrument. Study 5 Confirmatory Factor Analyses and Concurrent Validity Confirmatory Factor Analysis The Spanish instruments attitudinal construct (see Table 7) was evaluated using confirmatory factor analysis (CFA). The Calis statistical program in SAS version 8.2 (SAS Institute, 2001) was used to perform CFA. The Calis procedure makes use of structural equations in the same manner as the LISREL program for the analysis of data with latent variables (Hatcher, 1994). In particular, the covariance structure of the items and the maximum likelihood method is used to determine the relationship of the items to the latent or unseen variables (Hatcher, 1994). The program follows Bentlers recommendations for writing structural equations in which the variables are assigned to one latent variable (as described in Hatcher, 1994). 71

PAGE 83

To determine if the data fit the model, several fit indices were consulted. Specifically, the model was tested for overall fit by the 2 root mean square error of approximation (RMSEA), Bentler and Bonetts Non-normed Index, and Bentlers comparative fit index (Hatcher, 1994). Further, CFA revealed the correlation among the latent variables, provided factor loadings, and determined the magnitude of the errors associated with the items. Therefore the factorial validity of the attitudinal dimension will be established by determining the internal structure of the instrument. Table 7 Hypothesized Five-factor Model of Attitudes from CDC Items Item # Statement Factor 1. Peer Pressure 1. If your friends want you to do something that you think might not be safe, you should at least try it. 6. To keep your friends, you should go along with most things your friends want you to do. 11. Teen-agers should learn how to resist pressures from their friends. 15. When friends want you to do things you don't feel like doing, there's no harm in going along. 19. Teen-agers should be more willing to resist pressures from their friends. Factor 2. Abstinence 2. It's okay not to have sex while you are a teen-ager. 7. People who don't have sex before they get married are strange. 12. It's a good idea for teen-agers not to have sex. 20. These days it makes a lot of sense to wait to have sex until you get married. Factor 3. Condom Use 8. It is not smart to have sex without using a condom. 16. Using a condom doesn't make sex less pleasurable. 21. If people think they might have sex during a date, they should carry a condom. 23. People who use condoms during sex don't trust the person they're with. 72

PAGE 84

Table 7 (continued) Factor 4. Drugs/Steroids 4 A teen-ager can inject drugs once in a while without a risk of getting infected with HIV. 9. Using needles to inject steroids or drugs is a bad idea. 13. People who share drug needles shouldn't worry because they probably won't get infected with HIV. 17. Anyone who shares needles is taking a chance of getting infected with HIV. 24. People who share drug needles should clean the needles with bleach. Factor 5. Threat of HIV infection 5. Teen-agers are at risk of getting infected with HIV if they engage in sex without a condom. 10. Its okay to have sex without a condom because your chance of getting infected with HIV is very low. 14. Teen-agers should realize that if they're not careful, they could get infected with HIV. 18. If teen-agers are careful about choosing sexual partners, they won't get infected with HIV. 25. HIV is something that teen-agers should think about when they date. Concurrent Validity The validity of the CDCs Condom and Abstinence subscales was established through a comparison of an existing model. A study by Basen-Engquist et al. (1999) determined the factorial validity of the six items described in Table 8. They used CFA to establish that three items loaded on an abstinence model and another three on a condom model. These items were translated into Spanish and treated in parallel with the translated items from the CDC. The attitudinal items that load on the abstinence and condom subscales were correlated with items from Basen-Engquist et al. (1999) and served to establish a measure of concurrent validity. 73

PAGE 85

Table 8. Basen-Engquist et al.s (1999) Condom and Abstinence Model Condom Model I believe condoms should always be used if a person my age has sex, even if the two people know each other very well. I believe condoms should always be used if a person my age has sex, even if the girl uses birth control pills. I believe condoms should always be used if a person my age has sex. Abstinence Model I believe its OK for people my age to have sex with several different people in the same month. I believe its OK for people my age to have sex with a steady boyfriend or girlfriend. I believe people my age should wait until they are older before they have sex. Data Collection Research Protocol 1. Permission for the investigation was obtained through the Department of Public Health and the Department of Education of El Salvador. The studys protocol, a version of the Spanish instrument, consent forms, and the application to the Institutional Review Board were provided to the appropriate officials in the Division of HIV of the Department of Public Health four months before the start of the study. Subsequently, the Department of Education granted a letter of permission (demonstrated in Appendix G) to conduct the study in 30 high schools in the 74

PAGE 86

metropolitan area of San Salvador. The 30 schools were separated according to their geographic location in San Salvador: Central, East, West, North, or South. Afterwards, 10 schools were randomly selected via a lottery-style drawing two different schools per geographic location. 2. The primary investigator met with the principals or an assistant principal from each school. The nature of the study, the governments letter of permission, the testing schedule, and an explanation of the consent forms were discussed at this meeting. The designated school official granted permission to conduct the study at the 10 schools. 3. The primary investigator and the principal of the school randomly selected the classrooms. A week before the test session, an explanation of the study was provided to the students in Spanish. Each student was given a general consent form to take home (see Appendix I). 4. As a backup measure, the principal investigator trained two Salvadorian proctors to assist with the instructions, answer student questions, provide the correct responses for the Knowledge items, and collect the instrument from each room. The trained proctors were not needed. The day of the test session, the principal investigator explained the study again. Anonymity was assured and they were reminded of their right not to participate. To assure anonymity, the students did not write their names on the instrument and they placed the completed instrument inside a box. At the end of the session, correct answers to the Knowledge section were provided and questions answered. The sample consisted of 483 students in grades 10, 11, and 12 from 10 different schools. 75

PAGE 87

5. A test-retest reliability study was carried out in 2 of the 10 schools. Permission from the principal was obtained; then the study was explained to the teachers and students. As described in number 3, the students were given a consent form (see Appendix J) and reminded of their right not to participate. To identify the responses and assure anonymity, the students were asked to write a six-digit number in the following sequence: two digits from their birth month, the last two digits of their primary residence or house number, and the last two digits of their phone number. Two classrooms were selected to obtain a sample of 29 students. At the end of the second session, they were provided with the correct answers to the Knowledge section and the student's questions were answered. Data Analysis of Specific Purposes This multi-phase study translated, cross-culturally adapted, and validated an instrument to evaluate the HIV/AIDS knowledge and attitudes of public high school students in El Salvador. The research purposes were analyzed as follows: Research purpose 1: Determine the success of the cross-cultural adaptation by contrasting the first translated version of the CDC instrument with the final Spanish version of the CDC instrument (as described in Table 9). The statements from the two versions were contrasted to determine cultural and conceptual differences. The differences between the first Spanish version and the final instrument were qualitatively recorded and interpreted. Research purpose 2: Determine using a Salvadorian panel the level of cultural appropriateness of the translated CDC HIV/AIDS knowledge and attitudinal items for use in El Salvador. The panelists scores (Study 3) were added together and an average 76

PAGE 88

was calculated for each item. The criterion of inappropriateness was an average of two or less. A score of 2 indicated an item was either slightly appropriate or not at all appropriate. In contrast, higher scores indicated the item was culturally acceptable to Salvadorian adolescents. Inappropriate items were edited and reevaluated by the Lay panel. Table 9 Analysis of the Six Specific Purposes Specific Purpose Variables Analysis 1. Determine the success of the cross-cultural adaptation by contrasting the first translated version of the CDC instrument with the final Spanish version of the CDC instrument. 2. Determine using a Salvadorian panel, the level of cultural appropriateness of the translated CDC HIV/AIDS knowledge and attitudinal items for use in El Salvador. 3. Ascertain the level of readability for the final Spanish version of the CDC instrument and evaluate content validity. 4. Evaluate the reliability of the scores from the Spanish version of the CDC instrument using a sample of Salvadorian students in grades 10, 11, and 12. 5. Determine if the Spanish version of the CDC instrument contains five attitudinal dimensions (Peer-pressure, Condom use, Abstinence, Drug use, and Threat of HIV) when analyzed by confirmatory factor analysis. 6. Evaluate the concurrent validity of the Abstinence and Condom use subscales from the Spanish version of the CDC instrument. Translated items Language Usage Cultural appropriateness of the items from the HIV/AIDS Knowledge and Attitude section. Expert review of the HIV/AIDS Knowledge and Attitude section. Scores from the instrument. Subscales scores: Peer-pressure, Abstinence, Condom use, Drugs use, and the Threat of HIV. Abstinence and Condom use subscale scores. Qualitative: cultural and conceptual differences. Items with an average score of 2 or less will be eliminated. Items with an average score of 2 or less will be eliminated. Test-retest correlation, Cronbachs alpha. Confirmatory Factor Analysis. Pearson product moment correlation coefficient. Research purpose 3: Ascertain the level of readability of the final Spanish version of the CDC instrument and evaluate content validity. 77

PAGE 89

The readability of the instrument was qualitatively assessed in El Salvador by the principal of the Montessori Academy and by an individual whose qualifications included the role of school psychologist and a teacher of the high schools sexuality curriculum. The Spanish instrument was reviewed independently, and the reviewer reported the reading level based on previous experience with written material for students in this population. Content validity was determined through a panel of Salvadorian experts (Study 3) who rated the knowledge and attitudinal items. An average score of 2 or less indicated the item was not valid for use in the instrument. Additional items were solicited to determine if the construct was not adequately represented. Research purpose 4: Evaluate the reliability of the scores from the Spanish version of the CDC instrument using a sample of Salvadorian students in grades 10, 11, and 12. Reliability was determined by the test-retest method at a one-week interval with 29 students in two different schools. The Pearson product-moment correlation coefficient was calculated for the test-retest scores (Gall et al., 1996). Internal consistency was determined by calculating Cronbachs alpha (Gall et al., 1996). Research purpose 5: Determine if the Spanish version of the CDC instrument contains five attitudinal dimensions (Peer-pressure, Condom use, Abstinence, Drug use, and Threat of HIV) when analyzed by confirmatory factor analysis. The Calis statistical program from the SAS Institute (2001) was used to perform CFA. Maximum likelihood estimation using the covariance matrix of the items was used to determine the factorial structure. The goodness of fit indices and standardized factor pattern coefficients were consulted to determine if the model fit the data. 78

PAGE 90

Research purpose 6: Evaluate the concurrent validity of the Abstinence and Condom use subscales from the Spanish version of the CDC instrument. The translated CDC Abstinence and Condom use subscales were correlated with their respective subscales from the Basen-Engquist et al. (1999) study. Summary of the Methods Section An instrument to evaluate HIV/AIDS knowledge and attitudes in an adolescent population in the United States was compiled by adding items from the CDC and Basen-Engquist et al. (1999). The instrument was translated by two experienced translators, reviewed by a bilingual committee of four individuals, and back-translated to English. The translated instrument was then cross-culturally adapted by a method recommended by Guillemin et al. (1993). A panel of content matter experts and teachers from El Salvador reviewed the instrument to establish the cross-cultural equivalence among the English and Spanish version. Afterwards a panel of Salvadorian experts established the cultural appropriateness and content validity of the instrument. Finally, a sample of 483 high school students in the metropolitan area of San Salvador answered the instrument. The scores of the instrument served to calculate the reliability and validity. 79

PAGE 91

CHAPTER FOUR RESULTS Purpose of the Study The purpose of this multi-phased study was to translate, cross-culturally adapt, and validate the scores from an instrument that assessed the HIV/AIDS knowledge and attitudes of high school students in El Salvador. The first phase utilized the back-translation method to obtain the first Spanish version of the instrument, which was developed by the CDC to evaluate HIV/AIDS knowledge and attitudes of students in grades 7 to 12. In the next phase, an expert panel of reviewers from El Salvador established the conceptual equivalence of the Spanish instrument. The third phase evaluated the instruments content validity and cultural acceptability through a panel composed of HIV experts, health professionals, schoolteachers, and students from El Salvador. The last phase involved collecting evidence of the validity of the scores from the Spanish version of the CDC instrument; 483 public high school students in the metropolitan area of San Salvador participated in the validation phase. Reliability of the scores was measured by the test-retest method and coefficient alpha. Confirmatory factor analysis was used to evaluate if the attitudinal construct had five different dimensions (Peer-pressure, Condom use, Abstinence, Drug use, and Threat of HIV). As part of the concurrent validation process, the Condom and Abstinence subscales from the translated CDC instrument were correlated with a set of existing attitudinal items from a published study by Basen-Engquist et al. (1999). Study 1 80

PAGE 92

Forward-translation, Synthesis, and Back-translation of Instrument Forward-translation Two translations were independently produced (see Appendix K) by two experienced translators. A common problem was reported during the translation process, item 15 from the Knowledge section was reported as problematic by both translators who were unsure of the translation of the words lambskin condoms. Translator 1 made the literal translation piel de cordero, and Translator 2 did not provide a translation. No other problems were reported during the translation. Synthesis Bilingual Reviewers 1, 2, and 4 systematically reviewed the two forward Spanish translations (see Appendix K) and synthesized these into one version. Bilingual Reviewer 3 verified and confirmed the final translation. Their selections, comments, and suggestions from the synthesis are reported in Appendix L. This section provides a description of the major findings. Bilingual Reviewer 1 commented (see Table 10) that item 12 was confusing due to the word infected and was unsure if contagiado or infectado was the appropriate selection. According to the reviewer, the words had a different interpretation: contagiado implied a meaning of an epidemic whereas infectado denoted having an infection. The word infectado was selected. Further, Knowledge item 15 was problematic due to the translation of the word lambskin. The reviewer did not know the correct translation of this word. Table 10 81

PAGE 93

Problematic Items and Comments Made by Bilingual Reviewer 1 Knowledge Items English Spanish Comment 11 Once you are infected with HIV, you are infected for life. Una vez que t has sido infectado con VIH, t ests infectado de por vida. Unsure of which is the correct word for infected Contagiado vs. infectado. 15 If you want to keep from getting HIV, using lambskin" condoms is just as good as using latex condoms. Si t quieres protegerte de contraer VIH, usando condones de piel de cordero es tan efectivo como usar condones de ltex. Unsure of how to translate lambskin. Attitudinal Items 1 If your friends want you to do something that you think might not be safe, you should at least try it Si un amigo quiere que t hagas algo y t crees que no es seguro, t deberas por lo menos tratar de hacerlo. Both translations do not retain the exact meaning. Difficult to translate double negative. 7 People who don't have sex before they get married are strange. Las personas que no tienen sexo antes de casarse son extraas Difficulty with the word strange. Raro o extrao can be used. Item does not retain the meaning of the original. Problematic item. 14 Teen-agers should realize that if they're not careful, they could get infected with HIV. Los adolescentes tienen que saber que si no son cuidadosos, ellos pueden ser infectados con VIH. Vague item. 15 When friends want you to do things you don't feel like doing, there's no harm in going along Cuando tus amigos quieren que t hagas cosas que t sientes no deben hacerse, no hay ningn riesgo si t las haces. Difficulty with Harm in going along This is a saying in the United States. Unsure of its translation. 21 If people think they might have sex during a date they should carry a condom. Si una persona piensa que podra tener sexo al salir con alguien ellos debern traer un condn. The translation for date in Spanish was problematic. Note Underlined words were problematic. Further, Bilingual Reviewer 1 reported the statement in Attitudinal item 1, If your friends want you to do something that you think might not be safe, you should at least try it, as having an unusual sentence structure. This made the Spanish translation difficult to read. Likewise, the word strange in Attitudinal item 7 did not translate well. 82

PAGE 94

The translators suggested the words raro and extrao, which were not good translations. Further, item 21 was problematic because of the word date. The reviewer was unsure of the term given by adolescents in El Salvador when they are in a relationship or are dating another person. Bilingual Reviewer 2 suggested the word encuesta in the directions to convey the process of answering a survey. As described in Table 11, it was suggested to use the words sern confidenciales for the words no one will know how you answered to assure anonymity. An additional recommendation included the substitution of the word SIDA (AIDS in English) instead of VIH when translating HIV. Other recommendations were to place el in front of the word SIDA for sentences denoting the masculine tense. Additionally, to improve the translations, this reviewer rewrote items 9, 15, and 16. Table 11 Problematic Items and Comments Made by Bilingual Reviewer 2 Number English Spanish Comment DO NOT put your name on this survey Your answers will be kept secret. No one will know how you answered these questions. No escriba su nombre en esta encuesta Sus respuestas sern confidenciales. Nadie sabr como contesto a las preguntas. Use word encuesta What is your gender ? Cul es tu sexo ? The word sexo is a better option than genero. Have you received HIV education in school? Has recibido educacin acerca de VIH en la escuela? Unsure of using term VIH o SIDA. 9 Using needles to inject steroids or drugs is a bad idea? Usar agujas para inyectarse esteroides no es mala idea. Item was reworded. Table 11 (continued) 83

PAGE 95

15 When friends want you to do things you don't feel like doing, there's no harm in going along. Cuando tus amigos quieren que t hagas cosas que t hagas cosas indebidas, no corrers riesgo si las haces. Item was reworded. 16 Using a condom doesn't make sex less pleasurable. Usar un condn no te impedir sentir placer. Offered a new translation. 20 These days it makes a lot of sense to wait to have sex until you get married. En esta poca tiene mas sentido esperarse a tener relaciones sexuales hasta casarse. Can use the word sentido o sensato. 25 HIV is something that teen-agers should think about when they date. SIDA es algo que los jvenes deben tener en cuenta cuando empiezan a socializarse. Unsure about the translation of date. It can be cita, socializarse o salieron alguien. Directions DIRECTIONS: This survey asks you to say whether you agree or disagree with a set of statements. Please read each statement, then indicate whether you: Instrucciones : En sta encuesta se te pide que digas cuando estas de acuerdo y cuando con la informacin. Por favor lea cada pregunta, entonces indica si usted esta: Instructions vs. directions. The translation is too long and wordy. Directions Strongly Agree (SA), Agree (A), are Not Sure (NS), Disagree (D), or Strongly Disagree (SD), by circling the answer you want. Completamente de Acuerdo (CA), De Acuerdo (DA), No estas Seguro (NES), No estoy De Acuerdo (NEDA), Completamente en Desacuerdo (CED), Haga un crculo en la respuesta elegida. The translation is too long and wordy. Note Underlined words were problematic. Bilingual Reviewer 4 reported that the term HIV could be equally translated as VIH or SIDA, as described in Table 12. A minor problem was reported for Knowledge item 9, which stated, A person can get HIV by sharing drug needles. The reviewer was unsure if the translation should have been, Puede una persona adquirir VIH al compartir agujas para la droga or Una persona puede adquirir VIH al compartir agujas para la droga. The subtle difference is that the former is a statement and the latter is an implied question. Likewise, the use of the word t and not usted was raised as an issue for Attitudinal item 1. The word t was used excessively, which made the statement sound redundant (see Table 12). 84

PAGE 96

Table 12 Problematic Items and Comments Made by Bilingual Reviewer 4 Knowledge Items English Spanish Comment 2 A person can "pass" an HIV antibody test (test negative) but still be infected with HIV. Una persona puede pasar un examen de anticuerpos de VIH (examen es negativo) pero puede estar infectado con VIH. Unsure about VIH or SIDA. 9 A person can get HIV by sharing drug needles. Puede una persona adquirir VIH al compartir agujas para la droga. Unsure if the correct translation is Una persona puede o puede una persona. 14 You can get HIV from a mosquito bite. Puedes t adquirir VIH de una picada de mosquito. Unsure if to include t and not usted. Attitudinal Items 1 If your friends want you to do something that you think might not be safe, you should at least try it. Si sus amigos desean que Ud. haga algo que Ud. no considera seguro, Ud. por lo menos debe de tratar de hacerlo Suggested the use of Si tus amigos and not tu amigos to start the sentence. 2 It's okay not to have sex while you are a teen-ager. Est correcto no tener sexo mientras sea un(a) adolescente. Change bien for correcto. 8 It is not smart to have sex without using a condom. No es inteligente tener sexo sin usar un condn. Add the word muy before inteligente. 11 Teen-agers should be more willing to resist pressures from their friends. Los adolescentes deben tener ms buena voluntad para resistir las presiones de sus amigos. Change to tienen que aprender and not use deben tener. 5 Teen-agers should realize that if they're not careful, they could get infected with HIV. Los adolescentes tienen que saber que si no son cuidadosos, ellos pueden ser infectados con VIH. Meaning is difficult to capture. The words tienen que saber is not the same as realize. Problematic item. Note Underlined words were problematic. Bilingual Reviewer 4 noted that Attitudinal item 1 started with the statement, Si sus amigos and suggested Si tus amigos (see Table 12). The addition of tus made the item grammatically correct. Item 2 stated Est correcto no tener sexo mientras seas un (a) adolescente, and the suggestion was to use the word bien instead of correcto. 85

PAGE 97

The word bien denotes a choice between good and bad and is similar to OK in English. In contrast, correcto is a formal translation, which can be inferred as a constriction imposed by society. Another change was recommended for the words its not smart in item 8 because this is a unique saying in the United States. The words muy inteligente were added to provide a similar interpretation in Spanish. Subsequently, the principal investigator (Bilingual Reviewer 1) and Bilingual Reviewer 4 reviewed all the translations to form a synthesis. To find the best translation, the items were read aloud in English and Spanish to compare the meaning and to listen to the grammatical structure of the sentences. As described in Table 13, items that did not sound right were discussed and then edited. Then the selections were spell-checked with the Spanish version of Microsoft Word XP (Microsoft, 2003). The information in Table 13 describes the changes made to the instrument. The directions were reworded into a concise statement to avoid the redundancy of information in the consent form. Other changes were to keep as separate entities the translation of the virus HIV to VIH and the disease AIDS to SIDA (the Spanish term). Moreover, both reviewers decided to use the word infectado over contagiado for Knowledge item 11, since it referred to someone who had been infected. Both reviewers were unsure of the proper translation of lambskin and selected piel de cordero. Attitudinal item 1, Si un amigo quiere que t hagas algo y t crees que no es seguro, t deberas por lo menos tratar de hacerlo, was found to be problematic due to an uncertainty of the 86

PAGE 98

translation of the word safe and the excessive use of the word t. The translations meaning was unclear and the item was considered problematic. Table 13 Problematic Items Discussed by Bilingual Reviewers 1 and 4 Item English Spanish Comment Directions Read each question. Carefully check the one answer that fits best. Some of the questions use the phrase "having sex." This means sexual intercourse. DO NOT put your name on this survey. Your answers will be kept secret. No one will know how you answered these questions. Lea cada pregunta. Con cuidado marque la respuesta que le parezca adecuada. Algunas preguntas usan la frase teniendo sexo. Esto significa relaciones sexuales. Original directions were shortened. Knowledge Items 2 A person can "pass" an HIV antibody test (test negative) but still be infected with HIV. Una persona puede pasar una prueba de anticuerpos del SIDA(resultados negativos) y aun estar infectado con VIH. Will use the term VIH for HIV and SIDA for AIDS. 9 A person can get HIV by sharing drug needles. Puede una persona adquirir VIH al compartir agujas para la droga. Decided to leave as a question in Spanish. 11 Once you are infected with HIV, you are infected for life. Una vez que t has sido infectado con VIH, t ests infectado de por vida. Infectado was preferred over contagiado. 15 If you want to keep from getting HIV, using lambskin" condoms is just as good as using latex condoms. Si t quieres protegerte de contraer VIH, usando condones de piel de cordero es tan efectivo como usar condones de ltex. Left piel de cordero in the statement. Attitudinal Items 1 If your friends want you to do something that you think might not be safe, you should at least try it Si un amigo quiere que t hagas algo y t crees que no es seguro, t deberas por lo menos tratar de hacerlo. Frequent use of t. 2 It's okay not to have sex while you are a teen-ager Est correcto no tener sexo mientras sea un(a) adolescente. Correcto vs. bien. 87

PAGE 99

Table 13 (continued) 6 To keep your friends, you should go along with most things your friends want you to do. Para mantener sus amistades, Ud. debe hacer la mayora de las cosas que tus amigos quieren que Ud. haga. Debated the use of usted or t. 7 People who don't have sex before they get married are strange. Las personas que no tienen sexo antes de casarse son extraas Unsure of how to translate strange. 14 Teen-agers should realize that if they're not careful they could get infected with HIV. Los adolescentes tienen que saber que si no son cuidadosos ellos pueden ser infectados con VIH Uncertain of the translation of careful and how it will be interpreted in Spanish. Note Underlined words were problematic. The translation of the word OK in items 10, 22, and 26 could be translated as bien or correcto. The decision was to use the word correcto to denote a choice between right or wrong. Lastly, item 7 was problematic because the word strange (Spanish word extraos) was difficult to translate and it did not retain the original meaning. The first Spanish version was then submitted to the last Bilingual Reviewer on the committee. Table 14 Bilingual Reviewer 3s Comments about the First Spanish Version Spanish Recommendation Has recibido educacin sobre El VIH/SIDA? a. Si____ b. No____ Add accent on Si. Knowledge Items 1 T no puedes contraer el SIDA si t tienes sexo una o dos veces sin un condn. Use the word usted instead of t. 3 Los condones son 100% efectivos en prevenir el VIH. Use para instead of en. 5 T puedes adquirir VIH al sentarte en un inodoro que fue usado por una persona con VIH. Use the word usted instead of t. 15 Si t quieres protegerte de contraer VIH, usando condones de piel de cordero es tan efectivo como usar condones de ltex. Use si quieres protegerte. 88

PAGE 100

Bilingual Reviewer 3 confirmed the soundness of the first Spanish version by comparing the following: (1) original English version; (2) two forward-translations; and (3) the first Spanish version. As reported in Table 14, Bilingual Reviewer 3 suggested the following revisions: (1) put an accent on Si; (2) use Ud. and not t in knowledge item 1; (3) change en for para; and (4) change to Ud. in item 5. In summary, four bilingual reviewers synthesized the first Spanish version by selecting translated items that retained the same meaning as the ones in the English instrument (as described in Appendix L). Additionally, the items were checked for the correct grammatical structure, spelling, and placements of accents. The first Spanish version was then submitted to a professional translator for a back-translation. Back-translation The Spanish items were back-translated into English by a professional translator, as reported in Appendix M. Afterwards, the meaning of both English items (the original and back-translated) were compared. The results of the comparison, described in Table 15, revealed that four items were slightly different from the original and that only minor modifications were needed to correct the items. For example, as described in Table 15, the word OK was incorrectly translated. The Spanish word bien was added to the instrument to address this. The back-translation of item 21 was problematic because the word date was lost in the translation. A new translation was suggested,Si las personas piensan que podran tener sexo al salir con alguien, ellos debern traer un condn. This translation implies that two people are going out. Other departures include the substitution of OK for better and the word think for believe. Both departures were acceptable. With the exception of 89

PAGE 101

the four items, no other differences were found. The first Spanish version is presented in Appendix N. Table 15 Items That Differed in the Original Version and the Back-translation Attitudinal Items Original Item Back-translation Comment 2 It's okay not to have sex while you are a teen-ager. Its right not to have sex while you are a teen-ager. Right is not the same as OK, use the Spanish word bien. 3 I believe condoms should always be used if a person my age has sex, even if the two people know each other very well. I think that condoms should always be used if a person my age has sex, even if the two people know each other well. The word believe has been used instead of think. Both words are acceptable. 9 It's okay to have sex without a condom because your chance of getting infected with HIV is very low. Its right to have sex without using a condom because the chance of becoming infected with AIDS is very low. Right is not the same as OK. Use the Spanish word bien. 21 If people think they might have sex during a date, they should carry a condom. If someone thinks that they are going to have sex with someone, they should bring a condom. The concept of going out on a date has been lost. 26 I believe its OK for people my age to have sex with a steady boyfriend or girlfriend. I think its better for people my age to have sex with their boyfriend or girlfriend. OK has been back translated as better. Study 2 Conceptual Equivalence by Salvadorian Expert Panel Results from the Salvadorian Expert Panel The semantic, idiomatic, and conceptual equivalence of the English and Spanish versions were rated on a pass/fail scale by five professionals from El Salvador. The scores ranged from 0 to 5 and were collected individually from each professional. The passing and failing scores were summed for the semantic, idiomatic, and conceptual categories by item, as reported in Appendix O. As previously described, a passing score 90

PAGE 102

3 for each category indicated item equivalence. Additionally, the panelists provided suggestions and comments that ultimately improved the instruments use in El Salvador. Table 16 Suggestions and Changes to the Knowledge Items Knowledge Items English Spanish Edited Spanish Version 7 You can get HIV from drinking from the same glass or water fountain that a person with AIDS drank from. Puedes adquirir VIH en persona con SIDA. Puedes adquirir VIH en el chorro o usando un vaso que ha sido usado por una persona con SIDA. 10 HIV can be found in breast milk. VIH puede estar presente en la leche maternal. VIH puede estar presente en leche materna. 12 People infected with HIV are usually very thin and sickly. Personas infectadas con VIH son bien delgadas y se notan enfermizas. Personas infectadas con VIH usualmente se ven bien delgadas y enfermizas 13 Some people have gotten HIV by swimming in the same pool as someone with AIDS. Algunas personas han contrado el VIH al nadar en la misma piscina que una persona que lo tiene. Algunas personas han contrado el VIH al nadar en la misma piscina donde ha nadado una persona que lo tiene. 14 You can get HIV from a mosquito bite. T puedes adquirir VIH de una picada de mosquito T puedes adquirir VIH de una picada de zancudo 15 If you want to keep from getting HIV, using "lambskin" condoms is just as good as using latex condoms. Si t quieres protegerte de contraer VIH, usando condones de piel de cordero es tan efectivo como usar condones de ltex. Si t quieres protegerte de contraer VIH, condones de ltex son los mejores The findings indicated that all items received a passing score in the three categories, and the scores ranged from 3 to 5. As illustrated in Table 16, comments about the items were provided by some of the panelists and they were recorded. Item 16,Using 91

PAGE 103

a condom doesn't make sex less pleasurable, had the lowest passing score of 3 for the semantic, idiomatic, and conceptual equivalence. One of the panelists determined this item was improperly translated. In contrast, higher scores of 4 and 5 were observed for other items in each of the three categories. Therefore, the semantic, idiomatic, and conceptual equivalence of the items was determined to be acceptable. Despite the fact that the panelists determined that all items were acceptable, they identified minor changes that improved the cultural acceptability of the items (see Table 16). As a result of including words recognized by Salvadorian adolescents, seven Knowledge items were reworded. The phrase in item 7, "los tomaderos de agua o tomando agua de un vaso usado por una was changed to en el chorro o usando un vaso que ha sido usado por una persona con SIDA. Likewise, they suggested other minor changes to the phrases in items 10, 12, 13, and 14 (see Table 16). A panelist noted that lambskin condoms (item 15) should be eliminated since they are not used in El Salvador. Therefore, the phrase in item 15 was reworded as Si t quieres protegerte de contraer VIH, condones de ltex son los mejores (To protect yourself from getting HIV, latex condoms are the best). This was the most extensive change among the Knowledge items. The six Attitudinal items reported in Table 17 were edited as a result of their suggestions. Three panelists recommended improving the sentence structure of item 1. Likewise, the words injecting steroids in item 9 were considered irrelevant to students in El Salvador. It was suggested to add the word tattoos to reflect the new trend among students. Next, item 16 was retranslated by changing it to Usar condones no hace el sexo menos placentero. This new translation had a more direct meaning. 92

PAGE 104

Table 17 Suggestions and Changes to the Attitudinal Items Attitudinal Items English Spanish Edited Spanish Version 1 If your friends want you to do something that you think might not be safe, you should at least try it. Si tus amigos quieren que t hagas algo y t crees que no es seguro t deberas por lo menos tratar de hacerlo. Si tus amigos quieren que t hagas algo y t crees que no es seguro, para complacerlos, t deberas de hacerlo 2 Its okay not to have sex while you are a teen-ager. Est bien no tener sexo mientras sea un(a) adolescente. Est bien no tener sexo mientras seas un(a) adolescente. 9 Using needles to inject steroids or drugs is a bad idea. Usar agujas para inyectarse esteroides o drogas es una mala idea. Usar agujas para inyectarse drogas o hacerse tatuajes es una mala idea. 16 Using a condom doesn't make sex less pleasurable. Es menos el placer del sexo si se usan condones. Usar condones no hace el sexo menos placentero 21 If people think they might have sex during a date they should carry a condom. Si las personas piensan que podran tener sexo al salir con alguien, ellos debern traer un condn. Si las personas piensan que podran tener sexo al salir con alguien, ellos debern llevar un condn. 26 I believe its OK for people my age to have sex with a steady boyfriend or girlfriend. Creo que est bien que las personas de mi edad tengan sexo con el novio o novia. Creo que est bien que las personas de mi edad tengan sexo con su novio o novia. Review by Lay Panel A lay panel of three individuals from El Salvador proofread the instruments directions, items, and responses. They recommended making the directions more explicit in the Knowledge and Attitudinal sections. As described in Table 18, the directions in the Knowledge section were modified to inform the students about the anonymity of their 93

PAGE 105

responses and to clarify how to mark their answers. Additionally, directions in the Attitudinal section were modified to include an underline in the phrase, cuando no estas de acuerdo. The panel identified Knowledge item 13 as unclear and suggested it be edited to reflect Salvadorian culture. Also, Attitude item 1 was not clear and was edited as follows, Si tus amigos quieren que t hagas algo y t crees que no es seguro, para complacerlos, t deberas de hacerlo. The new statement placed a focus on peer-pressure. As described in Table 18, minor changes were reported for items 5 and 26. The panelists reported that these changes improved the instruments readability among adolescents. Table 18 Recommendations Made by the Lay Panel Item English Version First Spanish Version Edited Spanish Version INSTRUCCIONES INSTRUCCIONES Directions Read each question. Carefully check the one answer that fits best. Some of the questions use the phrase "having sex." This means sexual intercourse. DO NOT put your name on this survey. Your answers will be kept secret No one will know how you answered these questions. Lea cada pregunta. Con cuidado marque la respuesta que le parezca adecuada. Algunas preguntas usan la frase teniendo sexo. Esto significa relaciones sexuales. No pongas tu nombre en la encuesta, es annima y nadie sabr que respondiste Lea las afirmaciones y marque la respuesta que le parezca adecuada con una X. Directions DIRECTIONS: This survey asks you to say whether you agree or disagree with a set of statements. Instrucciones: En sta encuesta se te pide que digas cuando estas de acuerdo y cuando no estas de acuerdo con lo que se dice aqu. Instrucciones: En sta encuesta se te pide que digas cuando estas de acuerdo y cuando no estas de acuerdo con lo que se dice aqu. Knowledge Item 13 Some people have gotten HIV by swimming in the same pool as someone with AIDS. Algunas personas han contrado el VIH al nadar en la misma piscina que una persona que lo tiene. Algunas personas han contrado el VIH al nadar en la misma piscina donde han permanecido personas con SIDA (Change made in final revision). Table 18 (continued) Attitudinal Items 94

PAGE 106

1. If your friends want you to do something that you think might not be safe, you should at least try it. Si tus amigos quieren que t hagas algo y t crees que no es seguro t deberas por lo menos tratar de hacerlo. Si tus amigos quieren que t hagas algo y t crees que no es seguro, para complacerlos, t deberas de hacerlo 5 Teen-agers are at risk of getting infected with HIV if they engage in sex without a condom. Los adolescentes tienen el riesgo de ser infectados con el virus del SIDA si tienen relaciones sexuales sin usar un condn. Los adolescentes corren el riesgo de ser infectados con el virus del SIDA si tienen relaciones sexuales sin usar un condn. 26 I believe its OK for people my age to have sex with a steady boyfriend or girlfriend. Creo que est bien que las personas de mi edad tengan sexo con el novio o novia. Yo creo que est bien que las personas de mi edad tengan sexo con su novio o novia. (Revision made in final versin). Study 3 Cultural Acceptability, Content Validation, and Readability Cultural Acceptability A seven-member panel of Salvadorian professionals evaluated the instruments cultural appropriateness and established its content validity using a four-point scale. As reported in Appendix P, the individual scores and comments are described for each item. The score for each item was added together and an average was calculated. The procedure, as discussed in the Methods section, was to eliminate or rewrite an item with an average score equal to, or less than, 2. The average score of cultural appropriateness among the Knowledge items ranged from 2.7 to 4.0. As demonstrated in Table 19, an average of 2.7, 3.4, and 3.4 were calculated for items 1, 2, and 5, respectively. An average score of 3.6 was observed for items 12 and 14. Additionally, items 3, 4, 6, 7, 8, 9, 10, 11, 13, and 15 had the highest averages. Therefore, all the Knowledge items were rated as culturally appropriate for use with Salvadorian high school students. 95

PAGE 107

Table 19 Average Rating of Appropriateness for Salvadorian Culture by Seven Panel Members Knowledge Items Cultural Appropriateness Attitudinal Items Cultural Appropriateness 1 2.7 1 2.8 2 3.4 2 3.4 3 4.0 3 3.9 4 4.0 4 3.3 5 3.4 5 3.7 6 3.7 6 2.7 7 3.7 7 3.0 8 3.9 8 3.6 9 3.8 9 3.3 10 3.7 10 3.4 11 3.7 11 3.9 12 3.6 12 3.3 13 3.9 13 3.6 14 3.6 14 3.6 15 3.7 15 2.9 16 3.4 17 3.7 18 3.6 19 3.7 20 3.7 21 3.4 22 3.6 23 3.9 24 2.6 25 3.7 26 3.9 27 3.9 28 3.1 29 2.8 Note Scale 1 = not at all appropriate, 2 = slightly appropriate, 3 = moderately appropriate and 4 = very appropriate The next category, the Attitudinal items, had average scores that ranged from 2.8 to 4.0. Low mean scores of 2.8, 2.7, 2.6, and 2.8 were reported for items 1, 6, 24, and 29, respectively. Additionally, mean scores of 3.4, 3.3, 3.0, 3.3, 3.4, 3.3, and 3.4 were reported for items 2, 4, 7, 9, 10, 12, and 21, respectively. The highest scores were observed for items 3, 11, 19, 20, 23, 27, 28, and 29 and they were 3.9, 3.9, 3.7, 3.7, 3.9, 3.7, 3.7, and 3.6, respectively. These results indicate that all the attitudinal items were appropriate for Salvadorian culture. Content Validation 96

PAGE 108

The content validity scores for the Knowledge items ranged from 2.8 to 4.0 (described in Table 20). The lowest average of 2.8 was reported for item 1. The other items had mean scores over 3; for example items 3, 5, 7, 12, and 14 had average scores of 3.3, 3.3, 3.4, 3.2, and 3.2, respectively. Items 4, 6, 8, 9, 10, and 11 had mean scores of 3.8, 4.0, 3.8, 3.8, 4.0, and 3.8, respectively. The ratings determined that the content validity for the knowledge items was acceptable. The average for the Attitudinal items ranged from a low of 2.8 to a high of 4.0. Items 1 and 6, both peer-pressure items, received the lowest scores of 2.3 and 2.8, respectively. Items 4, 7, 9, 12, 15, 22, 24, and 26 had means of 3.2, 3.2, 3.2, 3.0, 3.2, 3.2, 3.2, and 3.2, respectively. Items 11, 17, 20, 25, 27, and 28 had higher mean scores of 4.0, 3.7, 3.8, 3.8, 3.7, and 3.7, respectively. Therefore, the content validity of the attitudinal items was judged as acceptable. Once again the panel made recommendations that would improve the instrument. They suggested modifying items 1, 9, 11, 15, and 16 from the Knowledge section to improve their use with adolescents. They suggested the substitution of the word sexo for relaciones sexuales to describe the term to have sex. Additionally, they suggested using both terms HIV/AIDS when referring to either HIV or AIDS because more students would be familiar with this term. Another minor suggestion was to place the word el (the) in front of VIH/SIDA. The panel stated that double negatives are difficult to understand in Spanish and suggested the word no taken out of item 1. 97

PAGE 109

Table 20 Mean Scores for Content Validation by Seven Panel Members Knowledge Item Average Content Scores Attitudinal Items Average Content Score 1 2.8 1 2.8 2 3.8 2 3.7 3 3.3 3 3.5 4 3.4 4 3.2 5 3.3 5 3.8 6 4.0 6 2.3 7 3.4 7 3.2 8 3.8 8 3.5 9 3.8 9 3.2 10 4.0 10 3.0 11 3.8 11 4.0 12 3.2 12 3.0 13 3.5 13 3.5 14 3.2 14 3.5 15 3.8 15 3.2 16 3.6 17 3.7 18 3.0 19 4.0 20 3.8 21 3.5 22 3.2 23 3.6 24 3.2 25 3.8 26 3.2 27 3.7 28 3.7 29 3.6 Note. Scale 1 = not at all appropriate, 2 = slightly appropriate, 3 = moderately appropriate, and 4 = very appropriate. A new phrase was suggested for item 9 that would make drug use the focal point of the sentence (see Table 21). The item was changed to Una persona puede adquirir VIH/SIDA al compartir agujas en el consumo de diversas drogas. Additionally, item 11 was changed to, Una vez que t has sido infectado con VIH/SIDA, t eres portador del virus toda tu vida. The change emphasizes that you will have the virus for the rest of your life. Likewise, the panel recommended modifying items 15 and 16. 98

PAGE 110

Table 21 Changes to Knowledge Items as Suggested by Salvadorian Panel During the Cultural Review Knowledge Items English Spanish Edited For Pretest Version 1 You can't get AIDS if you have sex only once or twice without a condom. T no puedes contraer el SIDA si t tienes sexo una o dos veces sin un condn. T puedes contraer el SIDA si t tienes sexo una o dos veces sin un condn. 9 A person can get HIV by sharing drug needles. Una persona puede adquirir VIH al compartir agujas para la droga. Una persona puede adquirir VIH/SIDA al compartir agujas en el consumo de diversas drogas 11 Once you are infected with HIV, you are infected for life. Una vez que t has sido infectado con VIH, t ests infectado de por vida. Una vez que t has sido infectado con VIH/SIDA, t eres portador del virus toda tu vida 15 If you want to keep from getting HIV, using "lambskin" condoms is just as good as using latex condoms. Si t quieres protegerte de contraer VIH condones de lates son los mejores. Una forma de prevencin del VIH/SIDA es el uso de condones de ltex. Note. Underlined items were included in the final version. Attitudinal items 1, 6, 16, and 19 were problematic and edited according to the panels recommendations. As described in Table 22, item 1 was depicted as too wordy, and subsequently it was reworded to make it more concise by reducing the number of ts in the statement. Similarly, a suggestion was made to eliminate or rewrite the word t in item 6 because it was repeated four times. Another revision was suggested by one of the panelists, who rewrote item 16 as follows: Usar condones no hace la relacin sexual menos placentera. He stated this new translation was more accurate. Item 24 was labeled as problematic by one of the panelists who said the topic should not be mentioned 99

PAGE 111

to students in high school. However, when the content rating (3.2) and cultural appropriateness (2.8) of item 24 was compared, the evidence did not evoke a problem. Table 22 Changes to Attitudinal Items as Suggested by the Salvadorian Panel During the Cultural Review Attitudinal Items English Spanish Edited For Pretest Version Comment 1 If your friends want you to do something that you think might not be safe, you should at least try it. Si tus amigos quieren que t hagas algo y t crees que no es seguro, t deberas por lo menos tratar de hacerlo. Si tus amigos quieren que t hagas algo incorrecto y t crees que no es seguro, deberas por lo menos tratar de hacerlo. Make it more direct, Too many Ts, (You's). 6 To keep your friends, you should go along with most things your friends want you to do. Para mantener tus amistades, t debes hacer la mayora de las cosas que tus amigos quieren que t hagas. Para mantener tus amistades, debes hacer la mayora de las cosas que tus amigos quieren que hagas. One suggestion to rewrite it. Too many Ts. 16 Using a condom doesn't make sex less pleasurable. Usar condones no hace el sexo menos placentero. Usar condones no hace la relacin sexual menos placentera. 19 Teen-agers should be more willing to resist pressures from their friends. Los adolescentes deben tener ms buena voluntad para resistir las presiones de sus amigos. Los adolescentes deben tener ms fuerza de voluntad para resistir las presiones de sus amigos. 24 People who share drug needles should clean the needles with bleach. Las personas que comparten agujas para drogas debern limpiarlas con leja. A panelist said it was not a good question. Note Underlined words were problematic. Readability The readability of the instrument was assessed independently in El Salvador by the principal of the Montessori Academy and by an individual who was a school 100

PAGE 112

psychologist/sexuality education teacher. They reported the words were written at the high school reading level. This finding was corroborated by a similar review conducted by the lay panel. However, the reliability of this method has not been reported and because of this, results have to be interpreted with caution. Study 4 Pretest with Salvadorian High School Students High School 5 was selected for a pretest of the instrument. Two groups of students, consisting of two male and two female students per group, answered the Spanish instrument. The students were asked to circle items that were difficult to understand or hard to read. Additionally, they were asked to write comments next to the problematic items. The students took approximately 35 minutes to complete the instrument. Afterwards, the probing technique was used to detect the meaning of each question and to identify problematic items. Items were read aloud by the principal investigator and the students were asked to provide a brief interpretation of the item. The students took turns responding back to the items and made general comments about the instrument. They said that the instrument was easy to read and that the content would be easily understood by other high school students. They stated that the items were appropriate for an HIV/AIDS survey and that the content was similar to what they had been exposed to in their own HIV class. Positive comments were also made in regard to the cultural appropriateness of the content. Table 23 describes the findings of the probing technique for the Knowledge section. The students found item 1 to be ambiguous because it was unclear if any of the 101

PAGE 113

persons depicted in the statement were infected with HIV. They suggested the sentence be clarified by indicating if one of the persons having sex had HIV. Item 2 was the only other problematic item from the knowledge section. This item made reference to testing, however, none of the students understood what was being asked. They did not know the meaning of the term anticuerpos (antibodies). When the term was explained to them, they stated that a blood test would be a better term. To improve the item's understanding, the Spanish words una prueba de sangre or its English translation blood test was added to item 2. Table 23 Pretest Results of the Knowledge Section Knowledge Items English Spanish Edited For Final Version Student comments Group 1. 1 You can't get AIDS if you have sex only once or twice without a condom. T puedes contraer el VIH/SIDA si t tienes sexo una o dos veces sin un condn. T puedes contraer el VIH/SIDA si t tienes sexo una o dos veces sin un condn. Two students from group 1 and two from group 2 wanted to know if one of the persons is infected with HIV. They said the question could be interpreted in several ways. 2 A person can "pass" an HIV antibody test (test negative) but still be infected with HIV. Una persona puede pasar una prueba de anticuerpos de VIH/SIDA (resultados negativos) y aun estar infectado con VIH. Una persona puede pasar una prueba de sangre para el VIH/SIDA (resultados negativos) y aun estar infectado con VIH. All of the students did not know the meaning of antibodies. Use blood test. The probing technique determined that Attitudinal items 1, 7, 8, 10, 14, 19, 21, and 24 were not interpreted the same as the original version or were problematic. As described in Table 24, the students found item 1 (Peer-pressure subscale) confusing and 102

PAGE 114

unclear. One students suggestion was to describe a specific situation in which the teens have to make a choice under the influence of their friends that is right or wrong. However, two students said they vaguely interpreted the situation as related to peer-pressure, but they were uncertain if others would do the same. Table 24 Pretest Results of the Attitudinal Section Attitudinal Item English Spanish Edited Version Student comments 1 If your friends want you to do something that you think might not be safe, you should at least try it. Si tus amigos quieren que hagas algo incorrecto y t crees que no es seguro, para complacerlos, deberas por lo menos tratar de hacerlo. No change. Interpreted as not going along with your friends. A student suggested giving a specific example. Another student said this was related to peer-pressure. One of the students found it confusing. 7 People who don't have sex before they get married are strange. Las personas que no tienen sexo antes de casarse son extraas. Las personas que no tienen relaciones sexuales antes de casarse son extraas. A student responded that if you do not have sex then you are on the other side of the fence. Another said there was something wrong with the person o algo mal en si. 8 It is not smart to have sex without using a condom. No es muy inteligente tener sexo sin usar un condn. No es muy inteligente tener relaciones sexuales sin usar un condn. A student left a note. They did not understand this question. 10 It's okay to have sex without a condom because your chance of getting infected with HIV is very low. Esta bien tener sexo sin usar un condn porque es muy bajo el chance de ser infectado con VIH/SIDA. Esta bien tener relaciones sexuales sin usar un condn porque el riesgo de ser infectado con VIH/SIDA es muy bajo. Suggested to not use the word chance. Use muy pocas las posibilidades o riesgo. 14 Teen-agers should realize that if they're not careful, they could get infected with HIV. Los adolescentes tienen que saber que si no son cuidadosos, ellos pueden ser infectados con VIH/SIDA. Los adolescentes tienen que saber que si no son cuidadosos, ellos pueden ser infectados con VIH/SIDA. One of the students said the question is too general and open to several interpretations. Another student said that not all people know how HIV is transmitted. 103

PAGE 115

Table 24 (continued) 19 Teen-agers should be more willing to resist pressures from their friends. Los adolescentes deben tener ms buena voluntad para resistir las presiones de sus amigos. Los adolescentes deben tener ms fuerza de voluntad para resistir las presiones de sus amigos. One of the students commented it was the same as question 11. Two students said it was a confusing question since it did not mention a specific situation. 21 If people think they might have sex during a date, they should carry a condom. Si las personas piensan que podran tener sexo al salir con alguien, ellos debern llevar un condn. Si las personas piensan que podran tener sexo al salir con su pareja, ellos debern llevar un condn. A student commented the question is not appropriate. Teens do not go out on dates and expect to have sex. They said sex is only in a relationship. Suggested to add the word pareja. 24 People who share drug needles should clean the needles with bleach. Las personas que comparten agujas para drogas debern limpiarlas con leja. Las personas que comparten agujas para drogas debern limpiarlas con leja. One of the students said it was a bad question and not something students are concerned with. Item 7 was problematic because it had several interpretations. According to one student, the word extrao was interpreted as a person who is on the other side of the fence, and this student made the clarification that they perceived the term as "being gay." Another student concurred and said they interpreted the word as someone who is a homosexual. Yet another student made the comment that there is something wrong inside of the person if they do not have sex. It was therefore noted that item 7 (Abstinence subscale) could have many interpretations and may not be representing the dimension to which it was assigned. A minor modification was suggested for item 10 (see Table 24) that included replacing the word chance with either riesgo or muy pocas las posibilidades. Items 14 and 19 were said to be too general and the students recommended specific examples of situations faced by teen-agers. They stated that the examples be more concrete and not 104

PAGE 116

open to interpretation. The other problem was with items 11 and 19; a student commented that the items were the same when they were not. This indicated that a similar sentence structure was affecting the interpretation of the items. The students commented that item 21 was irrelevant to adolescents in El Salvador. They said that in El Salvador young persons do not have sex when they go out on a date and that casual sex rarely happens. Instead, sexual relations occur if two people are in a long-term relationship. They suggested rewording the item to reflect this situation and recommended adding the word pareja as a reference to the couple. Further, a comment was made that item 24 was a bad item. According to one of the students, It is not important for adolescents to know how to clean drug needles. As described earlier, a panelist made the same comment and this indicated the item was problematic. Finally, the eight completed instruments were reviewed after the students left and the following items were circled: 2, 11, 19, 20, and 24 from the Attitudinal section. A note was written on the instrument, that said item 8 was not understood. The results were then used to edit the instrument, and with the assistance of the Lay Panel, the final instrument (see Appendix Q) was prepared for the reliability and validity study. Study 5 Reliability Study, Confirmatory Factor Analysis, and Concurrent Validity Reliability Study Reliability of Knowledge Scores Participants. Ten schools were randomly selected from 30 public high schools in the metropolitan area of San Salvador. The schools were similar in that they were all public and were composed of students from a lower socioeconomic and social status. 105

PAGE 117

Nine of the schools were co-ed; however, two of the schools had different characteristics, School 5 was predominantly male, and School 8 admitted only females. A random selection of classrooms yielded a total of 496 students, and 13 either refused or were not permitted to participate. The final sample for the analysis consisted of 483 students. The sample consisted of 41% males and 59% females (as described in Table 25). Additionally, the distribution by grade level was 31% in 10 th 42% in 11 th and 27% in 12 th The classrooms were randomly selected by the school principal, and they ranged in size from 23 to 54 students. The students ages consisted of < 1% in the age group 13 to 14, 25% in the 15 to 16 age group, and 74% in the 17 or older age group. Finally, 90% of the students reported having received HIV education. Table 25 Characteristics of the 10 Public High Schools School Students Selected Grade Percentage of Males Percentage of Females 1 45 12 27 73 2 40 11 50 50 3 85 10 and 11 26 74 4 54 12 15 85 5 86 10, 11, and 12 97 3 6 28 10 42 58 l7 31 10 39 61 8 47 11 0 100 9 38 11 50 50 10 23 10 30 70 Total 483 10, 11, and 12 41 59 106

PAGE 118

Internal consistency. Prior to looking at internal consistency, descriptive statistics were computed for the 15-item Knowledge test. A total of 483 students answered the Knowledge test. The instrument was scored by allowing two correct answers per item, and, as described in the Methods section, non-missing responses were not considered for the final score. The information in Table 26 describes the distribution of the scores for the 10 schools. Schools 1 and 7 had the lowest mean score (62%) and School 4 had the highest mean score (74%). The mean score for the 10 schools was 69% correct and the standard deviation was 16. The mode was 80% and the median was 73%. The correct scores ranged from 20% to 100%. The scores had a negative skewness (-0.47) and a negative kurtosis (-0.16). The remaining characteristics were similar for each of the 10 schools. Table 26 Score Distribution of the 15 True and False Knowledge Items According to School School Number of Students Mean % Standard Deviation Maximum Score Minimum Score Skewness Kurtosis 1 47 62 18 93 26 0.12 -1.02 2 40 65 17 100 20 -0.21 -0.13 3 86 64 17 93 20 -0.60 -0.17 4 55 79 12 100 47 -0.46 0.13 5 86 70 14 100 33 -0.33 -0.21 6 28 74 11 87 53 -0.54 1.08 7 32 62 11 80 33 -0.73 0.17 8 48 75 12 93 46 -0.43 -0.46 9 38 70 14 93 33 -0.45 -0.10 10 23 65 16 93 33 -0.13 -0.82 Total 483 69 16 100 20 -0.47 -0.16 The scores of the 483 students had an internal consistency (Cronbach alpha) of .57. As described in Table 27, the item-to-total correlation ranged from .07 to .35. The observed alpha may have resulted from a combination of factors known to lower the 107

PAGE 119

reliability of scores. These factors include the variance of an item, item discrimination, and test length (Crocker & Algina, 1986). The item-to-total correlation is an index of discrimination and provides information about how strongly the item relates to the total score. Items that had an item-to-total correlation less than or equal to .2 are marginal, and items with a value greater than or equal to .3 are acceptable. The item-to-total correlations of items 6, 7, 8, and 13 in Table 27 indicated they had an acceptable index of discrimination. In contrast, the item-to-total correlation of items 2, 3, 9, 10, 11, 12, and 15 indicated a low discrimination. Table 27 Mean Score and Item-to-total Correlation of the 15 True and False Knowledge Items Item Mean Item-to-total correlation (Discrimination) 1 0.81 .21 2 0.46 .19 3 0.71 .07 4 0.78 .23 5 0.65 .25 6 0.79 .30 7 0.75 .36 8 0.92 .33 9 0.95 .13 10 0.63 .18 11 0.93 .17 12 0.22 .17 13 0.65 .35 14 0.53 .20 15 0.49 .15 Items answered correctly by most of the students lowered the variance. Because the calculation of alpha is affected by the variance in the sample (Crocker & Algina, 1986), the observed alpha was directly affected by a lower variance. Items 8, 9, and 11 were answered correctly by more than 90% of the students, and according to their item-to-total correlation, they should be reworded or eliminated to improve the reliability of 108

PAGE 120

the scores. These items were assessing knowledge that was important for students to know and were therefore an integral component of the instrument. In contrast, items 2 and 12 were the most difficult for the students. Item 2 (A person can "pass" an HIV antibody test (test negative) but still be infected with HIV) was correctly answered by 46% of the students. As in items 8, 9, and 11, this item assessed an important dimension. Similarly, item 12 (People infected with HIV are usually very thin and sickly) was correctly answered by 22% of the students. As illustrated in Table 27, there were not enough difficult items. A characteristic of a well-designed instrument was an equal number of difficult and non-difficult items. Nevertheless, the disproportion of difficult items was not a concern because of the general purpose of the assessment. The intent of the Knowledge section was to assess the global characteristics of HIV transmission, identify risky behaviors, and dispel misconceptions. The instrument was similar to a criterion-based test where the researcher is concerned with mastery of a particular domain. Test-retest reliability. The reliability study was conducted with a subset of 39 students who were randomly selected from the studys larger population of 483 students. The initial group for the subset consisted of 26 students from School 1 and 35 students from School 5. Unfortunately, during the last administration of the instrument, many of the students were not present in their classroom and did not participate. Fourteen students from School 1 and six from School 5 were absent. One of the teachers commented that it was unusual for so many students to be absent. Additionally, two more students incorrectly filled out the identification codes and were dropped from the reliability study. 109

PAGE 121

The final sample for the test-retest study consisted of 39 individuals. There was a disproportionate number of male students (74 %) in the sample, and the distribution by grade level was 67% in 10 th and 33% in 12 th Additionally, 72% of the students reported receiving HIV education. These students were given the same instrument at two different times during a one-week period. The scores from the two occasions were correlated to obtain a Pearson product-moment correlation coefficient. Table 28 Characteristics of the 15 True and False Knowledge Scores at Time 1 and Time 2 Scale Time 1 (n = 39) Time 2 (n = 39) Mean 63% 60% Mode 47 67 Median 66 60 Standard Deviation 17 16 Skewness -0.35 0.08 Kurtosis -.091 -0.11 Cronbach's alpha .62 .52 The test-retest reliability coefficient for the scores of the 15 Knowledge items was .59. This implied that 59% of the variance was due to the true score variance for this group of students. Additionally, 31% of the variance can be attributed to measurement errors associated with temporary changes in the examinees. As described in Table 28, the means (correct answers) for Time 1 and Time 2 were 63% and 60%, respectively. Additionally, the alpha at Time 1 and Time 2 was very similar, .62 and .52, respectively. It was noted that the alpha for the larger study was .57. 110

PAGE 122

Reliability of the Attitudinal Scores Descriptive statistics of Attitudinal items. The 23 translated items from the CDC were administered to a sample of 483 students. It was noted that the instrument had many unanswered items. Particularly, items 3, 16, 23, 24, and 25 were recorded as having 16, 17, 11, 14, and 14 missing responses, respectively. The largest number of non-responses was observed for items 3, 16, and 23, which were from the Condom subscale. A non-response may have come from students who were either sexually inexperienced or who were uncomfortable answering the questions about condom use. Missing data were not used in the calculation of the mean score or in any of the other calculations. The items in the instrument included statements that were a combination of negative or positive statements, and this mixture was expected to avoid a response set. The negative statements were scored so that strongly agree = 1, agree = 2, not sure = 3, disagree = 4, and strongly disagree = 5. Positive statements were scored in reverse so that strongly disagree = 1, disagree = 2, not sure = 3, agree = 4, and strongly agree = 5. The mean, standard deviation, skewness, and kurtosis for the translated Attitudinal items are reported in Table 29. The means for the 23 attitudes ranged from 2.30 to 4.60. Univariate tests showed that for many of the items the distribution was negatively skewed. This indicated that for most of the students the scores were mostly positive. The large value for skewness was -2.72. Kurtosis values ranged from -1.10 (item 24 in the Drug use subscale) to 7.80 (item 14 in the Threat of HIV subscale). 111

PAGE 123

Table 29 Descriptive Statistics for the 23 CDC Translated Attitudinal Items Factor N Number of Missing Scores Mean Standard Deviation Skewness Kurtosis Peer-pressure 1 479 4 4.34 0.80 -1.46 2.84 6 481 2 4.34 0.87 -1.63 2.96 11 475 8 3.84 1.49 -0.91 -0.77 15 479 4 4.02 1.08 -1.19 0.97 19 477 6 4.39 1.05 -1.87 2.63 Abstinence 2 480 3 3.87 1.34 -0.87 -0.56 7 480 3 4.27 1.12 -1.66 1.96 12 479 4 3.99 1.27 -1.05 -0.13 20 479 4 4.25 1.14 -1.44 0.97 Condom use 8 475 8 3.82 1.33 -0.93 -0.37 16 466 17 2.95 1.25 -0.03 -0.77 21 477 6 4.23 1.03 -1.57 2.04 23 472 11 3.46 1.33 -0.42 -0.95 Drugs 4 475 8 4.01 1.12 -1.02 0.35 9 479 4 4.26 1.27 -1.63 1.29 13 478 5 4.33 1.09 -1.82 2.61 17 477 6 4.56 0.86 -2.49 6.41 24 469 14 2.54 1.41 0.41 -1.10 Threat of HIV 5 480 3 4.39 1.11 -2.03 3.19 10 479 4 4.31 1.09 -1.79 2.53 14 477 6 4.60 0.84 -2.72 7.80 18 476 7 2.36 1.32 0.66 -0.69 25 469 14 4.45 0.89 -2.10 4.57 Note Positive statements: Items 2, 5, 8, 9, 11, 12, 14, 17, 19, 20, 21, 24, and 25. Negative statements: Items 1, 3, 4, 6, 7, 10, 13, 15, 16, 18, 22, and 23. The negative statements were scored so that strongly agree = 1, agree = 2, not sure = 3, disagree = 4, and strongly disagree = 5. Positive statements were scored in reverse to compute the mean. Descriptive statistics of the Attitudinal subscales. As described in Table 30, the 23 translated items from the CDCs Booklet #6 (CDC, 2003) were hypothesized to represent five factors. The items were grouped according to the original designation given by the authors of the instrument. A mean score for each subscale was calculated by adding the responses for each item belonging to the subscale, and then dividing it by the total 112

PAGE 124

number of items in that subscale. A higher subscale mean score indicated a favorable attitude for that construct, and a lower mean score indicated an unfavorable attitude. The Condom subscale had the lowest mean (M = 3.63, SD = 0.61), and this indicated the majority of the responses were from students with a neutral attitude ( I am not sure). Likewise, the Drug use subscale had a mean of 3.94 and a standard deviation of 0.58, which was very close to an overall favorable attitude. The other subscales, Peer-pressure, Abstinence, and Threat of HIV, were considered to be favorable as they had mean responses greater than 4. In particular, the response for the Peer-pressure (M = 4.19, SD = 0.64) was the highest of the five subscales. Table 30 Descriptive Statistics of Attitudinal Subscales Factor N Mean Standard Deviation Skewness Kurtosis Peer-pressure 465 4.19 0.64 -0.61 -0.21 Abstinence 472 4.09 0.81 -0.77 0.01 Condom 453 3.63 0.61 -0.12 0.07 Drugs 453 3.94 0.58 -0.57 0.29 Threat of HIV 458 4.03 0.54 -0.56 0.57 Note Sample was 483 students. N is the number of responses. Pearson product-moment correlation coefficients were calculated to determine the relationship among the five subscales. As described in Table 31, the largest correlation (.41) was calculated for the Peer-pressure and Threat of HIV infection subscales. The Abstinence and Peer-pressure subscales had a medium correlation of .39. The correlation among the Abstinence and Drug use subscales was .36. The Condom and Abstinence subscales had the lowest correlation (.12). A medium to moderate correlation was expected due to the content that was being assessed (HIV and AIDS) and because items that assess attitudes were known to have a moderate correlation. 113

PAGE 125

Table 31 Pearson Correlation Matrix of the Five Attitudinal Subscales (n = 393) Peer-pressure Abstinence Condom Drugs Threat of HIV Peer-pressure 1.00 Abstinence .35 1.00 Condom .32 .18 1.00 Drugs .36 .32 .21 1.00 Threat of HIV .41 .29 .33 .25 1.00 Note. p < .0001. Internal consistency. Internal consistency coefficients (Cronbach's alpha) for the translated CDCs attitudinal subscales are presented in Table 32. The subscales had reliability coefficients that ranged from 0 to .58 (set to zero because reliabilities have a minimum value of 0). The Peer-pressure and Abstinence subscales were found to have a reliability coefficient of .55 and .58, respectively. The reliability coefficients of the Drug use and Threat of HIV subscales were .24 and .30, respectively. In contrast, the Condom subscale did not have an acceptable level of internal consistency (0), and this indicated that the scores for this scale were not reliable. 114

PAGE 126

Table 32 Cronbachs Alpha and Item-to-total Correlation of the 23 Translated CDC Attitudinal Items Subscale N Mean Standard Deviation Item-to-total Correlation Peer-pressure 1 479 4.34 0.80 .34 6 481 4.34 0.87 .33 11 475 3.84 1.49 .27 15 479 4.02 1.08 .28 19 477 4.39 1.05 .42 Abstinence 2 480 3.87 1.34 .42 7 480 4.27 1.12 .16 12 479 3.99 1.27 .44 20 479 4.25 1.14 .44 Condom use 8 475 3.82 1.33 -.007 16 466 2.95 1.25 -.13 21 477 4.23 1.03 .06 23 472 3.46 1.33 .07 Drugs 4 475 4.01 1.12 .16 9 479 4.26 1.27 .20 13 478 4.33 1.09 .12 17 477 4.56 0.86 .31 24 469 2.54 1.41 -.12 Threat of HIV 5 480 4.39 1.11 .17 10 479 4.31 1.09 .22 14 477 4.60 0.84 .14 18 476 2.36 1.32 .02 25 469 4.45 0.89 .18 Note Positive statements: Items 2, 5, 8, 9, 11, 12, 14,16, 17, 19, 20, 21, 24, and 25. Negative statements: Items 1, 3, 4, 6, 7, 10, 13, 15, 18, 22, and 23. The negative statements were scored so that strongly agree = 1, agree = 2, not sure = 3, disagree = 4, and strongly disagree = 5. Positive statements were scored in reverse to compute the mean. The range of the item-to-total correlations for the subscales were from -.13 to .45 (see Table 33). The item-to-total correlation identifies how the items measure the scale. Items with a negative or low value suggest possible weak items (Hatcher, 1994). The item-to-total correlations in the Peer-pressure and Abstinence subscales were determined to be acceptable. The item-to-total correlations for the Threat of HIV (.02 to .23) were 115

PAGE 127

very low. Additionally, the range of the item-to-total correlation for the Condom and Drug use subscales was low and indicated that some of the items were weakly related to their respective construct. Table 33 Cronbach Alpha and Item-to-total Correlation of the Subscales Subscale N Mean Number of Items in the Subscale Cronbach Alpha Range of Item-to-total Correlation Peer-pressure 465 4.19 5 .55 .27 to .42 Abstinence 472 4.09 4 .58 .16 to .45 Condom use 453 3.63 4 0.0 -.13 to .08 Drug use 453 3.94 5 .24 -.13 to .31 Threat of HIV 458 4.03 5 .30 .02 to .23 Note N is the number of responses from a sample of 483 students. The non-response for the subscales were 18 (Peer-pressure), 11 (Abstinence), 30 (Condom use), 30 (Drug use), and 25 (Threat of HIV). Test-retest. The responses from the subset of students (39) were used to calculate a reliability coefficient between the two scores after a one-week period. The Peer-pressure subscale had the highest reliability coefficient (.62). In decreasing order, the next smallest reliability coefficient was observed for the Drug use subscale (.58). And the next two smallest were calculated for the Threat of HIV (.49) and Abstinence (.43) subscales, respectively. The lowest reliability coefficient (.17) was observed for the Condom use subscale. Overall, the reliability of the scores was moderate for four of the subscales and inadequate for the Condom use subscale. The information in Table 34 indicated that the subsets mean score for each subscale was similar at Time 1 and Time 2. The standard deviation was similar for both groups. 116

PAGE 128

Table 34 Characteristics of Attitudinal Items at Different Times Subscale Time 1 (n = 39) Time 2 (n = 39) Peer-Pressure Mean 3.84 4.01 Standard Deviation 0.67 0.53 Skewness -0.35 -0.21 Kurtosis -0.21 -0.69 Missing Responses 1 0 Abstinence Mean 3.58 3.49 Standard Deviation 0.71 0.75 Skewness -0.63 0.03 Kurtosis 0.90 -0.59 Missing Responses 0 0 Condom use Mean 3.26 3.52 Standard Deviation 0.65 0.62 Skewness 0.80 0.89 Kurtosis 1.19 0.56 Missing Responses 2 4 Drug use Mean 3.74 3.79 Standard Deviation 0.61 0.58 Skewness -0.08 0.05 Kurtosis -0.42 0.45 Missing Responses 0 0 Threat of HIV Mean 3.89 3.80 Standard Deviation 0.47 0.49 Skewness -0.13 -0.45 Kurtosis 0.43 -0.23 Missing Responses 0 1 Note Response scaled ranged from 1 to 5. N is the number of responses from a sample of 483 students. Reliability of the Attitude Scores from Basen-Engquist et al.s (1999) Study Descriptive Statistics. The mean, standard deviation and other descriptive statistics for the items from Basen-Engquist et al. (1999) are presented in Table 35. The means for the six items ranged from 3.64 (SD = 1.35) to 4.49 (SD = 0.89). Additionally, the distribution of the scores was negatively skewed, and this distribution was similar to 117

PAGE 129

the one from the translated CDC items. Additionally, missing responses were found in all of the items. Item 3 (Condom subscale) had 16 non-responses and item 26 (Abstinence subscale) had 14 non-responses. As described earlier, the reasons for a non-response may be due to the lack of sexual experience and unfamiliarity with condoms. However, these reasons will have to be investigated in future research as they may affect the reliability and validity of the scores. When the items were combined to form Condom and Abstinence subscales, the mean was 4.22 (SD = 0.86) and 4.12 (SD = 0.89), respectively. Both subscales had distributions that were negatively skewed (see Table 36). The mean of Basen-Engquist et al.s (1999) Abstinence subscale was similar to the translated CDCs Abstinence subscale (4.09). In contrast, Basen-Engquist et al.s (1999) Condom subscale (4.22) was higher than the translated CDC Condom use (3.63) subscale. Overall, the reliability of the scores from Basen-Engquist et al.s Condom and Abstinence subscales was found to be adequate. Table 35 Descriptive Statistics for Basen-Engquist et al.s (1999) Items Item N Mean Standard Deviation Skewness Kurtosis Abstinence 22 475 4.49 0.89 -2.17 4.89 26 469 3.64 1.35 -0.59 -0.92 27 471 4.25 1.12 -1.54 1.42 Condom 3 467 4.14 1.16 -1.27 0.55 28 474 4.33 0.97 -1.66 2.41 29 473 4.17 1.12 -1.37 1.04 Note Positive statements: Items 22, 27, 28, and 29. Negative statements: Items 3 and 26. Response scaled ranged from 1 to 5. N is the number of responses from a sample of 483 students. 118

PAGE 130

Table 36 Descriptive Statistics of Basen-Engquist et al.s (1999) Attitudinal Subscales Factor N Mean Standard Deviation Skewness Kurtosis Condom 458 4.22 0.86 -1.31 1.49 Abstinence 463 4.12 0.89 -0.97 0.29 Note. N is the number of responses from a sample of 483 students. Internal consistency. As described in Table 37, the Basen-Engquist et al. (1999) Condom and Abstinence subscales, which were used as part of the concurrent validity analysis, had three items each. Internal consistency was .68 for the Abstinence and .69 for the Condom subscales. Therefore, the reliability of the scores for the items in this subscale indicated that the items were functioning together as a group. Table 37 Cronbach Alpha and Item-to-total Correlation of Basen-Engquist et al.s (1999) Subscales Scale N Number of Items Cronbachs Alpha Item-to-total Correlation (Range) Basen-Engquists Abstinence 463 3 .68 .45 to .57 Basen-Engquists Condom 458 3 .69 .40 to .58 Note. N is the number of responses from a sample of 483 students. Test-retest. The scores of the sample of 39 students had a reliability coefficient of .78 for the Abstinence subscale and .67 for the Condom subscale. Both reliability coefficients were greater than the ones observed for the translated CDC subscales. The 119

PAGE 131

means were compared for Time 1 and Time 2 (see Table 38). Both means were very close in value and had a similar standard deviation. Table 38 Comparison of Basen-Engquist et al.s (1999) Condom and Abstinence Subscales at Different Times Factor Time 1 (n = 39) Time 2 (n = 39) Abstinence Mean 3.58 3.49 Standard Deviation 0 .71 0.75 Skewness 0.63 0.03 Kurtosis 0.90 -0.59 Condom use Mean 3.20 3.52 Standard Deviation 0.65 0.62 Skewness 0.80 0.89 Kurtosis 1.19 0.56 Study 5 Confirmatory Factor Analysis Model 1 To evaluate the factorial validity of the five-factor attitudinal model underlying the CDC instrument, the Calis statistical program from the SAS Institute (2001) was used to perform the confirmatory factor analysis (CFA). The Calis procedure makes use of structural equations in the same manner as the LISREL program for the analysis of data with latent variables (Hatcher, 1994). Maximum likelihood estimation, using the variance-covariance matrix of the observed variables, was used to determine the relationship of the items with the latent or unseen variables (Hatcher, 1994). Fit indices. Figure 3 presents the five-factor correlated model that was tested. The results from CFA indicated that a five-factor attitudinal model did not fit the data. Several fit indices were consulted as recommended by Bryne (1994) and Hatcher (1994). As 120

PAGE 132

described in Table 39, the first index, the 2 (220, N = 408) = 586, p < .0001, was statistically significant. Therefore, the hypothesis of the model fitting the data was rejected. However, Bryne (1994) has noted that sample size affects the chi-square test and that other indices should be consulted before making a final decision. Table 39 included the goodness of fit indices such as Bentler and Bonetts Non-normed Index and Bentlers Comparative Fit Index. The values for the previous indices were .69 and .72, respectively. Values greater than .9 have generally been viewed as indicative that a model has an acceptable fit. Further, the Root Mean Square Error of Approximation (RMSEA) had a value of .064. An RMSEA value between 0 and .05 indicates the model fits the data (Bryne, 1994). Values between .05 and .08 indicate marginal fit. Therefore, the interpretation of the fit indices suggested the fit of the model was less than acceptable. Table 39 Goodness of Fit Indices to Examine Model Fit Fit index Value Chi-square 586 Chi-square degrees of freedom 220 p value .0001 Bentler & Bonetts (1980) Non-normed Index .66 Bentlers comparative fit index .71 RMSEA .064 Note. RMSEA is the Root Mean Square Error of Approximation. 121

PAGE 133

Threat of HIV Drug use Condom use Abstinence Pee r p ressure 1 6 11 15 19 2 7 12 20 16 21 23 4 9 13 17 24 5 10 8 14 18 Figure 3. Standardized Fact or Loadings for Model 1. 25 .43 E1 E6 E11 E15 E19 E2 E7 E12 E20 E8 E16 E21 E23 E4 E9 E13 E17 E24 E5 E10 E14 E18 E25 .43 .46 .40 59 .54 .26 .68 .63 .42 -.13 .47 .20 .35 47 .33 .66 -.06 .27 .31 .45 ..04 .43 122

PAGE 134

Standardized factor loadings. The standardized factor pattern coefficients (i.e., loadings), error variances, and t-test for significance were evaluated for each variable to determine sources of misfit. As demonstrated in Figure 3, the standardized factor loadings ranged in size from -.13 to .68. The data indicated a non-significant (p > .05) standardized factor loading for items 24 (Drug use subscale) and 18 (Threat of HIV subscale). The standardized factor loadings for items 1, 6, 11, and 15 from the Peer-pressure subscale were .43, .43, .46, .40, and .59, respectively. Items 2, 7, 12, and 20 from the Abstinence subscale had standardized factor loadings of .54, .26, .68, and .63, respectively. Further, items 8, 16, 21, and 23 from the Condom subscale had standardized factor loadings of .42, -.13, .47, and .20, respectively. In contrast, items 4, 9, 13, and 24 had loadings below .4 for the Drug subscale. Similarly, items 5, 10, 18, and 25 were weakly associated with the Threat of HIV subscale. Therefore, the items in the Abstinence subscale had a strong association to their factor and the rest of the items had a weak association to their factors. 123

PAGE 135

Table 40 Standardized Factor Loading, Error Variance, and Squared Multiple Correlation Coefficient of the Attitudinal Variables from Model 1 Subscale Standardized Loading Uniqueness t-Value R 2 Peer-pressure 1 0.43 0.04 7.77 .19 6 0.43 0.04 7.72 .18 11 0.46 0.08 8.31 .21 15 0.40 0.06 6.54 .14 19 0.59 0.05 10.83 .35 Abstinence 2 0.54 0.07 9.78 .30 7 0.26 0.06 4.53 .07 12 0.68 0.07 12.21 .47 20 0.63 0.06 11.32 .40 Condom use 8 0.42 0.08 6.75 .17 16 -0.13 0.07 -2.28 .02 21 0.47 0.06 7.35 .22 23 0.20 0.07 3.58 .04 Drug use 4 0.35 0.06 6.02 .12 9 0.47 0.07 8.27 .22 13 0.33 0.06 5.77 .11 17 0.66 0.04 11.45 .43 24 -0.06 0.08 -1.07* .004 Threat of HIV 5 0.27 0.05 5.30 .07 10 0.31 0.05 5.95 .09 14 0.45 0.04 8.27 .20 18 -0.04 0.06 -0.93* .002 25 0.43 0.04 8.06 .19 Note. Items 18 and 24 did not have a statistically significant factor loading (p > .05). Correlation of the factors. The confirmatory factor analysis (CFA) produced a correlation matrix of the latent factors which indicated that two correlations were inadmissible (see Table 41). The Peer-pressure and the Threat of HIV subscales had a correlation of 1.01, an inadmissible value. Similarly, the correlation between the Threat of HIV and the Condom use subscale was 1.26 (see Table 41). A correlation this high was problematic, and it indicated a potential problem with the solution derived for the five-factor model. 124

PAGE 136

Table 41 Correlation of the Five Factors Peer-pressure Abstinence Condom Drugs Threat of HIV Peer-pressure 1.00 Abstinence .52 1.00 Condom .76 .76 1.00 Drugs .74 .52 .72 1.00 Threat of HIV 1.01 .64 1.26 .95 1.00 Modification of the model. Major sources of misfit occur when an item wants to load on more than one item, and when there is correlation of the errors for pairs of items on more than one factor. The model assumes that measurement error for pairs of items are uncorrelated. Theoretically, the measurement error is random and random errors by definition are not correlated. Sources of misfit can be identified by using the modification feature in the Calis program. The Lagrange test is a modification index that estimates the reduction in chi-square if additional paths are added to the model (Hatcher, 1994). It predicts how a model will change by altering the paths among the variables; however, new paths must be inspected to determine if they are theoretically sound. Because the original model was inadequate, five modifications were proposed and the new models were subsequently tested. The new models were: (1) Model 2, removed items 18 (Threat of HIV) and 24 (Drug use); (2) Model 3, the exact replica of Model 2 with the exception of allowing 10 measurement errors to be correlated; (3) Model 4, separate analyses of each of the five subscales with CFA; (4) Model 5, eliminated the 125

PAGE 137

Condom subscale from the original model; (5) Model 6 similar to Model 5, with the exception of allowing nine measurement errors to be correlated. As summarized in Table 42, the Lagrange test suggested that a new path from an item to other factors and from one error to another correlated error would improve fit. Items that loaded on the Condom (item 21), Abstinence (item 7), and Drug use (item 13) factor were recommended by the Lagrange test to also load on the Peer-pressure factor. Specifically, item 21, which was associated with the Condom factor, had a secondary loading on both the Threat of HIV and Abstinence factor. Item 2, which was associated with the Abstinence factor, had a secondary loading on the Drug factor. Table 42 New Paths Suggested by the Lagrange Test New Paths from an Item to a Factor Item Subscale Change in 2 23 (Condom) Peer-pressure 16 21 (Condom) Peer-pressure 15 21 (Condom) Threat of HIV 14 7 (Abstinence) Peer-pressure 12 21 (Condom) Abstinence 12 13 (Drugs) Peer-pressure 9 2 (Abstinence) Peer-pressure 8 7 (Abstinence) Drugs 7 2 (Abstinence) Drugs 7 16 (Condom) Threat 6 New Paths from One Error to Another Error Error Error Change in 2 E6 (Peer-pressure) E1 (Peer-pressure) 35 E19 (Peer-pressure) E11 (Peer-pressure) 34 E10 (Threat) E7 (Abstinence) 29 E7 (Abstinence) E6 (Peer-pressure) 24 E10 (Threat) E13 (Drug) 24 E23 (Condom) E7 (Abstinence) 21 E13 (Drug) E7 (Abstinence) 18 E13 (Drug) E21 (Condom) 16 E9 (Drug) E8 (Condom) 15 E18 (Threat) E23 (Condom) 14 Note. Factors in parentheses indicate the original designation. 126

PAGE 138

These new paths provided statistical evidence that items were associated to more than one factor and supported the conclusion that some of the items were multi-dimensional. Items that were multi-dimensional may have affected the correlation among the factors. Multi-dimensionality was evaluated by inspecting the paths for practical significance. Afterwards, it was determined that new paths with items 21 and 23 (both Condom use items) had a small degree of practical significance, but that it was not sufficient to include them in a new model. Table 43 Goodness of Fit Indices of Model 1 and Model 3 Model 1 Model 3 Fit Index Value Value Chi-square 586 322 Chi-square degrees of freedom 220 170 p value .0001 .0001 Bentler & Bonetts (1980) Non-normed Index .66 .84 Bentlers Comparative Fit Index .71 .87 RMSEA .064 .04 Note Model 1 was the original five-factor model. Model 2, was similar to Mode1 l with the exception that items 18 (Threat of HIV) and 24 (Drug use) were removed. Model 3 was similar to Model 2 except that 10 measurement errors were allowed to be correlated. Model 2 and Model 3 Model 2 was similar to the first, but it differed in that two statistically non-significant items (18 and 24) were removed. The model was tested and the findings indicated Model 2 did not fit the data. A third model was tested, and, as with the second model, it did not include items 18 and 24, but it did allow the 10 errors in Table 42 to be 127

PAGE 139

correlated. The indices from CFA improved, but the fit was still marginal. Table 43 compares both models, and the information indicated that Model 3 was an improvement over the first model. Indices such as Bentler and Bonetts Non-normed Index (.84) and Bentlers Comparative Fit Index (.87) were higher in the third model than in the first model. Further, the Root Mean Square Error of Approximation (RMSEA) was .04, which indicated acceptable fit. Therefore, to make a confident conclusion, other indicators were examined to determine how well the model fit the data. Table 44 Standardized Factor Loadings for Model 1 and Model 3 Subscale/Item Model 1 Standardized Factor Loading Model 3 Standardized Factor Loading Peer-pressure 1 .43 .39 6 .43 .37 11 .46 .38 15 .40 .35 19 .59 .52 Abstinence 2 .54 .54 7 .26 .22 12 .68 .68 20 .63 .63 Condom use 8 .42 .42 16 -.13 -.12 21 .47 .48 23 .20 .20 Drug use 4 .35 .34 9 .47 .49 13 .33 .30 17 .66 .69 24 -.06 Item was deleted Threat of HIV 5 .27 .27 10 .31 .28 14 .45 .45 18 -.04 Item was deleted 25 .43 .45 128

PAGE 140

The standardized factor loadings for Model 1 and Model 3 were examined (see Table 44), and the values were very similar across the same items. The loadings were significant (p < .05) for all the items in Model 3. However, the standardized factor loadings in Model 3 did not improve, and therefore the strength of association between the items and their factors was not altered by allowing the errors to be correlated. Based on this evidence, correlated errors may be affecting the fit of the model, but including these parameters into the model was not enough to achieve excellent fit. Model 4. Confirmatory Factor Analysis of Each Subscale The five factors were tested separately with five different CFAs to determine how the model fit the data. Three of the five models were determined to have an adequate fit. The Non-normed Index, Comparative Fit Index, and the RMSEA for the Abstinence subscale indicated that the model fit the data (see Table 45). In particular, items 2, 12, and 20 were found to have a statistically significant factor loading and the loadings were strong. This information and the results from Model 1 suggested the three items were adequate for measuring the abstinence dimension. The CFA of the Condom use subscale indicated the model fit the data. The goodness of fit indices were .89 for the Non-normed index, .96 for the Comparative Fit Index, and .04 for the RMSEA. However, the standardized factor loadings for the Condom use subscale were unacceptable. Even after properly reversing the item, item 16 had a negative standardized loading, and item 23 did not have acceptable standardized factor loadings (< .4). The standardized factor loadings for items 8 and 21 in the Condom subscale were .50 and .40, respectively. The CFA of the Drug use subscale indicated the model fit the data, and a similar situation existed for the relationship among the Drug use 129

PAGE 141

subscale and its items. For example, items 13 and 24 did not have acceptable standardized factor loadings and only one item (# 17) had a standardized factor loading of .70. A stronger relationship between the factors and the items was expected for this analysis. Table 45 Summary of the Goodness of Fit Indices for the CFA of the Five Subscales Subscale/ Item p value Bentler & Bonetts Non-normed Index Bentlers Comparative Fit Index RMSEA Conclusion Peer-pressure (1, 6, 11, 15, 19) .0001 .43 .71 .17 Data did not fit Abstinence (2, 7, 12, 20) .78 1 1 0 Data fit Condom use (8,16, 21, 23) .18 .89 .96 .03 Data fit Drug use (4, 9, 13, 17) .06 .89 .94 .04 Data fit Threat of HIV (5, 10, 14, 25) .01 .67 .83 .06 Data did not fit As described in Table 46, goodness of fit indices associated with the Peer-pressure subscale indicated that the model did not fit the data. Items 1, 6, and 15 had moderate standardized factor loadings. The other two Peer-pressure items were less than .4. A final CFA was performed on the Threat of HIV subscale and the results indicated that the model did not fit the data. Three of the items in the Threat of HIV subscale had unacceptable standardized factor loadings, and the loadings for the other two items were small. 130

PAGE 142

Table 46 Summary of the Standardized Factor Loadings from Separate CFAs Factor/Item Standardized Factor Loading Uniqueness Peer-pressure 1 .63 .77 6 .64 .76 11 .23 .97 15 .42 .90 19 .35 .93 Abstinence 2 .58 .81 7 .19 .98 12 .65 .75 20 .62 .77 Condom use 8 .50 .86 16 -.36 .93 21 .40 .91 23* Drug use 4 .35 .93 9 .44 .89 13 .28 .95 17 .70 .70 24 -.14 .99 Threat of HIV 5 .31 .94 10 .31 .95 14 .40 .91 18* 25 .49 .86 Note indicates an item with a non-significant factor loading (p > .05). Model 5 and Model 6. Confirmatory Factor Analysis Model 5. The items from the Condom subscale were found to be inadequate for use as a subscale based on information from the adaptation, pretest, and the lack of internal consistency (alpha = 0). Those items were eliminated from Model 5. As illustrated in Figure 4, the 19 translated items in Model 5 were assumed to represent four dimensions: Peer-pressure, Abstinence, Drug use, and Threat of HIV. The goodness of fit indices from CFA determined the model did not fit the data. 131

PAGE 143

Threat of HIV Drug use Abstinence Pee r p ressure 1 6 11 15 19 2 7 12 20 9 13 17 24 10 14 18 25 5 4 Figure 4. Proposed four-dimensional at titudinal structure in Model 5. E1 E6 E11 E15 E19 E2 E7 E12 E20 E4 E9 E13 E17 E24 E10 E14 E18 E25 E5 132

PAGE 144

The results revealed that items 18 (Threat of HIV) and 24 (Drug use) had non-significant factor loadings. The largest standardized factor loadings were observed in item 2 (.55), item 12 (.66), item 17 (.62), and item 20 (.63). It was noted that a correlation among the Peer-pressure and Threat of HIV was 1.01 (an inadmissible value), and the correlation among the Drug use and Threat of HIV was .98. The findings were consistent with the high correlations in Model 1. Furthermore, the Lagrange test identified new paths from one error to another correlated error. It was hypothesized that adding new paths allowing pairs of errors to correlate would improve the models fit. Model 6. Model 6 was similar to the previous model, but it differed in that several measurement errors were allowed to correlate. As described in Table 47, nine errors were allowed to be correlated in the model. The Comparative Fit Index, Non-normed Index, and the RMSEA were .90, .88, and .04, respectively. Based on the fit indices, it was concluded that the model fit the data. The internal structure of the 19 items was therefore determined to have four different dimensions with nine measurement errors. Correlated errors. Correlated errors are common in instruments that assess attitudes. They can occur for two items within the same factor or in two items in two different factors. As described in Table 47, the Peer-pressure subscale had three sets of correlated errors: item 1 and item 6; item 11 and item 19; and item 11 and item 6. Correlated errors are common in items that are next to each other in questionnaires. In this study there were two sets of errors next to each other: item 6 and item 7; and item 17 and item 18. Other errors occurred for items in two different factors: Abstinence (item 7) and Threat of HIV (item 10), Threat of HIV (item 10) and Drug use (13), and Abstinence (7) and Drug use (13). 133

PAGE 145

Table 47 Description of New Paths Among the Errors in Model 6 Error(Subscale) Error(Subscale) Expected Change in 2 E6 (Peer-pressure) E1 (Peer-pressure) 38 E19 (Peer-pressure) E11 (Peer-pressure) 32 E10 (Threat of HIV) E7 (Abstinence) 28 E7 (Abstinence) E6 (Peer-pressure) 25 E10 (Threat of HIV) E13 (Drug use) 21 E18 (Threat of HIV) E17 (Drug use) 16 E13 (Drug use) E7 (Abstinence) 15 E11 (Peer-pressure) E6 (Peer-pressure) 12 E5 (Threat of HIV) E4 (Drug use) 12 Standardized factor loadings. Unlike a strong model that has high-standardized factor pattern coefficients (i.e., loadings), this model had many low standardized factor loadings. Peer-pressure items 1, 6, 11, and 15 had a weak association to the factor, and item 19 had a strong association with the factor. This information was consistent with the qualitative results of the adaptation and is useful in explaining the lower loadings. For example, items 1 and 6 were identified as marginally representing the Peer-pressure dimension. Three items from the Drug use subscale had non-significant loadings that were below .4. A similar pattern was observed for the Threat of HIV subscale: item 18 had a non-significant loading, and items 5 and 10 had unacceptable loadings. Only two items, 14 and 25, had marginal loadings for the Threat of HIV subscale. Based on the marginal standardized factor loadings, the model was not very strong, and the evidence related to the construct validity of the scores was therefore not strong. 134

PAGE 146

Table 48 Standardized Factor Loadings for Model 6 Subscale/ Item Standardized Factor Loading Uniqueness Peer-Pressure 1 .40 .91 6 .41 .91 11 .38 .92 15 .36 .92 19 .50 .86 Abstinence 2 .56 .82 7 .27 .96 12 .67 .73 20 .62 .77 Drug use 4 .36 .92 9 .45 .89 13 .35 .93 17 .65 .75 24 Threat of HIV 5 .26 .96 10 .33 .94 14 .47 .88 18 25 .46 .88 Note indicated a non-significant standardized factor loading. Study 5 Concurrent Validity To obtain a measure of concurrent validity, six items developed by Basen-Engquist et al. (1999) to measure attitudes toward condom use and abstinence were translated and treated in parallel with the translated CDC items. The attitudinal items from the translated CDCs Abstinence and Condom subscales were correlated with their respective subscales developed by Basen-Engquist et al. (1999). As reported in Table 49, a correlation of .44 was observed between the translated CDC Condom subscale and the Condom subscale from Basen-Engquist et al. (1999). 135

PAGE 147

Table 49 Correlation Between Basen-Engquists Subscales and CDCs Items (n = 483) CDC Scale Basen-Engquist Condom Basen-Engquist Abstinence Condom .44 Abstinence .67 Note. Pearson Product-moment Correlation Coefficients The correlation between the translated CDC Abstinence subscale and Basen-Engquist et al.s Abstinence subscale was .67. A strong correlation with the translated CDC Abstinence subscale provided evidence that the scales were measuring a similar construct. In contrast, the translated CDC Condom subscale had a smaller correlation with the Condom subscale from Basen-Engquist et al.'s study. Although this correlation was of low to moderate strength, it should be noted that with the low reliability of the translated CDC Condom subscale scores, this correlation has been weakened by the low reliability. Summary of the Results In conclusion, this section presented the results of a multi-phased study. The results of two forward-translations, a back-translation, and a synthesis were presented. Next, the results were presented for three different panels that reviewed the instrument to cross-culturally adapt it for use with Salvadorian adolescents. The instrument was pretested with two groups of students before it was administered to 483 high school students in San Salvador, El Salvador. Finally, the results of the reliability study, factorial validity, content validity, and concurrent validity were presented. 136

PAGE 148

CHAPTER FIVE DISCUSSION Purpose of the Study The purpose of this multi-phased study was to translate, cross-culturally adapt, and validate the scores from an instrument that assessed the HIV/AIDS knowledge and attitudes of high school students in El Salvador. The first phase utilized the back-translation method to obtain the first Spanish version of the instrument, which was developed by the CDC to evaluate HIV/AIDS knowledge and attitudes of students in grades 7 to 12. In the next phase, an expert panel of reviewers from El Salvador established the conceptual equivalence of the Spanish instrument. In the third phase, the instruments content validity and cultural acceptability were evaluated by a panel of HIV experts, health professionals, schoolteachers, and students from El Salvador. The last phase involved collecting evidence of the validity of the scores from the Spanish version of the CDC instrument; 483 public high school students in the metropolitan area of San Salvador were randomly selected for the validation phase. Reliability of the scores was measured by the test-retest method and coefficient alpha. Confirmatory factor analysis was used to evaluate if the attitudinal construct had five different dimensions (Peer-pressure, Condom use, Abstinence, Drug use, and Threat of HIV). As part of the concurrent validation process, the Condom and Abstinence subscales from the translated CDC instrument were correlated with a set of existing Attitudinal items from a published study by Basen-Engquist et al. (1999). 137

PAGE 149

The discussion is organized around six specific purposes. These purposes were to: (1) determine the success of the cross-cultural adaptation by contrasting the first translated version of the CDC instrument with the final Spanish version of the CDC instrument; (2) determine, using a Salvadorian panel, the level of cultural appropriateness of the translated CDC HIV/AIDS Knowledge and Attitudinal items for use in El Salvador; (3) ascertain the level of readability of the final Spanish version of the CDC instrument and evaluate content validity; (4) evaluate the reliability of the scores from the Spanish version of the CDC instrument using a sample of Salvadorian students in grades 10, 11, and 12; (5) determine if the Spanish version of the CDC instrument contains five Attitudinal dimensions (Peer-pressure, Condom use, Abstinence, Drug use, and Threat of HIV) when analyzed using confirmatory factor analysis; and (6) evaluate the concurrent validity of the Abstinence and Condom use subscales from the Spanish version of the CDC instrument. Purpose One: Determine the Success of the Cross-cultural Adaptation by Contrasting the First Translated Version of the CDC Instrument with the Final Spanish Version of the CDC Instrument A comparison of the first translated version of the instrument (see Appendix N) with the final translated (see Appendix Q) version provided information on cultural differences and served as a measure of the success of the adaptation. The first minor differences were in the written assurance to guarantee anonymity in the directions and the instructions about how to mark the answers on the document. These differences were due to the bilingual reviewers request not to duplicate the information in the informed consent and were not related to cultural differences. The final version reiterated the 138

PAGE 150

anonymity guaranteed throughout the study and perhaps improved the outcome since anonymity can be a precursor to providing honest answers. Differences in the Knowledge section. The Knowledge section had seven minor and three major differences. Minor differences in the final version included the addition of the term VIH/SIDA to portray the terms HIV and AIDS. The original English version differentiated HIV the virus from AIDS the disease. In El Salvador, the virus and disease are grouped under an umbrella term that facilitates understanding the concepts related to the disease. By combining both terms the focus of the items was shifted to assessing general and not specific knowledge. Another minor difference was the use of the term relaciones sexuales to denote having sex; the term not only defined the concept, but it also was culturally relevant and appropriate for the adolescent population. The term thus improved the instrument by identifying the correct risk factor for this population. Minor changes were made to the final version consisting of editing one or two words in the following items: 7 (chorro), 9 (consumo de drogas), 10 (maternal), 11 (portador), 12 (se ven enfermizas), and 14 (zancudo). The four bilingual reviewers overlooked these words. The panelists, on the other hand, identified them as colloquialisms from El Salvador, and these changes clarified the meaning and increased the instruments cultural appropriateness. The addition of these words met the idiomatic equivalence recommended by Guillemin et al. (1993) and substantiated the usefulness of the panels in the methodology. The major differences were observed in Knowledge items 1, 2, and 15. The double negative in the first version (item 1) was eliminated to improve readability and 139

PAGE 151

reduce the chance of students passing over it or guessing. This was less of a cultural difference and more of an attempt to improve a grammatically difficult item. A cultural difference existed in how HIV testing is promoted in the United States and El Salvador. The term antibody test in item 2 is common in the United States, yet in El Salvador it is not, and every student in the pretest misunderstood the Spanish translation. This exemplified the utility of the pretest in combination with a panel review and highlighted the difficulty of unilaterally obtaining a concepts semantic equivalence. Additionally, the use of the probing technique recommended by Guillemin el al. (1993) was invaluable in establishing semantic equivalence. The translation of the term lambskin condoms in Knowledge item 15 was difficult and labeled as problematic by the two translators and bilingual reviewers. A Salvadorian panelist indicated the term was not familiar in El Salvador because, unlike in the United States, only one type of condom is promoted in El Salvador. The cultural difference in this item was eliminated by rewording it to assess the role of latex condoms in the prevention of HIV. Students were therefore prevented from passing over it or guessing. Differences in the Attitude section. Differences between the two versions in Attitudinal items 1, 9, 16, and 21 suggested that difficulties resulted from a combination of the complexity in the original construct, cultural differences, and an inexact translation. The translation of item 1 did not completely retain the English versions concept of peer-pressure, and even after several revisions, the interpretation was still not clear. Item 1 was labeled as problematic and the Lay Panel attempted to refine the 140

PAGE 152

concept of peer-pressure by focusing on words that conveyed the idea of having to do something wrong for ones friends. This item was slightly problematic for the students in the pretest. They described it as vague, but did mention that it had a peer-pressure component to it. This suggested a cultural difference in how peer-pressure is interpreted among students. Students in El Salvador favored a concrete definition, one with vivid examples. Therefore, the differences appear to be related to an interaction of factors and provide support for Hui and Triandiss (1985) view that the conceptual/functional equivalence of a construct is dependent on its level of abstraction. In this case, the Peer-pressure item was given a global and abstract definition by the developers of the instrument, and, when tested, the students demanded a more concrete example. Another difference centered on the substitution of the word steroids for the word tattoos in item 9 (final version). The substitution, according to one panelist, made the item culturally relevant as it reflected a new trend among high school students in the public schools in El Salvador. Next, item 16 was improperly translated in the first version and was retranslated to adequately express the concept that condoms do not make sex less pleasurable. The new translation was adequate for sexually active students who use condoms, but it introduced a new dilemma. During the administration of the instrument, three female students said they did not understand the question; therefore, this item was determined to be inappropriate for sexually inexperienced youth. The divergence in item 21 was due to the navet of the four bilingual reviewers with regard to the dating practices of adolescents in El Salvador. The bilingual reviewers projected the view that adolescents are likely to have sexual intercourse during a date, 141

PAGE 153

which is more true in the United States than in El Salvador. Guillemin et al. (1993) have stressed the importance of capturing the experiential equivalence of the target culture to correctly adapt an instrument. The role of sexual relations was not raised as an issue by the panelists, but the students reported that sexual relations occurred mostly in long-term relationships. The students suggested changing the item to reflect the situation in El Salvador. This suggestion provides support for the methodology used to adapt the instrument in El Salvador. The studys methodology shared several components from the IQOLA (Bullinger et al., 1998; Gandek & Ware, 1988; Wagner et al., 1998). The use of experts was one of them. The experts ability to modify the instruments content so that the concepts were understood by the target population proved to be advantageous. Salvadorian experts made the necessary adjustments to reduce cultural differences. The differences revolved around colloquialisms that adolescents would understand and the clarification of the peer-pressure concept. Differences were also found in specific topics, such as the use of latex condoms, dating practices of adolescents, and the emergence of tattoos in students. These topics were adjusted to reflect the current situation in El Salvador, tested with students, and found to be culturally acceptable. The panels rating of the HIV/AIDS content provided evidence that supported the effectiveness of the methodology. The panel determined that the content was valid and culturally acceptable for use with Salvadorian adolescents. Furthermore, the differences between the two Spanish versions highlighted the necessity of including the Salvadorian panels to complement the work of the bilingual reviewers. 142

PAGE 154

Purpose Two: Determine, using a Salvadorian Panel, the Level of Cultural Appropriateness of the Translated CDC HIV/AIDS Knowledge and Attitudinal Items for use in El Salvador Brislin et al. (1973) and Guillemin et al. (1993) reported on the benefit of expert panels to culturally adjust an instrument's content. The procedure in this study made use of three different panels. The intent of the first two panel reviews in the adaptation phase was to improve the wording of the items and increase the cultural acceptability of the instrument. The inclusion of the last panel was based on the premise that an association with the target population provides the necessary knowledge about the cultural characteristics of the students and school environment. Therefore, a panel formed by Salvadorian teachers, health workers, and HIV professionals rated the cultural appropriateness of the adapted items for use in high schools in El Salvador. Culture encompasses a very broad definition (Brislin et al., 1973), and there is no agreement among scientists. A cultural adaptation is therefore dependent on persons who have experience living in the target culture. The advantage of including Salvadorian professionals is that they uphold the standards of the community by acting as the gatekeepers of information for the younger population. They have the responsibility to identify culturally inappropriate items that can be deemed as offensive for certain cultures. Also, they can identify items that are appropriate for the target population. This issue is particularly relevant for HIV/AIDS researchers who depend on instruments with sexually-related content. The panelists quantitative evaluation established the cultural appropriateness of the content for use with high school students in El Salvador. 143

PAGE 155

Purpose Three: Ascertain the Level of Readability for the Final Spanish Version of the CDC Instrument and Evaluate Content Validity Readability. Not having the correct reading level for the intended audience of the instrument can introduce systematic errors that result when items are misinterpreted. Additionally, when items are too difficult to read, a student may not answer the question or may guess. To minimize this effect, the translators and the four bilingual reviewers were asked to use simple words when representing the constructs. As an additional measure of quality, the Salvadorian panels substantiated the translation during the review process. When items were revised, the panelists were instructed to include words or sentences that were at a high school reading level. Because of this procedure, the final version was expected to be at an appropriate reading level. The readability of this instrument was assessed independently by the principal of the Montessori Academy in San Salvador and by a high school psychologist/sexuality education teacher. They concluded the reading level was appropriate for high school students. As an additional measure, the Salvadorian Lay Panel qualitatively analyzed the reading level and concurred that it was at a high school level. Content validity. The two-step process to establish content validity served to verify the importance of the items for use in El Salvador. First, the panelists rated each items appropriateness toward the HIV/AIDS constructs. Second, they were asked to provide additional items needed to represent the domains. The content found in the Knowledge and Attitude section was found to be acceptable. The high ratings indicated all the items were important to the domain, and the lack of additional items supported the conclusion that the domain was adequately represented in the final instrument. 144

PAGE 156

The inclusion of content matter experts in the review of an instrument is a strategy reported by Aday (1996). The process of identifying non-essential items was similar to Jemmott et al.s (1992) study in which experts used a Likert type-scale to accept or reject items. One advantage of this process was the combination of two strategies: the solicitation of additional items and a rating system. An additional advantage was the inclusion of 15 Knowledge and 29 Attitudinal items that were previously validated (content) and found by the CDC (2002b) to adequately cover the HIV/AIDS domain. Finally, the results of the content validation study demonstrated that the HIV/AIDS content was similar in both cultures. The results are in agreement with a prediction made by Hui and Triandis (1985) that a cross-cultural adaptation has a greater chance of success when the construct is familiar to both cultures. In this case, the ongoing AIDS epidemic has not only acquainted more individuals with the HIV/AIDS construct, but it has heightened awareness of the role of risk factors associated with HIV infection. Specifically, El Salvadors HIV/AIDS epidemic has been addressed through governmental and non-governmental educational programs that attempt to contain its transmission. As a result of these efforts, many qualified HIV professionals were available to assist with the adaptation and validation. According to Guillemin et al. (1993), the inclusion of experts familiar with the target culture is a vital part of a cross-cultural adaptation. In this study, their familiarity with the culture underscores their role in the content validation of items for use in the public high schools in El Salvador. 145

PAGE 157

Purpose Four: Evaluate the Reliability of the Scores from the Spanish Version of the CDC Instrument Using a Sample of Salvadorian Students in Grades 10, 11, and 12 Reliability of the Knowledge Scores. Two methods were used to calculate the reliability of the scores from the Knowledge section. The first was the test-retest method, and the second was the determination of internal consistency by calculating Cronbachs alpha. A subset (n = 39) of the population was administered the same instrument twice in a one-week period for the test-retest study. The Pearson product-moment correlation coefficient between the two scores was .59. According to Crocker and Algina (1986), a minimum acceptable value has not been set. Yet for this study, a higher value was expected because HIV/AIDS knowledge does not change during a one-week period unless a student has been exposed to a new program. Several factors were responsible for the low value of the test-retest correlation. One factor was the grading procedure that allowed a statement which was true to have two acceptable answers, I know it is true or I think it is true (the reverse was followed when the statement was false). The rationale for this scale was that it discourages a guess response and provides a global measure of HIV/AIDS knowledge. However, two correct scores reduced the variability of the scores and consequently lowed alpha. Another factor was that several items were answered correctly by more than 90% of the participants and this also lowers alpha. The scores from the Knowledge section (n = 483) had an internal consistency (Cronbachs alpha) of .57, and this value was lower than the traditionally accepted value of .70 (Hatcher, 1994). As previously described, two correct answers affects the variance and, as a consequence, the alpha value was lower than expected. The same analysis with 146

PAGE 158

one correct answer resulted in a higher Cronbachs alpha of .74. This value was acceptable and indicated that the scores were moderately reliable. Additionally, the results with one correct answer were similar to Jemmott et al.s (1992), who reported a coefficient alpha of .73 for scores in a population of adolescents. The use of two correct responses was responsible for lowering alpha. Item difficulty and discrimination also provided information that can be used to improve the internal consistency of the scores. Discrimination is a measure of how an item separates the higher scoring students who know the material against the scores of the lower scoring students or those who do not know the material (Crocker & Algina, 1986). The item-to-total correlation was used as an index of discrimination. Ideally, an item-to-total correlation > .3 was desirable, and it was noted that items 6, 7, 8, and 13 had a favorable index of discrimination. These items assessed general knowledge about the transmission of HIV, and the index suggested they differentiated among the low and high scoring groups. In contrast, items 3, 9, 11, 12, and 15 had marginal discrimination (item-to-total correlation < .2) that implied a problem with the items. Items 9 and 11 were answered correctly by most of the students, and because their variance was lower, also lowered alpha. If items 9 and 11 were eliminated, the value of alpha would drop from .57 to .56. The two items were retained because their removal would not have improved alpha and there were only 15 items in the instrument. Given high quality items, longer tests have a higher reliability coefficient than shorter tests (Crocker & Algina, 1986). The reliability of the scores could be increased by adding more items. The Spearman-Brown prophecy formula indicated that increasing the number of items to 50 would yield a reliability coefficient of .75. However, adding 35 147

PAGE 159

more items is not practical for an instrument that assesses HIV/AIDS knowledge in an adolescent population, especially when researchers and program evaluators often include other measures (e.g., attitude instruments) when studying adolescents. To summarize, the reliability of the 15 Knowledge scores was moderate for the sample of Salvadorian high school students. The weakness of the instrument was that it contained a mixture of easy and difficult items. That restricted the variance in these items. This was not a concern because the items were representative of the important sources of HIV transmission, risk behaviors, and myths that are taught in many HIV prevention programs. Reliability of the Attitudinal scores. Test-retest reliability and internal consistency reliability were calculated for the Attitudinal scores. The reliability coefficients obtained in the test-retest were found to be inconsistent across the different Attitudinal subscales. The scores from the Peer-pressure subscale yielded the highest reliability coefficient (.62). The reliability coefficients (in decreasing value) of the other subscales were .58 for the Drug use, .49 for the Threat of HIV, .43 for the Abstinence, and lastly, .17 for the Condom use (described in Table 50). In contrast, the reliability coefficients for the scores from Basen-Engquist et al.s (1999) Abstinence and Condom use were .78 and .67, respectively. The HIV/AIDS attitudes were not expected to change in a one-week period unless the students were exposed to an educational program during that week. The large amount of random measurement error is reflected in the low internal consistency reliabilities at each time point. Therefore, the low test-retest reliability may be due to the large amount of random measurement in the scales at each time point. As described in Table 50, the 148

PAGE 160

scores of the Condom and Abstinence subscales from Basen-Engquist et al. (1999) had the highest Cronbach alpha for all the subscales. The values were .69 and .68, respectively. Further, an alpha of .55 for the Peer-pressure and .58 for the Abstinence indicted the reliability of the scores was acceptable. However, these values were not as high as the alpha values reported by Carvajal et al. (1999) and Silva and Ross (2003). Both researchers measured HIV attitudes of adolescents in Latin America; the former reported an alpha of .70 and the latter reported an alpha of .88. Table 50 Test-retest and Internal Consistency of the Attitudinal Subscales Subscale Number of Items Cronbach Alpha Test-retest Correlation Coefficient Peer-pressure 5 .55 .62 Abstinence 4 .58 .43 Condom 4 0.0 .17 Drug use 5 .24 .58 Threat of HIV 5 .30 .49 Abstinence (Basen-Engquist) 3 .68 .78 Condom (Basen-Engquist) 3 .69 .67 Note. A negative reliability coefficient for the condom scale was set to .00. Unfortunately, the scores of the Condom use subscale had an alpha close to zero that placed the interpretation of the scores in doubt. This suggested the responses were not correlated to each other and were not internally consistent. Crocker and Algina (1986) reported that a low alpha is indicative of poor item quality and a lack of content homogeneity. Future research will need to re-examine the Condom use items. Moreover, the internal reliability of the Drug use and the Threat of HIV subscales indicated the items did not possess strong internal consistency. Because the CDC did not publish the 149

PAGE 161

reliability of the scores from either of the subscales, the values from this study could not be compared to the CDC values. The effect of a cross-cultural adaptation on measures of reliability is uncertain and the evidence is mixed. Theoretically, if the methodology of the cross-cultural adaptation is sound and reproduces the construct for the target culture, the reliability of the scores should be acceptable. Guillemin et al. (1993) observed that an adaptation is similar to constructing a new instrument. They stressed the importance of obtaining the psychometric characteristics of the instrument with the new populations. Davis et al. (1999) reported that a successful adaptation of an instrument with 35 items yielded positive results. However, they found it necessary to eliminate 17 items to reach alpha values of .73, .66, and .45 for HIV/AIDS myths, attitudes, and facts, respectively. The quality of the items and their cultural appropriateness were not evaluated by Davis et al. (1999). Thus the effect of the adaptation on the interpretation of the items cannot be ruled out, as it was necessary to eliminate one-half of the items to reach an acceptable level of reliability in the new population. Skevington (2002) believes that an adaptation will always affect reliability and stated that when the psychometric properties of a source instrument are compared with those of a new translation from a rigorously tested instrument in a target culture, the accepted parameters of reliability and validity tend to be notably poorer for the target culture than for the source (p. 137). A successful adaptation allows items to be used in a new population for the first time, and as the items are revised and improved by correctly matching them to the population, the effect will be to increase the quality of the instruments psychometric properties. 150

PAGE 162

To summarize, the results of the test-retest study and the internal consistency determined that the reliability of the scores from the Knowledge section was moderate. The reliability of the scores from the five Attitudinal subscales varied. The scores from the two subscales developed by Basen-Engquist et al. (1999) had reliability. The reliabilities of the scores from the Peer-pressure and Abstinence subscales were low. The least reliable scores were from the Threat of HIV and Drug use. The reliability of the scores from the Condom subscale was not acceptable. Consequently, some of the items representing the Condom use, Drug use, and Threat of HIV will need to be reworded so they represent one domain and re-tested to determine if the interpretation is adequate. In addition, more items with good discrimination could be added to improve the reliability of the scores. Purpose Five: Determine, Using Confirmatory Factor Analysis, if the Spanish Version of the CDC Instrument Contains Five Attitudinal Dimensions (Peer-pressure, Condom use, Abstinence, Drug use, and Threat of HIV) The construct validity of the attitudinal items was examined with confirmatory factor analysis (CFA) to determine how well the data fit the five-factor model consisting of: Peer-pressure, Abstinence, Condom use, Drug use, and Threat of HIV. Once it was determined that the five-factor model had a poor fit, CFA was used in an exploratory manner to evaluate alternative models. Given the post hoc nature of the analysis, the results need to be viewed with caution. These alternative models will have to be cross-validated in future research. Five additional models were proposed and when examined with CFA, only one of the models yielded an acceptable fit. 151

PAGE 163

Confirmatory Factor Analysis of Model 6 Results from CFAs, the reliability study, and the expert panel suggested that the Condom use subscale was problematic and should be eliminated from the instrument. The Condom use subscale was removed. A new model with four factors (Peer-pressure, Abstinence, Drug use, and Threat of HIV) and nine correlated errors was re-evaluated with CFA. The results showed that acceptable fit was achieved with the addition of nine correlated error terms. These results provide partial support for the four different attitudinal dimensions. As can be seen in Figure 5, the number of items (9 items) with a weak or low standardized factor loading ( .4) was almost equal to the number (10 items) of loadings that were moderate (> .4). This indicated that the model was not very strong and could be improved. Large secondary loadings affect the models fit. The Lagrange test identified 10 items in Model 6 that had a secondary loading on another factor; two of these items had practical significance. Drug use item 13 (People who share drug needles shouldn't worry because they probably won't get infected with HIV) had a secondary loading in the Threat of HIV subscale. This item can be interpreted as having a slight association to the risk dimension. Similarly, Drug use item 17 (Anyone who shares needles is taking a chance of getting infected with HIV) was associated with the Threat of HIV subscale. The translation of the item included the word risk (riesgo), and its interpretation was more directly perceived as a risk or threat of HIV. 152

PAGE 164

Threat of HIV Drug use Abstinence Pee r p ressure 1 6 11 15 19 2 7 12 20 9 13 17 24 10 14 18 25 5 4 Figure 5. Standardized factor loadings of the four-dimensi onal Attitudinal model. .40 E1 E6 E11 E15 E19 E2 E7 E12 E20 E4 E9 E13 E17 E24 E10 E14 E18 E25 E5 .41 .38 .36 50 .56 .27 .67 .62 .36 .45 .35 .65 .26 .33 .47 .007 .46 -.09 153

PAGE 165

The secondary loadings suggested that the items may need to be revised so they corresr an first was a correlation of .94 between the Peer-pressure and Threat of HIV subscat. The en the Peer-pressure and Threa ts pond to one subscale. The model assumed that each item loaded on only one factobecause each subscale was conceptually different. However, because the overall construct was about HIV/AIDS, correlations among the subscales were expected. Instruments that assess attitudes are also expected to have a certain amount of correlation among the subscales. A problem occurs when items can be theoretically associated with more thone factor. For example, the inclusion of multi-dimensional items in the instrument was problematic because it affected the internal structure of the instrument and produced a high correlation among subscales. Two unusually high correlations were identified in Model 6. The les. This suggested that the subscales, although designed to be conceptually different, overlapped significantly. Reasons for the high correlation were not evidenwords in the items did not yield clues for explaining the strong relationship. The words in the items from the Threat of HIV subscale did not make a specific reference to friends or peers. Items 5, 14, 18, and 25 used the word teen-agers in their statements (see Appendix 1), but the words were not written to directly involve a perceived risk involving sexual intercourse with a friend or with their peers. An alternative explanation for the high correlation betwe t of HIV subscales was the students perception of risk in some of the Peer-pressure items. The students in the pretest said that Peer-pressure items 1 and 6 weremaking a vague reference to a risk because they understood the statement as a studenchoice between a friend or a harmful situation. As an example, Peer-pressure item 1 (If 154

PAGE 166

your friends want you to do something that you think might not be safe, you should at least try it), may be indicative of a situation involving a risk. The same risk concept was implied in items 6 and 15. It is likely that such an interpretation may have played a role in producing the high correlation. The second high correlation of .9 4 was found between the Drug use and Threat of HIV ssed h an Apparent Association to More Than One Subscale ubscales. A direct relationship was apparent because items from the Drug use subscale made reference to the risk of injecting drugs and HIV. Similar words were uin items 4, 13, and 17. In particular, the words in item 4, A teen-ager can inject drugs once in a while without a risk of getting infected with HIV, were directly related to a risk dimension. Based on the fact that the word risk was included in these items, a higcorrelation was expected. Table 51 Items with Item Original Subscale Projected Subscale 4 Drug use Drug use & Risk/Th reat of HIV 5 Threat of H IV V IV V IV Condom use f HIV Condom use & Risk/Threat of HI 9 Drug use Drug use & Risk/Threat of HIV 10 Threat of H Condom use & Risk/Threat of HI 13 Drug use Drug use & Risk/Threat of HIV 17 Drug use Drug use & Risk/Threat of HIV 18 Threat of H Abstinence, Risk/Threat of HIV & 21 Condom use Abstinence & Condom use 25 Threat of HIV Abstinence & Risk/Threat o 155

PAGE 167

An examination of the words in items 4, 5, 9, 10, 13, 17, 18, 21, and 25 by the primary investigator indicated that the items could be interpreted as pertaining to more than one subscale (see Table 51). As discussed previously, items in the Drug use subscale were strongly associated with the Threat of HIV subscale. Furthermore, items (5 and 10) from the Threat of HIV could be associated with a condom dimension, and items 18 and 25 could be associated with an abstinence dimension. Seven items were associated with more than one subscale, and many of these items implied a risk or threat of HIV. The large number of items associated with assessing risk suggested that the instrument might be strongly assessing the Threat of HIV dimension. The model assumes that measurement errors for pairs of items are uncorrelated. Correlations between measurement errors could exist between pairs of items on the same factor and for items in different factors. In this study, Model 6 allowed the errors to be correlated to other errors (see Table 52) thus improving the fit of the model. The identification of correlated errors was useful because it provided information that can be used to improve the instrument. Specifically, Bryne (1994) reported that covariation, which results from correlated measurement errors, can be introduced by non-random errors associated with the items format. The type of scale in the instrument can introduce error. Furthermore, Basen-Engquist et al. (1999) reported that correlated errors affected their model, and they identified the grammatical structure as a source of error. In this study, items 1, 6, and 15 (Peer-pressure) had similar characteristics, such as positively worded statements and a similar grammatical structure. As described in Table 52, the correlation between the errors from items 1 and 6 was the greatest, and the 156

PAGE 168

similarity of the concept may have contributed to the correlation. The correlation between errors in items 11 and 19 was large, and both items were negatively worded and had similar grammar. During the pretest, a student inquired why items 11 and 19 were the same. The students perceived the two items to be the same. From this observation, it appeared the content and format were viewed to be the same by the students and substantiated Bryne (1994) and Basen-Engquist et al.s (1999) observations about the effect of correlated errors. Table 52 Paths in Model 6 Representing Correlated Error Error(Subscale) Error(Subscale) Expected Change in 2 E6 (Peer-pressure) E1 (Peer-pressure) 38 E19 (Peer-pressure) E11 (Peer-pressure) 32 E10 (Threat of HIV) E7 (Abstinence) 28 E7 (Abstinence) E6 (Peer-pressure) 25 E10 (Threat of HIV) E13 (Drug use) 21 E18 (Threat of HIV) E17 (Drug use) 16 E13 (Drug use) E7 (Abstinence) 15 E11 (Peer-pressure) E6 (Peer-pressure) 12 E5 (Threat of HIV) E4 (Drug use) 12 Correlated error is common when items are conceptually similar and persons provide a general response that reflects their overall attitude. Items in this study were assigned to one subscale, but they could be interpreted as belonging to another subscale. As previously described in Table 51, the possibility existed that it was difficult for students to discern the conceptual differences among some of the items, and this situation contributed to the observed correlated errors. Table 52 shows the correlation between errors that occurred for items from two different factors (E7 and E10, E10 and E13, E7 157

PAGE 169

and E13). The conceptual similarity reflected by multi-dimensional items in the Threat of HIV, Drug use, and Abstinence subscales may have been responsible for these correlated errors. Moreover, correlated measurement errors occur when similar items are placed next to each other and students answer in a set response. There were three sets of correlated errors from items that were next to each other (E4 and E5, E6 and E7, E17 and E18). The correlation between the error in item 4 (Drug use) and item 5 (Threat of HIV) may have been due to a similarity in both the words and concept. The correlation between the error in item 6 and item 7 was difficult to interpret and could not be explained. Different words and a different grammatical structure were observed in items 17 and 18; however, both items conveyed a risk for HIV, and this may explain the observed correlated error. To minimize the effect of errors, items that are grammatically and conceptually similar must be evenly dispersed in the instrument and not placed next to each other. In summary, Model 6, which tested the internal structure of a four-factor model (Abstinence, Drug use, Threat of HIV, and Peer-pressure) and included paths involving correlated errors, yielded an acceptable statistical fit. Additional research is needed to evaluate the robustness of this model. Purpose Six: Evaluate the Concurrent Validity of the Abstinence and Condom use Subscales from the Spanish Version of the CDC Instrument Pearson product-moment correlation coefficients were calculated to assess the relationship between the CDCs translated Condom and Abstinence subscales with their respective subscales developed by Basen-Engquist et al. (1999). The concurrent validity analysis showed a strong correlation (.67) between the two different translated 158

PAGE 170

Abstinence subscales. The strong correlation among the two subscales indicated the scales measured the Abstinence construct. The data from the reliability study and CFA provided additional evidence about the validity of the items to assess the abstinence subscale. The translated CDCs Abstinence subscale was therefore acceptable for measuring the dimension. A much lower correlation was observed between the two condom subscales. This lower correlation was not surprising because the coefficient alpha for the translated CDCs Condom subscale was not acceptable. The information from all phases of the study suggested that the items for the translated CDC Condom subscale need to be revised and retested. Significance and Implications of the Results Significance. This study provided an example of the methodology to translate, cross-culturally adapt, and validate the scores of an instrument for use in El Salvador. Guillemin et al.s (1993) guidelines for a cross-cultural adaptation were successfully implemented to provide program planners with a Spanish instrument for use with students in public high schools in El Salvador. In particular, an HIV/AIDS knowledge component can be assessed with a high degree of confidence and an attitudinal component can be assessed with a lower degree of confidence. The methodology for the cross-cultural adaptation was straightforward and advantageous because fewer resources, personnel, and time were needed compared to if a new instrument had been developed in El Salvador. Psychometric properties were also calculated to establish the baseline characteristics of the adapted instrument in the new population. The scores from the 159

PAGE 171

instrument were used to estimate reliability and establish construct validity. The psychometric properties of new instruments provide information that clarifies the interpretation of the scores from the target population and identifies components of the instrument that need to be improved. The combination of Guillemin et al.s (1993) guidelines and the IQOLA (Bullinger et al., 1998; Gandek & Ware, 1998) groups suggestions was advantageous in that it provided a measure of the instruments cultural adaptation. These procedures also provided the information about the reliability, content validity, and construct validity of the scores of the instrument. Two of the studys purposes were vital to the cross-cultural adaptation. The first was to construct an instrument with an acceptable level of content validity. The second was to achieve a high level of cultural acceptability of the items for use in high school. Both purposes were achieved with the assistance of three different Salvadorian panels and two pretests with high school students. They identified the content that had to be adjusted to meet the cultural views of Salvadorian adolescents. Additionally, their review lessened the concern about issues of human sexuality that could be deemed as offensive or not appropriate for use in high school. More importantly, assuring the content and cultural acceptability was essential in obtaining conceptual equivalence in the adapted version. When the HIV/AIDS content in the adapted version is equivalent, it will reduce the bias that occurs when instruments are interpreted differently by another cultural group (Hui & Triandis, 1985). Items related to peer-pressure, drug use, and dating practices were interpreted in different ways among Salvadorian adolescents. Peer-pressure items 1, 6, and 11 made a general reference to peer-pressure, and the students in the pretests had a different view of 160

PAGE 172

these items. They wanted a situation that focused on making a choice between a healthy behavior over the demands imposed by their friends. This perception raised the issue that peer-pressure consisted of two dimensions: an influence and a resistance. A new problem is introduced if the items are written to reflect a situation in which a friend influences a student into a harmful situation. The inclusion of a harmful situation involving an HIV risk, drugs, or abstinence increases the chance that the attitudinal concept will be perceived as multi-dimensional because two new factors were introduced, one associated with the peer/friend and the other to a potentially harmful situation. This becomes problematic in that a multi-dimensional item has the potential to confound the results. Conversely, the resistance dimension was typically represented by the following statement: Teen-agers should be more willing to resist pressures from their friends. These statements are straightforward and easy to interpret. Students may find statements about resistance to be repetitive. As demonstrated in this study, when items were similar to each another, it introduced correlated measurement errors. The addition of an equal number of positively and negatively worded "resistance" based peer-pressure items should be investigated in a future study. The results of this study provide a starting point of how peer-pressure is perceived among Salvadorian adolescents. Another observed cultural difference was the irrelevancy of item 24 (Drug use). The item made reference to cleaning drug needles with bleach to prevent the transmission of HIV. Students in the pretest and a panelist observed that it was unimportant to know how to clean drug needles. The prevalence of injecting drugs is not known for this population although a person on the Lay Panel said it was low. However, if injecting 161

PAGE 173

drugs is known to be high in a population, then it will substantiate the inclusion of item 24. At this point, the expert panel reported that the use of needles for tattoos was the wiser topic to assess because needles are a source of infection and tattoos were becoming fashionable. Based on this information, item 24 should be excluded from the instrument. The final difference centered on the sexual practices of adolescents who were dating. Item 21 assumed that teen-agers in the United States have sexual intercourse when they go out on a date. The item was relevant to teen-agers in the United States and was supported by data from the Youth Risk Behavior Survey, which estimated that 50% of high school students in the United States had sexual relations (Kann et al., 2000). The same data for Salvadorian high school students were not available. A recent survey in El Salvador estimated that 51% of women (15 to 25 years of age) were sexually active and that the percentage was higher for males in the same age (Associacion Demographica Salvadorena, 2003). From this information, it is apparent that young adults in El Salvador are sexually active, which emphasizes the need to investigate the cultural context of sexual relations. The Salvadorian students reported that sexual relations are not casual among them and rarely occur when two students go out on a date. They added that sexual relations occur when a couple has been in a long-term relationship. Items to assess abstinence and condom use must reflect the cultural characteristics of the dating practices of Salvadorian adolescents. The knowledge gained from exploring and assessing the context in which sexual relations occur in Salvadorian adolescents will be beneficial to HIV prevention. In particular, prevention programs can specifically target adolescents in a relationship. Attitudes related to abstinence or condom use can be strengthened by focusing on the 162

PAGE 174

characteristics and expectations related to relationships among adolescents. Also, the relevancy of other issues common in relationships, such as power, communication, and pregnancy prevention, should also be taught to couples. To summarize, the studys methodology for a cross-cultural adaptation provided an instrument in Spanish for use in El Salvador. In particular, the Knowledge component can be used with a high degree of confidence but the Attitudinal component can be used with less confidence. Validation techniques, such as factor analysis, that evaluate the internal structure of an instrument, are valuable to a cross-cultural adaptation. Factor analysis allows the factorial structure of cross-culturally adapted instruments to be evaluated. If the instruments yield the same structure, then it can be inferred that the same construct is being assessed, thus supporting the cross-cultural adaptation (Gandek & Ware, 1998). Achieving cross-cultural equivalence is the optimal goal in an adaptation. Ware et al. (1988) demonstrated the robustness of the IQOLAs methods to achieve equivalence by providing evidence--with the use of CFA--that their adapted instruments retained a two-factor structure across 10 different countries. Likewise, in another study, Daza et al. (2002) replicated the factorial structure of an instrument in Spain. The adaptation was judged to be successful because the internal structure of the original instrument was the same as the newly adapted instrument. An important point derived from the results of Daza et al. (2002) and the IQOLA group was that a previously validated instrument allows the factor structure of the adapted version to be compared. These two studies provide lessons for future adaptations and highlight the shortcoming of using items from an item bank. In this study, the CDCs items had a high degree of face validity, but other psychometric data were not available. 163

PAGE 175

The construct validity of the original CDC instrument was not determined by statistical analysis. In contrast, the items from Basen-Engquist et al. (1999) were used in several studies (Basen-Engquist et al., 2001; Coyle et al., 2001), and the scores were validated by an expert panel and through statistical procedures. These items were included in this study, treated in parallel to the CDC items and used in the concurrent validity study. The reliabilities of the scores for the adapted Basen-Engquist et al. (1999) Abstinence (alpha =.68) and Condom use (alpha =.69) subscales were very high. These results suggest that there are positive benefits of using a previously validated instrument on the outcome of the adaptation. Determining the factors responsible for the failure to replicate the original five-factor structure was difficult. A cross-cultural adaptation is a complex procedure. As discussed previously, extreme cultural differences and the use of a previously validated instrument can affect the results. Multi-dimensionality of the items is another factor that can affect the internal structure of an instrument. Ruiz et al. (2002) adapted the English version of the Weight Efficacy Life-Style Questionnaire for use in Spain. They reported that the new instrument did not retain the original five-factor model, and they stated that the original factor structure was not replicated due to items which were culturally irrelevant and ambiguous. In this study, one of the dimensions was not replicated, and multi-dimensionality affected an item's interpretation and subsequently affected the factor structure. The nature of the HIV/AIDS domain presented a challenge to writing quality items reflective of only one domain. Several items from the instrument were written to assess more than one domain. For example, the risk or threat of HIV could be interpreted 164

PAGE 176

as belonging to either the condom use or abstinence domain. The use of words that were associated with the Drug use, Condom use, and Abstinence subscales were used for items that assessed a Threat of HIV. Merging words from different constructs confuses the intended meaning and results in more than one interpretation. As an example, item 10 (Threat of HIV) stated: It's okay to have sex without a condom because your chance of getting infected with HIV is very low. The challenge for writing items similar to item 10 is to write statements without including words associated with the abstinence and condom use dimensions. The words in the Condom use items were also found to be problematic due to their interpretation and the fact that sexual experience was required to form an attitude. Condom use items have to be carefully written to assess the act of using a condom. It was important to distinguish the use of condoms from the act of having intercourse because it confounds the abstinence dimension. Another issue was the misunderstanding of the items by sexually inexperienced adolescents. During the administration of the instrument, four females discreetly said they did not understand the meaning of the item (Using a condom doesn't make sex less pleasurable). The formation of attitudes related to condom use may be dependent on previous sexual experience. Therefore, attitudes about condom use should be assessed in a population of older persons that may have a higher proportion of sexually experienced persons. Many HIV/AIDS prevention programs attempt to reduce or eliminate risk behaviors associated with sexual intercourse, delaying sexual activity, and condom use (Kirby et al., 1994; Kirby, 2001). Kirby (2001) reported that behavioral theories were the difference in every successful program because they systematically targeted a variety of 165

PAGE 177

psychosocial factors and provided information about HIV transmission. Knowledge and attitudes were selected because of their role as mediating variables in behavioral theories and because there is a need to evaluate HIV/AIDS prevention programs. Accordingly, curriculum planners in El Salvador will benefit by including psychosocial factors in their programs to reduce risky sexual behavior. The objectives should be directed at changing or modifying psychosocial factors, such as attitudes toward abstinence, attitudes about risk of HIV infection, sexual norms, attitudes toward peer-pressure, refusal skills, and attitudes about contraception or condom use. These programs must be evaluated with instruments that provide valid scores that can be used to determine the effectiveness of the interventions. Implications. The purpose of this study was to provide an instrument in Spanish to assess the HIV/AIDS knowledge and attitudes in the population of students in public high schools in El Salvador. The results from the reliability and validity of the scores of the Knowledge component suggested that this part of the instrument could be used with a high degree of confidence. Particularly, Knowledge items 2 to 15 were appropriate for use in El Salvador. In contrast, item 1 could be improved by restating it as: If you have sex without a condom and the other person has HIV, then you can get AIDS. After this minor modification, the instrument could be used by program evaluators to assess an HIV knowledge component. The Attitudinal section was theorized to assess five dimensions (Abstinence, Peer-pressure, Condom use, Threat of HIV, and Drug use); however, partial support for the validity of the scores was provided for only four of the dimensions. The validity of the scores for four-factor model was determined to be moderate when correlated errors 166

PAGE 178

were added to the model. Additionally, the following translated attitudinal items: 2, 4, 5, 9, 10, 11, 12, 13, 17, 19, 20, and 21 were judged to be acceptable for use in El Salvador. In contrast, the 11 Attitudinal items described in Table 53 should be reworded to improve and narrow their meaning so that they reflect one dimension; these items will then have to be retested in El Salvador. Table 53 Suggested Changes to the Attitudinal Items Subscale/Item Original Statement New Version Comment Peer-pressure 1 If your friends want you to do something that you think might not be safe, you should at least try it. Investigate how peer-pressure is perceived in El Salvador. 6 To keep your friends, you should go along with most things your friends want you to do. Investigate how peer-pressure is perceived in El Salvador and alter the grammatical structure to reduce correlated error. 11 Teen-agers should learn how to resist pressures from their friends. Leave as is and change 19 or vice versa. Change the grammatical structure of 11 and 19. 19 Teen-agers should be more willing to resist pressures from their friends. Similar to item 11. Abstinence 7 People who don't have sex before they get married are strange. People should wait to have sexual relations until marriage. Condom use 8 It is not smart to have sex without using a condom. A condom should be used every time a person has sexual relations. 16 Using a condom doesn't make sex less pleasurable. This question is appropriate with sexually experienced persons. 23 People who use condoms during sex don't trust the person they're with. 167

PAGE 179

Table 53 (continued) Threat of HIV 5 Teen-agers are at risk of getting infected with HIV if they engage in sex without a condom. Investigate how the threat of HIV is perceived in El Salvador. 10 Its okay to have sex without a condom because your chance of getting infected with HIV is very low. 14 Teen-agers should realize that if they're not careful, they could get infected with HIV. 18. If teen-agers are careful about choosing sexual partners, they won't get infected with HIV. 25. HIV is something that teen-agers should think about when they date. Investigate how the threat of HIV is perceived in El Salvador. Note : *An alternative version has to be developed and then pilot tested in El Salvador. The items from the Peer-pressure subscale were marginally associated with the dimension and should be reworded to strengthen this association. The recommendation from the pretest was to add a unique situation where students made a choice between a situation and their friends. However, the context of the situation has to be investigated in El Salvador, and, as discussed previously, the items must be carefully written to retain a peer-pressure dimension. Overall, the items that assessed the abstinence dimension were found to be adequate. The exception was item 7 which should be reworded because the word strange (extraos) was problematic. As presented in Table 53, the item can be improved by eliminating the word extraos and rewording the item as: People should wait to have sexual relations until marriage. If an adequate translation is not achieved, then the item should be eliminated. 168

PAGE 180

The Condom use subscale was not acceptable for use and items within this scale will have to be rewritten. Attitudes toward condom use were difficult to write because the global nature of the concept includes a dimension related to abstinence, relationships, and a partners pressure to use or not to use a condom. These items can be reworded either independently of sexual relations or written to reflect an attitude toward condom use immediately before sexual relations occur. For example, item 8 can be written specifically to assess condom use attitudes among adolescents by changing it to: A condom should be used if an adolescent has sexual relations. The statement attempts to assess a global attitude toward condom use. Additionally, the three condom items from Basen-Engquist et al. (1999) were worded differently than the ones from the CDC. The data from the study indicated Basen-Engquist et al.s items had an acceptable measure of reliability and were appropriate for use in El Salvador. New items can be modeled after these items; for example, item 28 (Basen-Engquist et al., 1999) stated, I believe condoms should always be used if a person my age has sex. The item directly assessed a global attitude toward condom use; however, a limitation of the three items was the use of the words I believe. The term will have to be used sparingly because a similar structure can introduce correlated measurement errors. To summarize, there was partial support that the instrument contained four Attitudinal dimensions (Abstinence, Peer-pressure, Threat of HIV, and Drug use). In particular, the construct validity of the Abstinence dimension was the strongest followed by Peer-pressure. More research is needed to provide additional validity and reliability evidence before this instrument can be recommended with a high degree of confidence to assess programs in El Salvador. 169

PAGE 181

Limitations and Directions for Future Research As an initial study, this research has contributed to the methodology of cross-cultural instrument adaptation and validation. Despite this contribution, additional research is needed that builds on the present design. The design of the study can be improved by increasing the size and diversity of the sample. Confirmatory factor analysis benefits from large samples. A larger sample can be obtained by including more schools. An advantage of a larger sample when using CFA is that it provides the flexibility to cross-validate the results. Cross-validation with two groups of 500 will allow an independent CFA with each sample. The factor structure of both groups can then be compared to verify the validity of the scores from the instrument. Additionally, the study can be extended by collecting data at more than one point. This would allow a comparison of the data during different times of the year. A limitation of the study was the small number of students (n = 39) who participated in the test-retest study. A larger sample size should be used in a test-retest. Additionally, the generalizability of the study was limited to Salvadorian students in public high schools. A future study should include private schools. Then the scores from the two groups could be compared to determine the differences among the two groups. The results of the study cannot be generalized to students who dropped out of school or who did not have the financial resources to stay in school. A future study should attempt to include members from these groups. The lack of published data related to the construct validity and the reliability of the scores of the CDCs items limited comparisons to the adapted version. Furthermore, no published studies have examined the construct validity of HIV/AIDS instruments for 170

PAGE 182

use in a Spanish-speaking population. This limitation precluded comparisons to the adapted version. Given the importance of measuring knowledge and attitudes related to HIV/AIDS in countries like El Salvador, future research will need to continue to collect validity evidence evaluating the success of translating and cross-culturally adapting instruments such as those developed by the CDC. 171

PAGE 183

REFERENCES Aday, L. A. (1996). Designing and conducting health surveys: A comprehensive guide (2nd ed.). San Francisco, CA: Jossey-Bass Publishers. Ajzen, I., & Fishbein, M. (1980). Understanding attitudes and predicting social behavior. Englewood Cliffs, NJ.: Prentice-Hall. Alan Guttmacher Institute. (1999). Facts in brief: Teen sex and pregnancy. Retrieved June, 2003, from http:/www.agi-usa.org/pubs/ib_teen_sex.html. Albarracin, D., Johnson, B. T., Fishbein, M., & Muellerleile, P. A. (2001). Theories of reasoned action and planned behavior as models of condom use: A meta-analysis. Psychological Bulletin, 127(1), 142-161. Alonso, J., Prieto, L., Ferrer, M., Vilagut, G., Broquetas, J. M., Roca, J., Batlle, J. S., & Anto, J. M. (1998). Testing the measurement mroperties of the Spanish version of the SF-36 Health Survey among male patients with chronic obstructive Pulmonary Disease. Journal of Clinical Epidemiology, 51(11), 1087-1094. Asociacion Demographica Salvadorea. (2003). Encuesta nacional de salud familiar. FESAL 2002/2003. Informe resumido. National survey of family health. Summarized version. San Salvador, El Salvador. Bandura, A. (1986). Social foundations of thought and action: a social cognitive theory. Englewood Cliffs, NJ.: Prentice-Hall. Basen-Engquist, K., Coyle, K. K., Parcel, G. S., Kirby, D., Banspach, S. W., Carvajal, S. C., & Baumler, E. (2001). Schoolwide effects of a multicomponent HIV, STD, and pregnancy prevention program for high school students. Health Education and Behavior, 28(2), 166-185. Basen-Engquist, K., Masse, L. C., Coyle, K., Kirby, D., Parcel, G. S., Banspach, S., & Nodora, J. (1999). Validity of scales measuring the psycho-social determinants of HIV/STD-related risk behavior in adolescents. Health Education Research, 14(1), 25-38. Basen-Engquist, K., & Parcel, G. S. (1992). Attitudes, norms, and self-efficacy: A model of adolescents' HIV-related sexual risk behavior. Health Education Quarterly, 19(2), 263-277. 172

PAGE 184

Basen-Engquist, K., Parcel, G. S., Harrist, R., Kirby, D., Coyle, K., Banspach, S., & Rugg, D. (1997). The safer choices project: Methodological issues in schoolbased health promotion intervention research. Journal of School Health, 67(9), 365-371. Beaton, D., Bombardier, C., Guillemin, F., & Ferraz, M. B. (2002). Recommendations for the cross-cultural adaptation of health status measures. Retrieved June, 2003 from the American Academy of Orthopedic Surgeons Website. Becher, J. C., Garcia, J. G., Kaplan, D. W., Gil, A. R., Li, J., Main, D., Herrera, J. A., Arias, L., & Bromet, A. (1999). Reproductive health risk behavior survey of Colombian high school students. Journal of Adolescent Health, 24(3), 220-225. Breckler, J. S., & Wiggins, E. S. (1989). On defining attitude and attitude theory: Once more with feeling. In A. R. Pratkanis, S. J. Breckler, & A. G. Greenwald, A. G. (Eds.), Attitude structure and function (pp. 407-427). Hillsdale, NJ: Lawrence. Erlbaum Associates. Brislin, R. W., Lonner, W. J., & Thorndike, R. M. (1973). Cross-cultural research methods. New York: J. Wiley. Bullinger, M., Alonso, J., Apolone, G., Leplege, A., Sullivan, M., Wood-Dauphinee, S., Gandek, B., Wagner, A., Aaronson, N., Bech, P., Fukuhara, S., Kaasa, S., & Ware, J. E., Jr. (1998). Translating health status questionnaires and evaluating their quality: The IQOLA Project approach. International Quality of Life Assessment. Journal of Clinical Epidemiology, 51(11), 913-923. Byrne, B. M. (1994). Structural equation modeling with EQS and EQS/Windows: Basic concepts, applications, and programming. Thousand Oaks, CA: Sage Publications. Carvajal, S., Parcel, G., Banspach, S., Chan, W., Coyle, K., Kirby, D., & Basen-Engquist, K. (1999). Psycho-social predictors of delay of first sexual intercourse by adolescents. Health Psychology, 18(5), 433-452. Centers for Disease Control and Prevention (CDC). (1988). Guidelines for effective school health education to prevent the spread of AIDS. Morbidity and Mortality Weekly Report, 37(S-2), 1-14. Centers for Disease Control and Prevention. (1995). HIV health education and risk reduction guidelines. Retrieved April, 2004, from http://www.cdc.gov/hiv/HERRG/HIV_HERRG.htm Centers for Disease Control and Prevention. (1998). Youth 97. [CD-ROM]. Springfield, VA: Centers for Disease Control and Prevention. 173

PAGE 185

Centers for Disease Control and Prevention. (1999). HIV/AIDS Prevention research synthesis project. Compendium of HIV prevention interventions with evidence of effectiveness interventions. Retrieved January, 2003, from http://www.cdc.gov/hiv/pubs/HIVcompendium/hivcompendium.pdf Centers for Disease Control and Prevention. (2002a). The global HIV and AIDS epidemic, 2001. Morbidity and Mortality Weekly Report, 50(21), 434-9. Centers for Disease Control and Prevention. (2002b). The handbook for evaluating HIV education-Booklet 6. Retrieved January, 2002, from http://www.cdc.gov/nccdphp/dash/publications/hiv_handbook/index.htm Centers for Disease Control and Prevention. (2003a). At a glance the HIV epidemic. Retrieved February, 2003, from http://www.cdc.gov/nchstp/od/news/At-aGlance.pdf Centers for Disease Control and Prevention. (2003b). Technical update for school health guidelines. Retrieved May, 2003, from http://www.cdc.gov/nccdphp/dash/sexualbehaviors/guidelines/technicalupdate Centers for Disease Control and Prevention. (2003c). Young people at risk: HIV/AIDS among Americas youth. Retrieved June, 2003, from http://www.cdc.gov/hiv/pubs/facts/youth.htm Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, & Adolescent and School Health. (2004). School health policies and program study (SHPPS). Retrieved April, 2004, from http://www.cdc.gov/nccdphp/dash/shpps/ Central Intelligence Agency. (2004). World Fact Book 2003. Retrieved April, 2004, from http://www.cia.gov/cia/publications/factbook/index.html Choi, K. H., & Coates, T. J. (1994). Prevention of HIV infection. AIDS, 8(10), 13711389. Cleland, J. G., Ferry, B., & World Health Organization. (1995). Sexual behaviour and AIDS in the developing world. London; Bristol, PA: Taylor & Francis. Contreras, A., Garcia-Alonso, R., Echenique, M., & Daye-Contreras, F. (1999). The SOL formulas for converting SMOG Readability scores between health education materials written in Spanish, English, and French. Journal of Health Communication, 4, 21-29. 174

PAGE 186

Coyle, K., Basen-Engquist, K., Kirby, D., Parcel, G., Banspach, S., Harrist, R., Baumler, E., & Weil, M. (1999). Short-term impact of safer choices: A multicomponent, school-based HIV, other STD, and pregnancy prevention program. Journal of School Health, 69(5), 181-188. Coyle, K., Basen-Engquist, K., Kirby, D., Parcel, G., Banspach, S., Collins, J., Baumler, E., Carvajal, S., & Harrist, R. (2001). Safer choices: Reducing teen pregnancy, HIV, and STDs. Public Health Reports, 116( S-1), 82-93. Crocker, L., & Algina, J. (1986). Introduction to classical and modern test theory. New York: Holt, Rinehart, and Winston. Davis, C., Tang, C. S.-K., Chan, S.-F. F., & Noel, B. (1999). The development and validation of the international AIDS Questionnaire-Chinese Version (IAQ-C). Educational and Psychological Measurement, 59(3), 481-491. Daza, P., Novy, D. M., Stanley, M. A., & Averill, P. (2002). The depression anxiety stress scale-21: Spanish translation and validation with a Hispanic sample. Journal of Psychopathology and Behavioral Assessment, 24(3), 195-205. Departamento de Educacion de El Salvador. (2002). Estado actual de la educacion. 2002. Retrieved April, 2004, from http://www.mined.gob.sv/ DiClemente, R. J., Zorn, J., & Temoshok, L. (1986). Adolescents and AIDS: a survey of knowledge, attitudes and beliefs about AIDS in San Francisco. American Journal of Public Health, 76(12), 1443-1445. Figueroa Erazo, D. M. (2001a). Psicologia del la adolescencia. Dessarrollo de programa oficial de Primer ao de Bachillerato.San Salvador, El Salvador: Talleres Graphicos, UCA. Figueroa Erazo, D. M. (2001b). Psicologia de la adolescencia. Desarrollo del programa oficial de Segundo ao de Bachillerato. San Salvador, El Salvador: Talleres Graphicos, UCA. Fisher, J. D., & Fisher, W. A. (2000). Theoretical approaches to individual-level change in HIV risk behavior. In J. L., Peterson, & R. J., DiClemente (Eds.), Handbook of HIV prevention (pp. 3-48). New York: Kluwer Academic/Plenum Publishers. Fundacion Fomento En Salud. (2001). Reporte final. Enquesta conductual de riesgos relacionada con las IT/VIH/SIDA. National survey of risks associated with HIV/AIDS. Tegucigalpa MDC: Fundacion Fomento En Salud Centro National de Concientizacion y Prevencion Del SIDA, Mercaplan. Gall, M. D., Borg, W. R., & Gall, J. P. (1996). Educational research: An introduction (6th ed.). White Plains, NY: Longman. 175

PAGE 187

Gandek, B., Ware, J. E., Jr., Aaronson, N. K., Alonso, J., Apolone, G., Bjorner, J., Brazier, J., Bullinger, M., Fukuhara, S., Kaasa, S., Leplege, A., & Sullivan, M. (1998). Tests of data quality, scaling assumptions, and reliability of the SF-36 in eleven countries: Results from the IQOLA Project. International Quality of Life Assessment. Journal of Clinical Epidemiology, 51(11), 1149-1158. Gandek, B., & Ware, J. E. (1998). Methods for validating and norming translations of health status questionnaires: The IQOLA Project approach. Journal of Clinical Epidemiology, 51(11), 953-959. Green, L. W., & Kreuter, M. W. (1999). Health promotion planning: An educational and ecological approach (3rd ed.). Mountain View, CA: Mayfield. Greenwald, A. G. (1989). Why are attitudes important? In A. R. Pratkanis, S. J. Breckler, & A. G. Greenwald, A. G. (Eds.), Attitude structure and function (pp.1-10). Hillsdale, NJ: Lawrence. Erlbaum Associates. Guillemin, F., Bombardier, C., & Beaton, D. (1993). Cross-cultural adaptation of healthrelated quality of life measures: Literature review and proposed guidelines. Journal of Clinical Epidemiology, 46(12), 1417-1432. Hatcher, L. (1994). A step-by-step approach to using SAS for factor analysis and structural equation modeling. Cary, NC: SAS Institute Inc. Henderson, A. C., Champlin, S., & Evashwick, W. (1998). Promoting teen health: Linking schools, health organizations, and community. Thousand Oaks, CA: Sage Publications. Herdman, M., Fox-Rushby, J., & Badia, X. (1998). A model of equivalence in the cultural adaptation of HRQoL instruments: The universalist approach. Quality of Life Research, 7(4), 323-335. Hui, H., & Triandis, H. (1985). Measurement in cross-cultural psychology. A review and comparison of strategies. Journal of Cross-Cultural Psychology, 16(2), 131-152. Jemmott, J. B., III, Jemmott, L. S., & Fong, G. T. (1992). Reductions in HIV riskassociated sexual behaviors among black male adolescents: Effects of an AIDS prevention intervention. American Journal of Public Health, 82(3), 372-377. Jemmott, J. B., III, Jemmott, L. S., & Fong, G. T. (1998). Abstinence and safer sex HIV risk-reduction interventions for African American adolescents: A randomized controlled trial. Jouranl of American Medical Association, 279(19), 1529-1536. 176

PAGE 188

Jemmott, J. B., & Jemmott, L. S. (2000). HIV behavioral interventions for adolescents in community settings. In J. L., Peterson, & R. J., DiClemente (Eds.), Handbook of HIV prevention (pp. 103-124). New York: Kluwer Academic/Plenum Publishers. Johnson, B. T., Carey, M. P., Marsh, K. L., Levin, K. D., & Scott-Sheldon, L. A. J. (2003). Interventions to reduce sexual risk for the human immunodeficiency virus in adolescents, 1985-2000: A research synthesis. Archives of Pediatrics and Adolescent Medicine, 157(4), 381-388. Kann, L., Kinchen, S. A., Williams, B. I., Ross, J. G., Lowry, R., Grunbaum, J. A., & Kolbe, L. J. (2000). Youth risk behavior surveillance--United States, 1999. Morbidity and Mortality Weekly Report, CDC Surveillance Summaries, 49(5), 132. Keller, S. D., Ware, J. E., Jr., Bentler, P. M., Aaronson, N. K., Alonso, J., Apolone, G., Bjorner, J. B., Brazier, J., Bullinger, M., Kaasa, S., Leplege, A., Sullivan, M., & Gandek, B. (1998). Use of structural equation modeling to test the construct validity of the SF-36 Health Survey in ten countries: Results from the IQOLA Project. International Quality of Life Assessment. Journal of Clinical Epidemiology, 51(11), 1179-1188. Kim, N., Stanton, B., Li, X., Dickersin, K., & Galbraith, J. (1997). Effectiveness of the 40 adolescent AIDS-risk reduction interventions: A quantitative review. Journal of Adolescent Health, 20(3), 204-215. Kirby, D. (1992). School-based programs to reduce sexual risk-taking behaviors. Journal of School Health, 62(7), 280-287. Kirby, D. (1999). Reflections on two decades of research on teen sexual behavior and pregnancy. Journal of School Health, 69(3), 89-94. Kirby, D. (2001). Emerging answers. Research findings on programs to reduce teen pregnancy. Washington, DC: National Campaign to Prevent Teen Pregnancy. Kirby, D. (2002). The impact of schools and school programs upon adolescent sexual behavior. Journal of Sex Research, 39(1), 27-33. Kirby, D., Barth, R. P., & Leland, N. (1991). Reducing the risk: Impact of a new curriculum on sexual risk-taking. Family Planning Perspectives, 23(6), 253-263. Kirby, D., Korpi, M., Adivi, C., & Weissman, J. (1997). An impact evaluation of project SNAPP: An AIDS and pregnancy prevention middle school program. AIDS Education and Prevention, 9(S-1), 44-61. 177

PAGE 189

Kirby, D., Short, L., Collins, J., Rugg, D., Kolbe, L., Howard, M., Miller, B., Sonenstein, F., & Zabin, L. S. (1994). School-based programs to reduce sexual risk behaviors: A review of effectiveness. Public Health Reports, 109(3), 339-360. Leplege, A., Ecosse, E., Verdier, A., & Perneger, T. V. (1998). The French SF-36 Health Survey: Translation, cultural adaptation and preliminary psychometric evaluation. Journal of Clinical Epidemiology, 51(11), 1013-1023. Main, D. S., Iverson, D. C., McGloin, J., Banspach, S. W., Collins, J. L., Rugg, D. L., & Kolbe, L. J. (1994). Preventing HIV infection among adolescents: Evaluation of a school-based education program. Preventive Medicine, 23(4), 409-417. Microsoft Corporation. (2003). Office XP Standard. Word XP [Computer software]. Miller, K. S., & Whitaker, D. J. (2001). Predictors of mother-adolescent discussions about condoms: Implications for providers who serve youth. Pediatrics, 108(2), E28. Ministerio de Salud Publica y Asistencia Social, & Programa Nacional de ITS/VIH/SIDA. (2003). Situacion epidemiologica de las ITS/VIH/SIDA en El Salvador. El Salvador. [Brochure]. San Salvador, El Salvador. Mueller, D. J. (1986). Measuring social attitudes: A handbook for researchers and practitioners. New York: Teachers College Press. ODonnell, L., Myint-U, A., ODonnell, C. R., & Stueve, A. (2003). Long-term influence of sexual norms and attitudes on timing of sexual initiation among urban minority youth. Journal of School Health, 73(2), 68-75. Power, M., Harper, A., & Bullinger, M. (1999). The World Health Organization WHOQOL-100: Tests of the universality of Quality of Life in 15 different cultural groups worldwide. Health Psychology, 18(5), 495-505. Quackenbush, M., Clark, K., & Nelson, M. (1995). The HIV challenge: Prevention education for young people (2nd ed.). Santa Cruz, CA: ETR Associates. Rodriguez-Kelly, M. (2001). Las mujeres y el VIH/SIDA: perspectivas y realidades. Washington, DC: Genmark. Ruiz, V. M., Berrocal, C., Lopez, A., & Rivas, T. (2002). Factor analysis of the Spanish version of the weight efficacy life-style questionnaire. Educational and Psychological Measurement, 62(3), 539-555. SAS Institute Inc. (2001). SAS version 8.2 [Computer program]. Carey, NC. SAS Institute Inc. (2004). SAS version 8.2 users guide. Carey, NC. 178

PAGE 190

Schmidt, S., & Bullinger, M. (2003). Current issues in cross-cultural quality of life instrument development. Arch Phys Med Rehabil (Archives of physical medicine and rehabilitation.), 84(4)(Suppl 2: S29-34). Siegel, D., DiClemente, R., Durbin, M., Krasnovsky, F., & Saliba, P. (1995). Change in junior high school students' AIDS-related knowledge, misconceptions, attitudes, and HIV-preventive behaviors: Effects of a school-based intervention. AIDS Education and Prevention, 7(6), 534-543. Silva, M., & Ross, I. (2003). Evaluation of a school-based sex education program for low-income male high school students in Chile. Evaluation and Program Planning, 26, 1-9. Skevington, S. (2002). Advancing cross-cultural research on quality of life: Observations drawn from the WHOQOL development. Quality of Life Research, 11, 135-144. Song, E. Y., Pruitt, B. E., McNamara, J., & Colwell, B. (2000). A meta-analysis examining effects of school sexuality education programs on adolescents' sexual knowledge, 1960-1997. Journal of School Health, 70(10), 413-416. St. Lawrence, J. S., Brasfield, T. L., Jefferson, K. W., Alleyne, E., O'Bannon, R. E., & Shirley, A. (1995). Cognitive-behavioral intervention to reduce African American adolescents' risk for HIV infection. Journal of Consulting and Clinical Psychology, 63(2), 221-237. Stephenson, J. (2003). Growing, evolving HIV/AIDS pandemic is producing social and economic fallout. Journal of the American Medical Association, 289(1), 31-33. Stevens, J. (1996). Applied multivariate statistics for the social sciences (3rd ed.). Mahwah, NJ: Lawrence Erlbaum Associates. Thompson, B., & Daniel, L. (1996). Factor analytic evidence for the construct validity of scores: A historical overview and some guideline. Educational and Psychological Measurement, 56(2), 197-208. Thompson, B., & Vacha-Haase, T. (2000). Psychometrics is datametrics: The test is not reliable. Educational and Psychological Measurement, 60(2), 174-195. UNAIDS (Joint United Nations Program on HIV/AIDS). (1997). Impact of HIV and sexual health education on the sexual behaviour of young people: A review update. Retrieved April, 2004, from http://www.unaids.org/publications/documents/children/schools/grunskme.pdf 179

PAGE 191

UNAIDS. (2000). El Salvador. Epidemiological fact sheet on HIV/AIDS and sexually transmitted infections. Retrieved April, 2004, from http://www.unaids.org/en/default.asp UNAIDS. (2001). The global strategy framework on HIV/AIDS. Retrieved May, 2003. Retrieved May, 2002, from http://www.unaids.org/publications/documents/care/general/JC637-GloblalFrame UNAIDS. (2002). The global HIV/AIDS epidemic 2002-2003. [CD-ROM]. From www.unaids.org UNAIDS. (2004). AIDS epidemic update: December 2003. Retrieved April, 2004, from www.unaids.org/unaids/en/Resources/Publications/corporate+publications/AIDS +epidemic+update+-+December+2003.asp United Nations Children Fund (UNICEF), Joint United Nations Program on HIV/AIDS (UNAIDS), & World Health Organization (WHO). (2002). Young people and HIV/AIDS: Opportunity in crisis. New York: United Nations Children's Fund (UNAIDS). World Health Organization. UNESCO. (2003). HIV/AIDS and education: A strategic approach. Paris: International Institute for Educational Planning. Retrieved August, 2003, from http:/portal.unesco.org/aids/iatt-education United States Agency for International Development (USAID). (2004). Country Profile. HIV/AIDS. El Salvador. Retrieved April, 2004, from http://www.usaid.gov/our_work/global_health/aids/Countries/lac/elsalvadbrief Vijver, F. J. R. v. d., & Leung, K. (1997). Methods and data analysis for cross-cultural research Fons van de Vijver, Kwok Leung. Thousand Oaks, CA: Sage Publications. Wagner, A. K., Gandek, B., Aaronson, N. K., Acquadro, C., Alonso, J., Apolone, G., Bullinger, M., Bjorner, J., Fukuhara, S., Kaasa, S., Leplege, A., Sullivan, M., Wood-Dauphinee, S., & Ware, J. E., Jr. (1998). Cross-cultural comparisons of the content of SF-36 translations across 10 countries: Results from the IQOLA Project. International Quality of Life Assessment. Journal of Clinical Epidemiology, 51(11), 925-932. Ware, J. E., Jr., Kosinski, M., Gandek, B., Aaronson, N. K., Apolone, G., Bech, P., Brazier, J., Bullinger, M., Kaasa, S., Leplege, A., Prieto, L., & Sullivan, M. (1998). The factor structure of the SF-36 Health Survey in 10 countries: Results from the IQOLA Project. International Quality of Life Assessment. Journal of Clinical Epidemiology, 51(11), 1159-1165. 180

PAGE 192

World Bank. (2003). HIV/AIDS in Central America. An overview of the epidemic and priorities for prevention. Retrieved June, 2003, from http://wbln0018.worldbank.org/LAC/LAC.nsf/ECADocByUnid2ndLanguage/A3 23B430E178853F85256DBF0054CE44?Opendocument 181

PAGE 193

APPENDICES 182

PAGE 194

Appendix A. The English Knowledge and Attitude Survey DIRECTIONS : Read each question. Carefully check the one answer that fits best. Some of the questions use the phrase "having sex." This means sexual intercourse. DO NOT put your name on this survey. Your answers will be kept secret. No one will know how you answered these questions. 1. What is your gender? a. Male____ b. Female____ 2. What grade are you in? a. 9 th ____ b.10 th ____ c. 11 th ____ d. 12 th ____ 3. What is your age? a. 13-14____ b. 15-16____ c.17-19____ 4. Have you received HIV education in school? a. Yes____ b. No____ 5. In what year was the education received? _________ 6. Where did you receive HIV education? a. School b. Community c. Church d. other __________ Please check one answer that best fits: I am sure it is True I think it is True I dont know I think it is False I am sure it is False 1. You can't get AIDS if you have sex only once or twice without a condom. 2. A person can "pass" an HIV antibody test (test negative) but still be infected with HIV. 3. Condoms are 100% effective in preventing HIV. 4. Males can pass HIV on to others through their semen. 5. You can get HIV by sitting on the seat of a toilet that a person with AIDS has used. 6. Abstinence from sex and drugs is the best way for teen-agers to avoid getting HIV. 7. You can get HIV from drinking from the same glass or water fountain that a person with AIDS drank from. 8. HIV can be found in semen, vaginal fluids, and blood. 183

PAGE 195

Appendix A (continued) I am sure it is True I think it is True I dont know I think it is False I am sure it is False 9. A person can get HIV by sharing drug needles. 10. HIV can be found in breast milk. 11. Once you are infected with HIV, you are infected for life. 12. People infected with HIV are usually very thin and sickly. 13. Some people have gotten HIV by swimming in the same pool as someone with AIDS. 14. You can get HIV from a mosquito bite. 15. If you want to keep from getting HIV, using "lambskin" condoms is just as good as using latex condoms. 184

PAGE 196

Appendix A (continued) DIRECTIONS: This survey asks you to say whether you agree or disagree with a set of statements. Please read each statement, then indicate whether you: Strongly Agree, Agree, are Not Sure, Disagree, or Strongly Disagree Indicate your response by marking an x in the box. Strongly Agree Agree Not Sure Disagree Strongly Disagree 1. If your friends want you to do something that you think might not be safe, you should at least try it. 2. Its okay not to have sex while you are a teen-ager. 3. I believe condoms should always be used if a person my age has sex, even if the two people know each other very well. 4. A teen-ager can inject drugs once in a while without a risk of getting infected with HIV. 5. Teen-agers are at risk of getting infected with HIV if they engage in sex without a condom. 6. To keep your friends, you should go along with most things your friends want you to do. 7. People who don't have sex before they get married are strange. 8. It is not smart to have sex without using a condom. 9. Using needles to inject steroids or drugs is a bad idea. 10. It's okay to have sex without a condom because your chance of getting infected with HIV is very low. 11. Teen-agers should learn how to resist pressures from their friends. 185

PAGE 197

Appendix A (continued) Strongly Agree Agree Not Sure Disagree Strongly Disagree 12. It's a good idea for teen-agers not to have sex. 13. People who share drug needles shouldn't worry because they probably won't get infected with HIV. 14. Teen-agers should realize that if they're not careful, they could get infected with HIV. 15. When friends want you to do things you don't feel like doing, there's no harm in going along. 16. Using a condom doesn't make sex less pleasurable. 17. Anyone who shares needles is taking a chance of getting infected with HIV. 18. If teen-agers are careful about choosing sexual partners, they won't get infected with HIV. 19. Teen-agers should be more willing to resist pressures from their friends. 20. These days it makes a lot of sense to wait to have sex until you get married. 21. If people think they might have sex during a date, they should carry a condom. 22. I believe its OK for people my age to have sex with several different people in the same month. 23. People who use condoms during sex don't trust the person they're with. 24. People who share drug needles should clean the needles with bleach. 186

PAGE 198

Appendix A (continued) Strongly Agree Agree Not Sure Disagree Strongly Disagree 25. HIV is something that teen-agers should think about when they date. 26. I believe its OK for people my age to have sex with a steady boyfriend or girlfriend. 27. I believe people my age should wait until they are older before they have sex. 28. I believe condoms should always be used if a person my age has sex. 29. I believe condoms should always be used if a person my age has sex, even if the girl uses birth control pills. Thank you for your time 187

PAGE 199

Appendix B. Sample Form for the Forward-translation into Spanish Directions : Please read the statement on the left. Write the correct Spanish translation on the right. If there is a problem in the translation look at the form at the end of the document. Write the item number and describe any difficulty involved in the translation. Thank you for your Help. Translator _____________. Original Version Item Spanish Translated Item. Comments The Knowledge and Attitude Survey: DO NOT put your name on this survey. Your answers will be kept secret. No one will know how you answered these questions. DIRECTIONS: Read each question. Carefully check the one answer that fits best. Some of the questions use the phrase "having sex." This means sexual intercourse. 1.What is your gender? A. Male B. Female 2. What grade are you in? A. 9 th B.10 th C. 11 th D. 12 th 3. What is your age? A. 13-14 B. 15-16. C. 17-19 4. Have you received HIV education in school? A. Yes B. No 5. You can't get AIDS if you have sex only once or twice without a condom. 6. A person can "pass" an HIV antibody test (test negative) but still be infected with HIV. Note: Form adapted from Beaton, D., Bombardier, C., Guillemin, F., & Ferraz, M.B. (2002). 188

PAGE 200

Appendix C. Sample Form Sent to Bilingual Reviewers for the Synthesis Directions: Read the statement in English. Then read the translated versions and select the best one and place it at the end of the form. If there is a discrepancy or word that does not fit the translation write a note (use red font) in the block. Original English Version Item Forward Translated Item. #1. Forward Translated Item. #2 Final Translated item. Write a note in red font if there is a discrepancy. The Knowledge and Attitude Survey DO NOT put your name on this survey. Your answers will be kept secret. No one will know how you answered these questions. DIRECTIONS: Read each question. Carefully check the one answer that fits best. Some of the questions use the phrase "having sex." This means sexual intercourse. 1. What is your gender? A. Male B. Female 2. What grade are you in? A. 9 th B.10 th C. 11 th D. 12 th 3. What is your age? A. 13-14 B. 15-16. C.17-19 4. Have you received HIV education in school? A. Yes B. No 1. You can't get AIDS if you have sex only once or twice without a condom. 2. A person can "pass" an HIV antibody test (test negative) but still be infected with HIV. 3. Condoms are 100% effective in preventing HIV. 4. Males can pass HIV on to others through their semen. 5. You can get HIV by sitting on the seat of a toilet that a person with AIDS has used. 6. Abstinence from sex and drugs is the best way for teen-agers to avoid getting HIV. 7. You can get HIV from drinking from the same glass or water fountain that a person with AIDS drank from. Note: Form adapted from Beaton, D., Bombardier, C., Guillemin, F., & Ferraz, M.B. (2002). 189

PAGE 201

Appendix D. Sample of Back-translation Form: Please translate the Spanish items and place the translation in the appropriate block. Please write a note and make a list (at the bottom of the document) if you encountered a difficult item. Cuestionario de Conocimientos y Actitudes de HIV/SIDA No escriba su nombre en esta encuesta. Sus respuestas sern guardadas completamente confidenciales. Nadie sabr como contesto a las preguntas. INSTRUCCIONES: Lea cada pregunta. Con cuidado marque la contestacin que sea apropiada. Algunas preguntas usan la frase teniendo sexo. Esto significa relaciones sexuales. 1. Cal es tu gnero? A. masculino B. Femenino 1. 2. En que grado ests? A. 9 B. 10 C. 11 D. 12 2. 3. Que edad tienes? A. 13-14 B. 15-16 C. 17-19 3. 4. Has recibido educacin sobre El SIDA en la escuela? A. Si B. No 4. 5. Tu no puedes contraer el SIDA si tu tienes sexo una o dos veces sin un condn 5. 6. Una persona puede pasar una prueba de anticuerpos del SIDA (resultados negativos) y aun estar infectado por el SIDA. 6. 7. Los condones son 100% efectivos en prevenir VIH 7. 8. Los hombres pueden pasar VIH a travs de su semen 8. Note: Form adapted from Beaton, D., Bombardier, C., Guillemin, F., & Ferraz, M.B. (2002). 190

PAGE 202

Appendix E. Form for Semantic Equivalence (condensed version) Dear Reviewer: Thank you for participating in this review. I have translated an instrument developed in the USA to assess HIV/AIDS knowledge and attitudes. I would like to use this survey with adolescent students in El Salvador. I need to find out if the translation was done properly and if it is linguistically the same. If you have questions about the process do not hesitate to ask. DIRECTIONS : I am asking that you compare the English and Spanish version and to mark in the Box X if you agree that the English and Spanish statements are equal in semantic, idiomatic and conceptual structure. If you have comments add them in the box. Please email this back to me at Czometaiii@aol.com or call me at 28-0871 (I will be in El Salvador September 18 to October 20th) and I will can pick it up. Could you please send me your name and a brief description of your professional qualifications? Thank you very much! The definition of semantic, idiomatic and conceptual equivalence. Variable Definition Semantic words and the composition of the sentences have the same meaning as the original item. Idiomatic Idioms and colloquialisms have been translated into expressions understood by Salvadorian adolescents. These are words that are unique to one culture but not understood by the culture. Conceptual The ideas/concepts are the same and will be understood by students in El Salvador. The concepts are relevant to the experiences of Salvadorian adolescents. Directions: After each statement circle Pass= ____ or mark an X (for computer version) next to pass if both statements are equal. Knowledge Semantic Idiomatic Conceptual 1 You can't get AIDS if you have sex only once or twice without a condom. Tu no puedes contraer el SIDA si tu tienes sexo una o dos veces sin un condn. Pass = ______ Fail = no response Pass = ______ Fail = no response Pass = ______ Fail = no respons e 2 A person can "pass" an HIV antibody test (test negative) but still be infected with HIV. Una persona puede pasar una prueba de anticuerpos de VIH (resultados negativos) y aun estar infectado con VIH. Pass = ______ Fail = no response Pass = ______ Fail = no response Pass = ______ Fail = no response 3 Condoms are 100% effective in preventing HIV. Los condones son 100% efectivos en prevenir el VIH. Pass = ______ Fail = no response Pass = ______ Fail = no response Pass = ______ Fail = no response 4 Males can pass HIV on to others through their semen. Los hombres pueden p asar VIH a travs de su semen. Pass = ______ Fail = no response Pass = ______ Fail = no response Pass = ______ Fail = no response 191

PAGE 203

Appendix E (continued) 5 You can get HIV by sitting on the seat of a toilet that a person with AIDS has used. T puedes adquirir VIH al sentarte en un inodoro que fue usado por una persona con VIH. Pass = ______ Fail = no response Pass = ______ Fail = no response Pass = ______ Fail = no response 6 Abstinence from sex and drugs is the best way for teen-agers to avoid getting HIV. Abstinencia de sexo y de las drogas es la mejor forma para que los adolescentes eviten contraer el VIH. Pass = ______ Fail = no response Pass = ______ Fail = no response Pass = ______ Fail = no response Attitudes Semantic Idiomatic Conceptual 1 If your friends want you to do something that you think might not be safe, you should at least try it. Si tus amigos quieren que tu hagas algo y t crees que no es seguro, t deberas por lo menos tratar de hacerlo. Pass = ______ Fail = no response Pass = ______ Fail = no response Pass = ______ Fail = no response 2 Its okay not to have sex while you are a teen-ager. Est bien no tener sexo mientras sea un(a) adolescente. Pass = ______ Fail = no response Pass = ______ Fail = no response Pass = ______ Fail = no response 3 I believe condoms should always be used if a person my age has sex, even if the two people know each other very well. Creo que los condones debern siempre ser usados si una persona de mi edad tiene sexo, aunque las dos p ersonas se conozcan muy bien. Pass = ______ Fail = no response Pass = ______ Fail = no response Pass = ______ Fail = no response 4 A teen-ager can inject drugs once in a while without a risk of getting infected with HIV. Un adolescente puede inyectarse drogas de vez en cuando sin riesgo de contraer VIH. Pass = ______ Fail = no response Pass = ______ Fail = no response Pass = ______ Fail = no response 5 Teen-agers are at risk of getting infected with HIV if they engage in sex without a condom. Los adolescentes tienen el riesgo de ser infectados con VIH si tienen relaciones sexuales sin usar un condn. Pass = ______ Fail = no response Pass = ______ Fail = no response Pass = ______ Fail = no response 6 To keep your friends, you should go along with most things your friends want you to do. Para mantener tus amistades, t debes hacer la mayora de las cosas que tus amigos quieren que t hagas. Pass = ______ Fail = no response Pass = ______ Fail = no response Pass = ______ Fail = no response 192

PAGE 204

Appendix F Cultural and Content Validity Form Estimado Panelista: A usted se le ha solicitado participar en un panel para revisar y evaluar un instrumento que medir los conocimientos y actitudes de VIH/SIDA entre adolescentes de bachillerato en San Salvador. Quiero saber si las palabras y las ideas de estas preguntas son culturalmente apropiadas o si van a hacer entendidas por estudiantes Salvadoreos. Tambin, necesito saber si las preguntas acerca de los conocimientos y actitudes miden estos dos conceptos apropiadamente. Solicito preguntas adicionales que podrn mejorar esta encuesta. Entonces las tres metas son: 1. Determinar si las preguntas son apropiadas para la cultura Salvadorea. Si no son, por favor corrija la pregunta o sugiera una nueva. 2. Determinar si el componente del conocimiento es apropiado. 3. Determinar si las actitudes son apropiadas. Las cinco dimensiones son: Abstinencia, uso de condones, drogas/esteroides, amenaza de infeccin de VIH e influencia de los amigos. Por favor lea cada pregunta y provea su opinin de acuerdo a la escala siguiente: No est apropiado = 1, Poco Apropiado = 2, Moderadamente apropiado = 3, Muy apropiado = 4. Preguntas Conocimientos de VIH/SIDA Cultura Salvadorea Comentarios o Preguntas Adicionales 1 Tu no puedes contraer el SIDA si tu tienes sexo una o dos veces sin un condn. 12 -3 4 12 -3 4 2 Una persona puede pasar una prueba de anticuerpos de VIH (resultados negativos) y aun estar infectado con VIH. 12 -3 4 12 -3 4 3 Los condones son 100% efectivos en prevenir el VIH. 12 -3 4 12 -3 4 4 Los hombres pueden pasar VIH a travs de su semen. 12 -3 4 12 -3 4 5 T puedes adquirir VIH al sentarte en un inodoro que fue usado por una persona con VIH. 12 -3 4 12 -3 4 6 Abstinencia de sexo y de las drogas es la mejor forma para que los adolescentes eviten contraer el VIH. 12 -3 4 12 -3 4 193

PAGE 205

Appendix G. Letter of permission from the government 194

PAGE 206

Appendix H. Informed Consent for Students in the Pretest INFORMACION PARA PRETEST. PERMISO PARA PARTICIPANTES Y PADRES DE FAMILIA Estimado Padre de Familia: Su hija/hijo est invitado a participar en una encuesta que forma parte de un proyecto de la Universidad del Sur de Florida (USF), de la ciudad de Tampa, Florida, Estados Unidos. A travs de sta carta le pedimos permiso para que su hijo(a) llene un cuestionario annimo y que los cientficos les puedan hacer una entrevista y preguntar sobre el entendimiento. Por favor lea esta carta que explica la naturaleza del estudio y si tiene preguntas por favor llame me y le explicar el estudio. Si usted est de acuerdo que ellos tomen parte en ste estudio por favor firme el papel y regrselo con su hijo(a). Hay un pago de $10 por participar. Tambin, si el joven no da su aprobacin ellos no tienen que tomar parte en el estudio. Muchas Gracias por su ayuda. Ttulo del Estudio : La traduccin, adaptacion y validacion de un instrumento del Centro de Control de Enfermedades (USA) para medir los conocimientos y actitudes de VIH/SIDA entre una poplacion de adolescentes en bachillerato. Investigador Principal : Carlos Salvador Zometa MSPH (estudiante de doctorado) Lugar : El Salvador, escuelas de San Salvador y sus alrededores. Objetivo General: Evaluar los conocimientos y actitudes del VIH/SIDA en los jvenes de bachillerato en San Salvador, as obtener un equivalente cultural de un taller desarrollado por el Centro de Control de Enfermedades (CDC). Poblacin : adolescentes estudiantes entre los 14 y 18 aos, de ambos sexos, de San Salvador y sus alrededores. Instrumento : Se pide al joven que complete una encuesta para poder medir el conocimiento y sus actitudes hacia el VIH/SIDA. Tiempo aproximado : una hora. Beneficios : No habr beneficios personales al participar en este estudio. Pero ayudara a incrementar los conocimientos sobre esta problemtica. Riesgos : No hay riesgos conocidos o anticipados con la encuesta. Normas establecidas : La participacin es voluntaria y si deciden o participar, no perdern ningn beneficio que recibirn. Una vez iniciado el proceso de llenado de la encuesta, es permitido abandonarla o no empezar a llenarla. La encuesta es annima, no es necesario poner el nombre del participante. Los datos son privados y la informacin ser procesada por el Departamento de Salud y Recursos Humanos y la Comisin Institucional de Revisin de la Universidad del Sur de La Florida. El estudio se publicara en combinacin con otros datos, y en ningn caso revelara informacin particular de los participantes. Al finalizar el estudio, los datos obtenidos sern destruidos. Mayor informacin con Carlos Salvador Zometa al telfono 28-0871. Cualquier reclamo llamar a la Divisin de Quejas de Investigacin de la Universidad del Sur de Florida al 00-144-00-1(813) 974-5638. 195

PAGE 207

Appendix H (continued) Aprobacin para que el jove n tome parte en el estudio : Yo libremente doy mi aprobacin para que mi hijo (a) tome parte en ste estudio. Me doy por informado de lo anterior y he recibido una copia de esta forma. Firma de los padres Nombre de los padres Fecha Declaracin del Investigador : Yo he explicado cuidadosamente el contenido y la natu raleza de ste protocolo. Y por lo tanto certifico que con todo mi conocimiento la persona que firma esta forma da consentimiento entiende la naturaleza, demandas, riesgos y beneficios al participar en este estudio. Firma del investigador ____________________ _______________ Nombre del investigador fecha Si el joven no quiere llenar la encuesta, por favor explique las razones aqu: ________________________ __________________________ ________ Firma de los padres del Escriba el nombre de los padres Fecha joven participantes ________________________ ________ Firma del jven participante Nombre del jven participante Fecha Forma de consentimiento para ser firmada por los jvenes participantes (Lic.Carlos S Zometa. MSPH) me ha explicado acer ca de sta estudio llamado (Adaptacin Cultural Combinada, Confiable y Vlida de una Encuesta para Medir el Conocimiento y la Actitud hacia el VIH de los jvenes de El Salvador). Yo estoy de acuerdo en tomar parte en ste estudio. _________________________ _____________________ ____________ Firma del jven tomando parte Escriba su nombre Fecha En este estudio ________________________ ____________________ _____________ Firma de los padres del jven Escriba su nombre Fecha Tomando parte en ste estudio _________________________ __________________ _____________ Firma de la persona que recibe Escriba su nombre Fecha consentimiento _________________ __________________ ______________ Firma de un testigo Escriba su nombre Fecha 196

PAGE 208

Appendix I. General Informed Consent Form INFORMACION GENERAL Y PERMISO PARA PARTICIPANTES Estimado Padre de Familia: Su hija/hijo est invitado a participar en una encuesta que forma parte de un proyecto de la Universidad del Sur de Florida (USF), de la ciudad de Tampa, Florida, Estados Unidos. A travs de sta carta le pedimos permiso para que su hijo(a) llene un cuestionario annimo en la escuela. Por favor lea esta carta que explica la naturaleza del estudio y si tiene preguntas por favor llame me y le explicar el estudio. Si usted est de acuerdo que ellos tomen parte en ste estudio nada mas es necesario que su hijo(a) llene el cuestionario en la escuela. Tambin, si el joven no da su aprobacin ellos no tienen que tomar parte en el estudio. Muchas Gracias por su ayuda. Ttulo del Estudio : La traduccin, adaptacion y validacion de un instrumento del Centro de Control de Enfermedades (USA) para medir los conocimientos y actitudes de VIH/SIDA entre una poplacion de adolescentes. Investigador Principal : Carlos Salvador Zometa MSPH (estudiante de doctorado) Lugar : El Salvador, escuelas de San Salvador y sus alrededores. Objetivo General: Evaluar los conocimientos y actitudes del VIH/SIDA en los jvenes de bachillerato en San Salvador, as obtener un equivalente cultural de un taller desarrollado por el Centro de Control de Enfermedades (CDC). Poblacin : adolescentes estudiantes entre los 14 y 18 aos, de ambos sexos, de San Salvador y sus alrededores. Instrumento : Se pide al joven que complete una encuesta para poder medir el conocimiento y sus actitudes hacia el VIH/SIDA. Tiempo aproximado : Media hora. Beneficios : No habr beneficios personales al participar en este estudio. Pero ayudara a incrementar los conocimientos sobre esta problemtica. Riesgos : No hay riesgos conocidos o anticipados con la encuesta. Normas establecidas : La participacin es voluntaria. Si deciden a participar o no participar, no perdern ningn beneficio que recibirn. Una vez iniciado el proceso de llenado de la encuesta, es permitido abandonarla o no empezar a llenarla. La encuesta es annima, no es necesario poner el nombre del participante Los datos son privados y la informacin ser procesada por el Departamento de Salud y Recursos Humanos y la Comisin Institucional de Revisin de la Universidad del Sur de La Florida. El estudio se publicara en combinacin con otros datos, y en ningn caso revelara informacin particular de los participantes. Al finalizar el estudio, los datos obtenidos sern destruidos. Mayor informacin con Carlos Salvador Zometa al telfono 28-0871. Cualquier reclamo llamar a la Divisin de Quejas de Investigacin de la Universidad del Sur de Florida al 00-144-00-1(813) 974-5638. 197

PAGE 209

Appendix I (continued) Firme si no aprueba que el jve n tome parte en el estudio : Yo no doy mi aprobacin para que mi hijo (a) tome parte en ste estudio. Me doy por informado de lo anterior y he recibido una copia de esta forma. Firma de los padres Nombre de los padres Fecha Declaracin del Investigador : Yo he explicado cuidadosamente el contenido y la natu raleza de ste protocolo. Y por lo tanto certifico que con todo mi conocimiento la persona que firma esta forma da consentimiento entiende la naturaleza, demandas, riesgos y beneficios al participar en este estudio. Firma del investigador ___________________ _______ Nombre del investigador Fecha 198

PAGE 210

Appendix J. Informed Consent for Test-Retest. INFORMACION GENERAL Y PERMISO PARA PARTICIPANTES Estimado Padre de Familia: Su hija/hijo est invitado a participar en una encuesta que forma parte de un proyecto de la Universidad del Sur de Florida (USF), de la ciudad de Tampa, Florida, Estados Unidos. A travs de sta carta le pedimos permiso para que su hijo(a) llene un cuestionario annimo en la escuela. Tambin un grupo pequeo van a tomar el cuestionario dos veces para un estudio de recopilacion. Por favor lea esta carta que explica la naturaleza del estudio y si tiene preguntas por favor llame me y le explicar el estudio. Si usted est de acuerdo que ellos tomen parte en ste estudio nada mas es necesario que su hijo(a) llene el cuestionario en la escuela. Tambin, si el joven no da su aprobacin ellos no tienen que tomar parte en el estudio. Muchas Gracias por su ayuda. Ttulo del Estudio : La traduccin, adaptacion y validacion de un instrumento del Centro de Control de Enfermedades (USA) para medir los conocimientos y actitudes de VIH/SIDA entre una poplacion de adolescentes. Investigador Principal : Carlos Salvador Zometa MSPH (estudiante de doctorado) Lugar : El Salvador, escuelas de San Salvador y sus alrededores. Objetivo General: Evaluar los conocimientos y actitudes del VIH/SIDA en los jvenes de bachillerato en San Salvador, as obtener un equivalente cultural de un taller desarrollado por el Centro de Control de Enfermedades (CDC). Poblacin : adolescentes estudiantes entre los 14 y 18 aos, de ambos sexos, de San Salvador y sus alrededores. Instrumento : Se pide al joven que complete una encuesta para poder medir el conocimiento y sus actitudes hacia el VIH/SIDA. Tambin un grupo pequeo va a tomar el cuestionario dos veces para medir la recopilacin de datos. Tiempo aproximado : Media hora. Beneficios : No habr beneficios personales al participar en este estudio. Pero ayudara a incrementar los conocimientos sobre esta problemtica. Riesgos : No hay riesgos conocidos o anticipados con la encuesta. Normas establecidas : La participacin es voluntaria. Si deciden a participar o no participar, no perdern ningn beneficio que recibirn. Una vez iniciado el proceso de llenado de la encuesta, es permitido abandonarla o no empezar a llenarla. La encuesta es annima, no es necesario poner el nombre del participante Los datos son privados y la informacin ser procesada por el Departamento de Salud y Recursos Humanos y la Comisin Institucional de Revisin de la Universidad del Sur de La Florida. El estudio se publicara en combinacin con otros datos, y en ningn caso revelara informacin particular de los participantes. Al finalizar el estudio, los datos obtenidos sern destruidos. Mayor informacin con Carlos Salvador Zometa al telfono 28-0871. Cualquier reclamo llamar a la Divisin de Quejas de Investigacin de la Universidad del Sur de Florida al 00-144-00-1(813) 974-5638. 199

PAGE 211

Appendix J (continued) Firme si no aprueba que el jve n tome parte en el estudio : Yo no doy mi aprobacin para que mi hijo (a) tome parte en ste estudio. Me doy por informado de lo anterior y he recibido una copia de esta forma. Firma de los padres Nombre de los padres Fecha Declaracin del Investigador : Yo he explicado cuidadosamente el contenido y la natu raleza de ste protocolo. Y por lo tanto certifico que con todo mi conocimiento la persona que firma esta forma da consentimiento entiende la naturaleza, demandas, riesgos y beneficios al participar en este estudio. Firma del investigador ___________________ _______ Nombre del investigador Fecha 200

PAGE 212

Appendix K. Results of Forward-translation of English Instrument Original English Version Item Forward-translation #1. Forward-translation #2. The Knowledge and Attitude Survey El Conocimiento y Actitudes de la Encuesta Cuestionario de Conocimiento y Actitudes DO NOT put your name on this survey. Your answers will be kept secret. No one will know how you answered these questions. NO ESCRIBA su nombre en sta forma, sus respuestas se mantendrn secretas. Nadie va a saber como usted ha contestado las preguntas. No escriba su nombre en esta encuesta. Sus respuestas sern guardadas completamente confidenciales. Nadie sabr como contesto a las preguntas. DIRECTIONS: Read each question. Carefully check the one answer that fits best. Some of the questions use the phrase "having sex." This means sexual intercourse. DIRECCIONES: Lea cada pregunta. Cuidadosamente marque la respuesta que le parezca adecuada. En algunas de las preguntas se us la frase teniendo sexo esto significa contacto sexual. INSTRUCCIONES Lea cada pregunta. Con cuidado marque la contestacin que sea apropiada. Algunas preguntas usan la frase teniendo sexo Esto significa relaciones sexuales. 1. What is your gender? A. Male B. Female Cal es tu gnero? A. masculino B. Femenino Cul es su gnero? A. Hombre B. Mujer 2. What grade are you in? A. 9 th B. 10 th C. 11 th D. 12 th En que grado ests? A. 9 B.10 C. 11 D. 12 En que grado ests? A. 9no B. 10 C. 11avo D. 12avo 3. What is your age? A. 13-14 B. 15-16. C.17-19 Qu edad tienes? A. 13-14 B. 15-16 C. 17-19 Que edad tienes? A. 13-14 B.15-16 C. 17-19 4. Have you received HIV education in school? A. Yes B. No Has recibido educacin sobre El SIDA en la escuela? A. Si B. No Has recibido educacin sobre El SIDA en la escuela? A. S B. No 1. You can't get AIDS if you have sex only once or twice without a condom. Tu no puedes contraer SIDA si tienes sexo una vez o dos veces solamente sin usar condones. Tu no puedes contraer el SIDA si tu tienes sexo una o dos veces sin un condn. 2. A person can "pass" an HIV antibody test (test negative) but still be infected with HIV. Unaexm persona puede pasar un en de anticuerpos de VIH (exmen es negativo) pero puede estar infectado con VIH. Una persona puede pasar una prueba de anticuerpos del SIDA (resultados negativos) y aun estar infectado por el SIDA. 3. Condoms are 100% effective in preventing HIV. Los condones son 100% efectivos en prevenir VIH. Los condones son 100% efectivos para prevenir el SIDA. 4. Males can pass HIV on to others through their semen. Los hombres pueden pasar VIH a travs de su semen. Los hombres pueden contagiar a otros con el SIDA por medio de su semen 5. You can get HIV by sitting on the seat of a toilet that a person with AIDS has used. Tu puedes adquirir VIH al sentarte en un inodoro que fu usado por una persona con VIH. Ud. Puede contraer el SIDA al sentarse en el asiento del inodoro donde una persona con SIDA se ha sentado. 201

PAGE 213

Appendix K. (continued) 6. Abstinence from sex and drugs is the best way for teen-agers to avoid getting HIV. Abstinencia de sexo y drogas es la major forma para que los jvenes eviten el VIH. Abstinencia del sexo y de las drogas es la mejor forma para que los adolescents eviten contraer el SIDA. 7. You can get HIV from drinking from the same glass or water fountain that a person with AIDS drank from. Puedes adquirir VIH en los tomaderos de agua o tomando agua de un vaso usado por una persona con VIH. Ud. puede contraer el SIDA al tomar liquidos en el mismo vaso o en la fuente del agua donde una persona con SIDA ha tomado. 8. HIV can be found in semen, vaginal fluids, and blood. VIH puede estar presente en el semen, fluidos vaginales, y sangre. El SIDA puede encontrarse en el semen, fluidos vaginales y la sangre. 9. A person can get HIV by sharing drug needles. Puede una persona adquirir VIH al compartir agujas para la droga. Una persona puede contagiarse con el SIDA compartiendo las agujas de las drogas. 10. HIV can be found in breast milk.VIH puede estar presente en la leche maternal. El SIDA puede encontrarse en la leche maternal 11. Once you are infected with HIV, you are infected for life. Una vez que tu has sido infectado con VIH, tu ests infectado de por vida. Una vez que has sido contagiado con el SIDA, estas contagiado de por vida. 12. People infected with HIV are usually very thin and sickly. Personas infectadas con VIH son bien delgadas y se notan enfermizas. Las personas contagiadas con el SIDA generalmente se ven muy delgados y enfermizos 13. Some people have gotten HIV by swimming in the same pool as someone with AIDS. Algunas personas han adquirido VIH nadando en la misma piscina con una persona con SIDA. Algunas personas han contraido el SIDA al nadar en la misma piscina que una persona que lo tiene 14. You can get HIV from a mosquito bite. Puedes t adquirir VIH de una picada de mosquito. Ud. puede contraer el SIDA del piquete de un mosquito 15. If you want to keep from getting HIV, using "lambskin" condoms is just as good as using latex condoms. Si tu quieres protejerte de contraer VIH, usando condones de piel de cordero es tan efectivo como usar condones de latex. Si Ud. desea evitar contraer el SIDA,usando condoneslambskin es tan eficiente como usando condones de latex 16. A new vaccine has been developed for the treatment of AIDS. Sabes si hay alguna vacuna que ha sido descubierta para curar el VIH? Una nueva vacuna ha sido desarrollada para tratar el SIDA 17. If a pregnant woman has AIDS, there is a chance it may harm her unborn baby. Si una mujer preada tiene SIDA, existe la posibilidad que el feto sufra daos? Si una mujer embarazada tiene el SIDA, hay un chance que contagiar a su beb que no ha nacido. 1. If your friends want you to do something that you think might not be safe, you should at least try it. Si un amigo quiere que tu hagas algo y t crees que no es seguro, t deberas por lo menos tratar de hacerlo. Si sus amigos desean que Ud. haga algo que Ud. no considera seguro, Ud. por lo menos debe de tratar de hacerlo 202

PAGE 214

Appendix K. (continued) 2. It's okay not to have sex while you are a teen-ager. Est bien no tener sexo cuando t eres an un jovencito. Est correcto no tener sexo mientras sea un(a) adolescente. 3. I believe condoms should always b e used if a person my age has sex, even if the two people know each other very well.*Yo creo que los condones debern siempre ser usados si una persona de mi edad tiene sexo, aunque las dos personas se conozcan muy bien. Considero que los condones siempre deben usarse aunque dos personas se conozcan bien. 4. A teen-ager can inject drugs once in a while without a risk of getting infected with HIV. Un joven puede inyectarse drogas de vez en cuando sin riesgo de contraer VIH. Un adolescente se puede inyectar drogas de vez en cuando sin correr el riesgo de obtener el SIDA 5. Teen-agers are at risk of getting infected with HIV if they engage in sex without a condom. Los jovenes tienen riesgo de infectarse con VIH si ellos practican sexo sin usar condones. Los adolescents tienen el riesgo de ser infectados con el virus del SIDA si tienes relaciones sexuales sin usar un condn. 6. To keep your friends, you should go along with most things your friends want you to do. Para no perder tus amigos, t tienes que complacerlos con todas las cosas que ellos quieren que t hagas. Para mantener sus amistades, Ud. debe hacer las mayora de las cosas que sus amigos quieren que Ud. haga 7. People who don't have sex before they get married are strange. Las personas que no tienen sexo ante s de casarse son raras. Raras as in frequency Las personas que no tienen sexo antes casarse son extraos. 8. It is not smart to have sex without using a condom. No es muy inteligente tener sexo sin usar condones. N o es inteligente tener sexo sin usar un condn 9. Using needles to inject steroids or drugs is a bad idea. Usar agujas para inyectarse esteroides no es una mala idea. Added drugs and took out no Las personas que compartes agujas p ara las drogas, no deben preocuparse p orque posiblemente no sern infectados por el SIDA. 10. It's okay to have sex without a condom because your chance of getting infected with HIV is very low. Esta bien tener sexo sin usar condones ya que tus chances de contraer VIH son muy bajos. Esta correcto tener sexo sin usar un condn porque es mu y bajo el chance de ser infectado con el SIDA. 11. Teen-agers should learn how to resist pressures from their friends. Los jvenes tienen que aprender a resistir las presiones de sus amigos. Los adolescents deben saber resistir las presiones de sus amigos 12. It's a good idea for teen-agers not to have sex. Es una buena idea si los jvenes no tienen sexo. Es una buena idea para los adolescentes no tener sexo. 13. People who share drug needles shouldn't worry because they probably won't get infected with HIV. Las personas que comparten agujas para drogas no deben preocuparse porque ellos probablemente no se infectaran de VIH. Las personas que compartes agujas p ara las drogas, no deben preocuparse p orque posiblemente no sern infectados por el SIDA. 203

PAGE 215

Appendix K. (continued) 14. Teen-agers should realize that if they're not careful, they could get infected with HIV. Los jvenes tienen que saber que si no son cuidadosos, ellos pueden ser infectados con VIH. Los adolescenters deben pensar que si no tienen cuidado, pueden ser infectados con el SIDA. 15. When friends want you to do things you don't feel like doing, there's no harm in going along. Cuando tus amigos quieren que t hagas cosas que t sientes no deben hacerse, no hay ningn riesgo si t las haces. Cuando los amigos quieren que haga cosas que Ud. no se siente con deseos de hacer, no hay peligro en seguir la corriente 16. Using a condom doesn't make sex less pleasurable. Es menos el placer del sexo si se usan los condones. Usando un condn, no hace el sexo menos placentero 17. Anyone who shares needles is taking a chance of getting infected with HIV. Cualquiera que comparte agujas se est tomando el riesgo de ser infectado con VIH. Cualquiera que comparte las agujas, est tomando el chance de ser infectados con el SIDA. 18. If teen-agers are careful about choosing sexual partners, they won't get infected with HIV. Si los jvenes son cuidadosos al escoger su pareja para sexo, ellos no sern infectados con VIH. Si los adolescents tienen cuidado seleccionando sus compaeros sexuales, ellos no serninfectados por el SIDA 19. Teen-agers should be more willing to resist pressures from their friends. Los jvenes deberan resistir las presiones de sus amigos. Los adolescents deben tener mas b uena voluntad para resistir las presiones de sus amigos. 20. These days it makes a lot of sense to wait to have sex until you get married. En sta poca hace mas sentido esperar a tener sexo al casarse. En estos tiempos es mas sensato esperarse a tener sexo hasta que uno se case 21. If people think they might have sex during a date, they should carry a condom. Si una persona piensa que debe tener sexo al salir con alguien, ellos deberan traer un condn. Si las personas consideran que pueden tener sexo durante una salida, es mejor llevar consigo un condn 22. I believe its OK for people my age to have sex with several different people in the same month.* Yo creo que est bien que las personas de mi edad tengan sexo con diferentes personas y en el mismo mes Considero que es correcto para personas de mi edad tener sexo con diferentes personas en el mismo mes 23. People who use condoms during sex don't trust the person they're with. La persona que usa condones cuando tiene sexo no confa en su pareja. Las personas que usan condones durante el sexo, no confianen la persona con quien estan. 24. People who share drug needles should clean the needles with bleach. Las personas que comparten agujas para drogas deberan limpiar stas con leja. Las personas que comparten las agujas para las drogas, deben limpiarlas con bleach. 25. HIV is something that teen-agers should think about when they date. VIH es algo en que los jvenes deberan meditar cuando van a tener una cita. El SIDA es un asunto que los adolescents deben de pensar cuando salen con alguien 204

PAGE 216

Appendix K. (continued) 26. I believe its OK for people my age to have sex with a steady boyfriend or girlfriend.*Yo creo que est bien que las personas de mi edad tengan sexo con el novio o novia. Creo que esta correcto para personas de mi edad tener sexo con un novio o una novia. 27. I believe people my age should wait until they are older before they have sex.*Yo creo que las personas de mi edad deberan esperar un poco ms de tiempo para tener sexo. Creo que las personas de mi edad deben esperar hasta que tengan mas edad antes de tener sexo. 28. I believe condoms should always be used if a person my age has sex.*Yo creo que siempre se debe de usar condones cuando una persona de mi edad tiene sexo. Creo que los condones siempre deben ser usados por una persona de mi edad para tener sexo. 29. I believe condoms should always be used if a person my age has sex, even if the girl uses birth control pills.*Yo creo que los condones debern siempre ser usados si una persona de mi edad tiene sexo, aunque la jven use pastillas para no salir embarazada Considero que los condones deben usarse si una persona de mi edad tiene sexo, aunque la joven use pastillas anticonceptivas. I am sure it's true. Yo estoy seguro que es verdad Estoy seguro que es verdad I think it's true. Yo creo que es verdad Creo que es verdad I don't know. Yo no s No se I think it's false. Yo creo que esto es falso Creo que es falso I am sure it's false Yo estoy seguro que es falso Estoy seguro que es falso YOUR VIEWS Tus Puntos de Vista SUS PUNTOS DE VISTA DIRECTIONS: This survey asks you to say whether you agree or disagree with a set of statements. Please read each statement, then indicate whether you: DIRECCIONES: En sta encuesta se te pide que digas cuando estas de acuerdo y cuando no estas de acuerdo con lo que se dice aqu. Por favor lea cada pregunta, entonces indica cuando t ests Este cuestionario le hace preguntas para para saber si Ud. est de acuerdo o no con ciertas aseveraciones. Por favor lea cada una y conteste si Ud. Strongly Agree (SA), Agree (A), are Not Sure (NS), Disagree (D),or Strongly Disagree(SD), by circling the answer you want. Completamente de Acuerdo (CA), De Acuerdo (DA), No estas Seguro (NES), No estoy De Acuerdo (NEDA), Completamente en Desacuerdo (CED), hacer un crculo en la respuesta elegida. Esta de Acuerdo completamente (CA), Esta de Acuerdo (EA), No esta segura (NS), Esta en desacuerdo (DSA), o esta en complete desacuerdo (CDA). Poniendo un circulo en la contestacion que prefiere. Strongly Agree (SA) Completamente de Acuerdo (CA) De acuerdo completamente (CA) Agree (A) De Acuerdo (DA) De Acuerdo (EA) Are Not Sure (NS) No Estoy Seguro (NES) No esta seguro (NS) Disagree (D) No Estoy de Acuerdo (NEDA) Desacuerdo (DSA) Or Strongly Disagree (SD) Completamente en Desacuerdo (CED) Completamente en desacurdo (CDA) Note. Indicates items from Basen-Engquist et al. (1999). Form adapted from Beaton, D., Bombardier, C., Guillemin, F., & Ferraz, M.B. (2002). 205

PAGE 217

Appendix L. Results of Synthesis of Translations by the Bilingual Reviewers. Original English Version Item Bilingual Reviewer #1. Bilingual Reviewer #2. Bilingual Reviewer #4 The Knowledge and Attitude Survey Cuestionario de Conocimientos y Actitudes de HIV/SIDA El Conocimiento y Actitudes de la Encuesta DO NOT put your name on this survey. Your answers will be kept secret. No one will know how you answered these questions. No escriba su nombre en esta encuesta. Sus respuestas sern guardadas completamente confidenciales. Nadie sabr como contesto a las preguntas. NO ESCRIBA su nombre en encuesta, sus respuestas sern confidenciales secretas. Nadie va a saber como usted ha contestado las preguntas. NO ESCRIBA su nombre en sta forma, sus respuestas se mantendrn secretas. Nadie va a saber como usted ha contestado las preguntas. DIRECTIONS: Read each question. Carefully check the one answer that fits best. Some of the questions use the phrase "having sex." This means sexual intercourse. INSTRUCCIONES Lea cada pregunta. Con cuidado marque la contestacin que sea apropiada. Algunas preguntas usan la frase teniendo sexo Estosignifica relaciones sexuales DIRECCIONES: Lea cada pregunta. Cuidadosamente marque la respuesta que le parezca adecuada. En algunas de las preguntas se us la frase teniendo sexo esto significa contacto sexual. INSTRUCCIONES Lea cada pregunta. Con cuidado marque la contestacin que sea apropiada. Algunas preguntas usan la frase teniendo sexo. Esto significa relaciones sexuales. 1. What is your gender? A. Male B. Female Cal es tu gnero? A. masculino B. Femenino Cal es tu sexo? A. masculino B. Femenino 1.Cul es su gnero? A. Hombre B. Mujer 2. What grade are you in? A. 9th B. 10th C. 11th D. 12th En que grado ests? A. 9 B.10 C. 11 D. 12 .En que grado ests? A. 9 B.10 C. 11 D. 12 2. En que grado ests? A. 9 B.10 C. 11 D. 12 3. What is your age? A. 13-14 B. 15-16. C.17-19 Que edad tienes? A. 13-14 B.15-16 C. 17-19 Qu edad tienes? A. 13-14 B. 15-16 C. 17-19 3. Qu edad tienes? A. 13-14 B. 15-16 C. 17-19 4. Have you received HIV education in school? A. Yes B. No Has recibido educacin sobre El SIDA en la escuela? A. Si B. No Has tu recibido educacin sobre del SIDA en la escuela? A. Si B. No 4. Has recibido educacin sobre El SIDA en la escuela? A. S B. No 1. You can't get AIDS if you have sex only once or twice without a condom. Tu no puedes contraer el SIDA si tu tienes sexo una o dos veces sin un condn Tu no puedes contraer SIDA si tienes sexo una vez o dos veces solamente sin usar condones. 5. Tu no puedes contraer SIDA si tienes sexo una vez o dos veces solamente sin usar condones. 206

PAGE 218

Appendix L. (continued) 2. A person can "pass" an HIV antibody test (test negative) but still be infected with HIV. Una persona puede pasar una prueba de anticuerpos del SIDA (resultados negativos) y aun estar infectado por el SIDA. Una persona puede pasar un exmen de anticuerpos de SIDA (exmen es negativo) pero puede estar infectado con SIDA. Una persona puede pasar un examen de anticuerpos de VIH (examen es negativo) pero puede estar infectado con VIH. 3. Condoms are 100% effective in preventing HIV. Los condones son 100% efectivos en prevenir VIH Los condones son 100% efectivos en prevenir el SIDA. Los condones son 100% efectivos en prevenir VIH. 4. Males can pass HIV on to others through their semen. Los hombres pueden pasar VIH a travs de su semen Los hombres pueden pasar el SIDA a travs de su semen. Los hombres pueden pasar VIH a travs de su semen. 5. You can get HIV by sitting on the seat of a toilet that a person with AIDS has used. Tu puedes adquirir VIH al sentarte en un inodoro que fu usado por una persona con VIH Tu puedes adquirir VIH al sentarte en un inodoro que fue usado por una persona con VIH. Tu puedes adquirir VIH al sentarte en un inodoro que fu usado por una persona con VIH. 6. Abstinence from sex and drugs is the best way for teen-agers to avoid getting HIV. Abstinencia del sexo y de las drogas es la mejor forma para que los adolescents eviten contraer el VIH Abstinencia de sexo y drogas es la major forma para que los jvenes eviten el SIDA. Abstinencia de sexo y drogas es la mejor forma para que los jvenes eviten el VIH. 7. You can get HIV from drinking from the same glass or water fountain that a person with AIDS drank from. Puedes adquirir VIH en los tomaderos de agua o tomando agua de un vaso usado por una persona con VIH Puedes adquirir VIH en los tomaderos de agua o tomando agua de un vaso usado por una persona con VIH. Puedes adquirir VIH en los tomaderos de agua o tomando agua de un vaso usado por una persona con VIH. 8. HIV can be found in semen, vaginal fluids, and blood. VIH puede estar presente en el semen, fluidos vaginales, y sangre VIH puede estar presente en el semen, fluidos vaginales, y sangre. VIH puede estar presente en el semen, fluidos vaginales, y sangre. 9. A person can get HIV by sharing drug needles. Puede una persona adquirir VIH al compartir agujas para la droga Puede una persona adquirir VIH al compartir agujas para la droga. Puede una persona adquirir VIH al compartir agujas para la droga. 10. HIV can be found in breast milk. VIH puede estar presente en la leche maternal VIH puede estar presente en la leche maternal. VIH puede estar presente en la leche maternal. 207

PAGE 219

208 Appendix L. (continued) 11. Once you are infected with HIV, you are infected for life. Una vez que tu has sido infectado con VIH, tu ests infectado de por vida Contagiado vs infectado Una vez que tu has sido infectado con VIH, tu ests infectado de por vida. Una vez que tu has sido infectado con VIH, tu ests infectado de por vida. 12. People infected with HIV are usually very thin and sickly. Personas infectadas con VIH son bien delgadas y se notan enfermizas Personas infectadas con VIH son bien delgadas y parecen enfermizas. Personas infectadas con VIH son bien delgadas y se notan enfermizas. 13. Some people have gotten HIV by swimming in the same pool as someone with AIDS. Algunas personas han contraido el VIH al nadar en la misma piscina que una persona que lo tiene Algunas personas han adquirido VIH nadando en la misma piscina con una persona con SIDA. Algunas personas han adquirido VIH nadando en la misma piscina con una persona con SIDA. 14. You can get HIV from a mosquito bite. Puedes t adquirir VIH de una picada demosquito Puedes t adquirir VIH de una picada de mosquito. Puedes t adquirir VIH de una picada de mosquito. 15. If you want to keep from getting HIV, using "lambskin" condoms is just as good as using latex condoms. Si tu quieres protejerte de contraerVIH, usando condones de piel de cordero es tan efectivo como usar condones de latexSi tu quieres protejerte de contraer VIH, usando condones de piel de cordero es tan efectivo como usar condones de latex. Si tu quieres protegerte de contraer VIH, usando condones de piel de cordero es tan efectivo como usar condones de ltex. 16. A new vaccine has been developed for the treatment of AIDS. Una nueva vacuna ha sido desarrollada para tratar el SIDA Sabes si hay alguna vacuna que ha sido descubierta para curar el VIH? Una nueva vacuna ha sido desarrollada para tratar el SIDA 17. If a pregnant woman has AIDS,there is a chance it may harm her unborn baby Si una mujer preada tiene SIDA, existe la posibilidad que el feto sufra daos? Si una mujer preada tiene SIDA, existe la posibilidad que el feto sufra daos? Si una mujer preada tiene SIDA, existe la posibilidad que el feto sufra daos? 1. If your friends want you to do something that you think might not be safe, you should at least try it. Si un amigo quiere que tu hagas algo y t crees que no es seguro, t deberas por lo menos tratar de hacerlo Si un amigo quiere que tu hagas algo y t crees que no es seguro, t deberas por lo menos tratar de hacerlo. Si sus amigos desean que Ud. haga algo que Ud. no considera seguro, Ud. por lo menos debe de tratar de hacerlo 2. It's okay not to have sex while you are a teen-ager. Est correcto no tener sexo mientras sea un(a) adolescente Est bien no tener sexo cuando t eres an un adolescente. Est correcto no tener sexo mientras sea un(a) adolescente.

PAGE 220

Appendix L. (continued) 3. I believe condoms should always be used if a person my age has sex, even if the two people know each other very well.*52.Yo creo que los condones debern siempre ser usados si una persona de mi edad tiene sexo, aunque las dos personas se conozcan muy bien. NA Yo creo que los condones debern siempre ser usados si una persona de mi edad tiene sexo, aunque las dos personas se conozcan muy bien. 4. A teen-ager can inject drugs once in a while without a risk of getting infected with HIV. Un joven puede inyectarse drogas de vez en cuando sin riesgo de contraer VIH Un joven puede inyectarse drogas de vez en cuando sin riesgo de contraer VIH. Un joven puede inyectarse drogas de vez en cuando sin riesgo de contraer VIH. 5. Teen-agers are at risk of getting infected with HIV if they engage in sex without a condom. Los adolescents tienen el riesgo de ser infectados con el virus del SIDA si tienen relaciones sexuales sin usar un condn. Los jovenes tienen riesgo de infectarse con VIH si ellos practican sexo sin usar condones. Los jovenes tienen riesgo de infectarse con VIH si ellos practican sexo sin usar condones. 6. To keep your friends, you should go along with most things your friends want you to do. Para mantener sus amistades, Ud. debe hacer las mayora de las cosas que sus amigos quieren que Ud. haga Para no perder tus amigos, t tienes que complacerlos con todas las cosas que ellos quieren que t hagas. Para no perder tus amigos, t tienes que complacerlos con todas las cosas que ellos quieren que t hagas. 7. People who don't have sex before they get married are strange. Las personas que no tienen sexo antes casarse son extraos. Las personas que no tienen sexo antes de casarse son raras. Las personas que no tienen sexo antes de casarse son extraas. 8. It is not smart to have sex without using a condom. N o es muy inteligente tener sexo sin usar condones No es muy inteligente tener sexo sin usar condones. No es inteligente tener sexo sin usar un condn 9. Using needles to inject steroids or drugs is a bad idea. Usar agujas para inyectarse esteroides o drogas es una mala idea Usar agujas para inyectarse esteroides es una mala idea. Usar agujas para inyectarse esteroides no es una mala idea. 10. It's okay to have sex without a condom because your chance of getting infected with HIV is very low. Esta correcto tener sexo sin usar un condn porque es muy bajo el chance de ser infectado con el SIDA Esta bien tener sexo sin usar condones ya que tus chances de contraer VIH son muy bajos. Esta bien tener sexo sin usar condones ya que tus chances de contraer VIH son muy bajos. 11. Teen-agers should learn how to resist pressures from their friends. Los adolescents deben saber resistir las presiones de sus amigos Los jvenes tienen que aprender a resistir las presiones de sus amigos Los adolescentes deben saber resistir las presiones de sus amigos 209

PAGE 221

Appendix L. (continued) 12. It's a good idea for teen-agers not to have sex. Es una buena idea para los adolescentes no tener sexo Es una buena idea que los jvenes no tengan sexo. Es una buena idea para los adolescentes no tener sexo. 13. People who share drug needles shouldn't worry because they probably won't get infected with HIV. Las personas que comparten agujas para drogas no deben preocuparse porque ellos probablemente no se infectaran de VIH Las personas que comparten agujas para drogas no deben preocuparse porque ellos probablemente no se infectaran de VIH. Las personas que comparten agujas para drogas no deben preocuparse porque ellos probablemente no se infectaran de VIH. 14. Teen-agers should realize that if they're not careful, they could get infected with HIV. Los adolescentes tienen que saber que si no son cuidadosos, ellos pueden ser infectados con VIH. (Realize problematic definition.) Los jvenes tienen que saber que si no son cuidadosos, ellos pueden ser infectados con VIH. Los jvenes tienen que saber que si no son cuidadosos, ellos pueden ser infectados con VIH. 15. When friends want you to do things you don't feel like doing, there's no harm in going along. Cuando tus amigos quieren que t hagas cosas que t sientes no deben hacerse, no hay ningn riesgo si t las haces.( Harm in going along is a USA saying?) Cuando tus amigos quieren que t hagas cosas indebidas, no correras riesgo si las haces. Cuando los amigos quieren que haga cosas que Ud. no se siente con deseos de hacer, no hay peligro en seguir la corriente 16. Using a condom doesn't make sex less pleasurable. Es menos el placer del sexo si se usan los condones Usar un condon no te impidira entir placer.. Es menos el placer del sexo si se usan los condones. 17. Anyone who shares needles is taking a chance of getting infected with HIV. Cualquiera que comparte agujas se est tomando el riesgo de ser infectado con VIH Cualquiera que comparte agujas se est tomando el riesgo de ser infectado con VIH. Cualquiera que comparte agujas se est tomando el riesgo de ser infectado con VIH. 18. If teen-agers are careful about choosing sexual partners, they won't get infected with HIV. Si los adolescentes son cuidadosos al escoger su pareja para sexo, ellos no sern infectados con VIH. Si los jvenes son cuidadosos al escoger su pareja para sexo, ellos no sern infectados con VIH. Si los jvenes son cuidadosos al escoger su pareja para sexo, ellos no sern infectados con VIH. 19. Teen-agers should be more willing to resist pressures from their friends. Los jvenes deberan resistir las presionesde sus amigos. Los jvenes deberan resistir las presiones de sus amigos. Los adolescentes deben tener ms buena voluntad para resistir las presiones de sus amigos. 20. These days it makes a lot of sense to wait to have sex until you get married. Los jvenes deberan resistir las presiones de sus amigos. En sta poca tiene mas sentido esperar a tener relciones exuales hast casarse. En estos tiempos es ms sensato esperarse a tener sexo hasta que uno se case 210

PAGE 222

Appendix L. (continued) 21. If people think they might have sex during a date, they should carry a condom. Si una persona piensa que podria tener sexo al salir con alguien, ellos deberan traer un condn. (What is the word for Date in Spanish) Si una persona piensa que tal vez tenga sexo al salir con alguien, ellos deberan traer un condn. Si las personas consideran que pueden tener sexo durante una salida, es mejor llevar consigo un condn 23. People who use condoms during sex don't trust the person they're with. Las personas que usan condones durante el sexo, no confan en la persona con quien estan. La persona que usa condones cuando tiene sexo no confa en su pareja. Las personas que usan condones durante el sexo, no confan en la persona con quien estn. 24. People who share drug needles should clean the needles with bleach. Las personas que comparten agujas para drogas debern limpiar stas con leja Las personas que comparten agujas para drogas debern limpiar stas con blanqueador. Las personas que comparten agujas para drogas debern limpiar stas con leja. 25. HIV is something that teen-agers should think about when they date. El SIDA es un asunto que los adolescentes deben de pensar cuando salen con alguien. SIDA es algo que los jvenes deben tener en cuenta cuando empiezan a socializar. VIH es algo en que los jvenes debern meditar cuando van a tener una cita. 26. I believe its OK for people my age to have sex with a steady boyfriend or girlfriend.*Yo creo que est bien que las personas de mi edad tengan sexo con el novio o novia. Yo creo que est bien que las personas de mi edad tengan sexo con el novio o novia. Yo creo que est bien que las personas de mi edad tengan sexo con el novio o novia. 27. I believe people my age should wait until they are older before they have sex.*Yo creo que las personas de mi edad deberan esperar un poco ms de tiempo para tener sexo. .* Yo creo que las personas de mi edad deberan esperar un poco ms de tiempo para tener sexo. Creo que las personas de mi edad deben esperar hasta que tengan mas edad antes de tener sexo. 28. I believe condoms should always be used if a person my age has sex.*Yo creo que siempre se debe de usar condones cuando una persona de mi edad tiene sexo Yo creo que siempre se debe de usar condones cuando una persona de mi edad tiene sexo. Creo que los condones siempre deben ser usados por una persona de mi edad para tener sexo. 211

PAGE 223

Appendix L. (continued) 29. I believe condoms should always be used if a person my age has sex, even if the girl uses birth control pills.*. Yo creo que los condones debern siempre ser usados si una persona de mi edad tiene sexo, aunque la jven use pastillas para no salir embarazada. *NA Yo creo que los condones debern siempre ser usados si una persona de mi edad tiene sexo, aunque la jven use pastillas para no salir embarazada I am sure it's true. Estoy seguro que es verdad Yo estoy seguro de que es verdad Estoy seguro que es verdad I don't know No se Yo no s No se I think it's false. Creo que es falso Yo creo que esto es falso Creo que es falso I am sure it's false Estoy seguro que es falso Yo estoy seguro de que es falso Estoy seguro que es falso DIRECTIONS: This survey asks you to say whether you agree or disagree with a set of statements. Please read each statement, then indicate whether you: Insrtuciones: En sta encuesta se te pide que digas cuando estas de acuerdo y cuando no estas de acuerdo con lo que se dice aqu. Por favor lea cada pregunta, entonces indica si usted esta: DIRECCIONES: En sta encuesta se te pide que respondas i estas de acuerdo o desacuerdo con la informacin ofrecida. Este cuestionario le hace preguntas para saber si Ud. est de acuerdo o no con ciertas aseveraciones. Por favor lea cada una y conteste si Ud. Strongly Agree (SA), Agree (A), are Not Sure (NS), Disagree (D), or Strongly Disagree (SD), by circling the answer you want. Completamente de Acuerdo (CA), De Acuerdo (DA), No estas Seguro (NES), No estoy De Acuerdo (NEDA), Completamente en Desacuerdo (CED), Haga un crculo en la respuesta elegida. Completamente de Acuerdo (CA), De Acuerdo (DA), No estas Seguro (NES), No estoy De Acuerdo (NEDA), Completamente en Desacuerdo (CED), hacer un crculo en la respuesta elegida. Completamente de Acuerdo (CA), De Acuerdo (DA), No estas Seguro (NES), No estoy De Acuerdo (NEDA), Completamente en Desacuerdo (CED), hacer un crculo en la respuesta elegida. Strongly Agree (SA) Completamente de Acuerdo (CA) Completamente de Acuerdo (CA) Completamente de Acuerdo (CA) Agree (A) De Acuerdo (DA De Acuerdo (DA) De Acuerdo (DA) Are Not Sure (NS) No Estoy Seguro (NES) No Estoy Seguro (NES) No Estoy Seguro (NES) Disagree (D) No Estoy de Acuerdo (NEDA) No Estoy de Acuerdo (NEDA) No Estoy de Acuerdo (NEDA) Or Strongly Disagree (SD) Completamente en Desacuerdo (CED) Completamente en Desacuerdo (CED) Completamente en Desacuerdo (CED) Note. Items from Basen-Engquist et al. (1999). Form adapted from Beaton, D., Bombardier, C., Guillemin, F., & Ferraz, M.B. (2002). 212

PAGE 224

Appendix M. Results of the Back-translation. Original English English Back-translation Spanish items for the first version DO NOT put your name on this survey. Your answers will be kept secret. No one will know how you answered these questions. Do not write your name on this survey. Your answers will be kept completely confidential. No one will now how you answered the questions. Note: contest No escriba su nombre en esta encuesta. Sus respuestas sern guardadas completamente confidenciales. Nadie sabr como contesto a las preguntas DIRECTIONS: Read each question. Carefully check the one answer that fits best. Some of the questions use the phrase "having sex." This means sexual intercourse. Instructions: Read each question. Take care to mark the appropriate answer. Some questions use the phrase having sex. This means sexual intercourse. Note: teniendo sexo would tener sexo be more correct? INSTRUCCIONES: Lea cada pregunta. Con cuidado marque la contestacin que sea apropiada. Algunas preguntas usan la frase teniendo sexo. Esto significa relaciones sexuales 1.What is your gender? A. Male B. Female What is your gender? Male B. Female Note: Doesnt genero/masculine/femenino refer just to grammar? So says my dictionary. Would not Cal es to sexo? varn/hembra or hombre/mujer be more correct? Femenino Cal es tu gnero? A. masculino B. 2. What grade are you in? A. 9th B. 10th C. 11th D. 12th What grade are you in? A. 9th. B. 10th C. 11th D. 12th En que grado ests? A. 9 B.10 C. 11 D. 12 3. What is your age? A. 13-14 B. 15-16. C. 17-19 How old are you? A. 13-14 B. 15-16 C. 17-19 Que edad tienes? A. 13-14 B.15-16 C. 17-19 4. Have you received HIV education in school? A. Yes B. No Have you received AIDS education at your school? A. Yes B. No Has recibido educacin sobre El SIDA en la escuela? A. Si B. No 213

PAGE 225

Appendix M. (continued) English Back translation, Spanish items for 1st version 1. You can't get AIDS if you have sex only once or twice without a condom. You cannot contract AIDS if you have sex one or two times without a condom. Note:Are these true or false questions? I think new instructions are needed on how the respondent is to answer questions 6. Tu no puedes contraer el SIDA si tu tienes sexo una o dos veces sin un condn 2. A person can "pass" an HIV antibody test (test negative) but still be infected with HIV. A person can pass an AIDS antibodies test (negative result) and still be infected with AIDS Una persona puede pasar una prueba de anticuerpos del SIDA (resultados negativos) y aun estar infectado por el SIDA 3. Condoms are 100% effective in preventing HIV. Condoms are 100% effective in the prevention of HIV Los condones son 100% efectivos en prevenir VIH 4. Males can pass HIV on to others through their semen. Men can pass on HIV through their semen Los hombres pueden pasar VIH a travs de su semen 5.You can get HIV by sitting on the seat of a toilet that a person with AIDS has used. You can acquire HIV by sitting down on a toilet that was used by a person with HIV. Tu puedes adquirir VIH al sentarte en un inodoro que fu usado por una persona con VIH 6. Abstinence from sex and drugs is the best way for teen-agers to avoid getting HIV. Abstinence from sex and drugs is the best way that teen-agers can avoid contracting HIV. Abstinencia del sexo y de las drogas es la mejor forma para que los adolescentes eviten contraer el VIH 7. You can get HIV from drinking from the same glass or water fountain that a person with AIDS drank from. You can acquire HIV from drinking fountains or by drinking water from a glass used by a person with HIV. Puedes adquirir VIH en los tomaderos de agua o tomando agua de un vaso usado por una persona con VIH 8. HIV can be found in semen, vaginal fluids, and blood. HIV is present in semen, vaginal fluids, and blood. VIH puede estar presente en el semen, fluidos vaginales, y sangre 9. A person can get HIV by sharing drug eedles. Can a person acquire HIV from sharing drug needles? Note: Or should the Spanish be: Una persona puede adquirir Puede una persona adquirir VIH al compartir agujas para la droga 214

PAGE 226

Appendix M. (continued) English Back translation, Spanish items for 1st version 10. HIV can be found in breast milk. HIV can be present in breast milk. VIH puede estar presente en la leche maternal 11. Once you are infected with HIV, you are infected for life. Once you are infected with HIV, you are infected for the rest of your life. Una vez que tu has sido infectado con VIH, tu ests infectado de por vida. 12. People infected with HIV are usually very thin and sickly. People infected with HIV are very thin and are notably very sick Personas infectadas con VIH son bien delgadas y se notan enfermizas 13. Some people have gotten HIV by swimming in the same pool as someone with AIDS. Some people have contracted HIV by swimming in the same swimming pool as someone else who has it. Algunas personas han contraido el VIH al nadar en la misma piscina que una persona que lo tiene 14. You can get HIV from a mosquito bite. Can you acquire HIV from a mosquito bite? Puedes t adquirir VIH de una picada de mosquito 15. If you want to keep from getting HIV, using "lambskin" condoms is just as good as using latex condoms. If you want to protect yourself from contracting HIV, using lamb skins condoms is just as effective as using latex condoms. Note: would el usar or el uso be more correct than the gerund usando? Si tu quieres protejerte de contraer VIH, usando condones de piel de cordero es tan efectivo como usar condones de latex 16. A new vaccine has been developed for the treatment of AIDS A new vaccine has been developed to treat AIDS. Una nueva vacuna ha sido desarrollada para tratar el SIDA 17. If a pregnant woman has AIDS,there is a chance it may harm her unborn baby If a pregnant woman has AIDS, is there a possibility her fetus could be hurt? Si una mujer preada tiene SIDA, existe la posibilidad que el feto sufra daos? 1. If your friends want you to do something that you think might not be safe, you should at least try it. If a friend wants you to do something and you think its not safe, you should at least try it. Si un amigo quiere que tu hagas algo y t crees que no es seguro, t deberas por lo menos tratar de hacerlo 215

PAGE 227

Appendix M. (continued) English Back translation, Spanish items for 1st version 2. It's okay not to have sex while you are a teen-ager. Its right not to have sex while you are a teen-ager. Est correcto no tener sexo mientras sea un(a) adolescente 3. I believe condoms should always be used if a person my age has sex, even if the two people know each other very well. I think that condoms should always be used if a person my age has sex, even if the two people know each other well. Creo que los condones debern siempre ser usados si una persona de mi edad tiene sexo, aunque las dos personas se conozcan muy bien 4. A teen-ager can inject drugs once in a while without a risk of getting infected with HIV. A young person can inject himself with drugs once in a while without running the risk of contracting HIV. Un joven puede inyectarse drogas de vez en cuando sin riesgo de contraer VIH 5. Teen-agers are at risk of getting infected with HIV if they engage in sex without a condom. Teen-agers run the risk of being infected with the AIDS virus if they have sex without using a condom. Los adolescents tienen el riesgo de ser infectados con el virus del SIDA si tienen relaciones sexuales sin usar un condn. 6. To keep your friends, you should go along with most things your friends want you to do.In order to have friendships, you should do most of the things that your friends want you to do. Para mantener sus amistades, Usted. debe hacer las mayora de las cosas que sus amigos quieren que usted haga 7. People who don't have sex before they get married are strange. People who do not have sex before getting married are strange. Las personas que no tienen sexo antes de casarse son extraos. 8. It is not smart to have sex without using a condom. Its not very smart to have sex without using condoms. No es muy inteligente tener sexo sin usar condones 9. Using needles to inject steroids or drugs is a bad idea. Using needles to inject steroids or drugs is a bad idea. Usar agujas para inyectarse esteroides o drogas es una mala idea 10. It's okay to have sex without a condom because your chance of getting infected with HIV is very low. Its right to have sex without using a condom because the chances of becoming infected with AIDS is very low. Esta correcto tener sexo sin usar un condn porque es muy bajo el chance de ser infectado con el SIDA 216

PAGE 228

Appendix M. (continued) English Back translation, Spanish items for 1st version 11. Teen-agers should learn how to resist pressures from their friends. Teen-agers should know how to resist peer pressure. Los adolescents deben saber resistir las presiones de sus amigos 12. It's a good idea for teen-agers not to have sex. Its a good idea for teen-agers to have sex. (Forgot word NOT) Es una buena idea para los adolescentes no tener sexo 13. People who share drug needles shouldn't worry because they probably won't get infected with HIV. People who share drug needles shouldnt worry because they probably wont be infected with HIV. Las personas que comparten agujas para drogas no deben preocuparse porque ellos probablemente no se infectaran de VIH 14. Teen-agers should realize that if they're not careful, they could get infected with HIV.Teen-agers should know that if they arent careful, they can become infected with HIV. Los adolescentes tienen que saber que si no son cuidadosos, ellos pueden ser infectados con VIH. 15. When friends want you to do things you don't feel like doing, there's no harm in going along. When your friends want you to do something and you dont feel you should do it, theres no risk in not doing it. Cuando tus amigos quieren que t hagas cosas que t sientes no deben hacerse, no hay ningn riesgo si t las haces. 16. Using a condom doesn't make sex less pleasurable. Sex is less pleasurable if you use condoms. Personal Note: Well, of course its less pleasurable, Carlos, but thats no reason for teen-agers not to use one. Is this a trick question? Es menos el placer del sexo si se usan los condones 17. Anyone who shares needles is taking a chance of getting infected with HIV. Anyone who shares needles is taking the risk of being infected Sex is less pleasurable if you use condoms. Cualquiera que comparte agujas se est tomando el riesgo de ser infectado con VIH 18. If teen-agers are careful about choosing sexual partners, they won't get infected with HIV. If teen-agers arent careful in choosing their sex partner, they will be infected with HIV. Si los adolescentes son cuidadosos al escoger su pareja para sexo, ellos no sern infectados con VIH. 19. Teen-agers should be more willing to resist pressures from their friends. Young people should resist peer pressure.Note: debern Los jvenes deberan resistir las presiones de sus amigos. 217

PAGE 229

Appendix M. (continued) English Back translation, Spanish items for 1st version 20. These days it makes a lot of sense to wait to have sex until you get married. NA NA 21. If people think they might have sex during a date, they should carry a condom. If someone thinks that they are going to have sex with someone, they should bring a condom. Si una persona piensa que podria tener sexo al salir con alguien, ellos deberan traer un condn. 22. I believe it's OK for people my age to have sex with several different people in the same month. I think it's okay for people my age to have sex with different people in the same month. Creo que est bien que las personas de mi edad tengan sexo con diferentes personas en el mismo mes. 23. People who use condoms during sex don't trust the person they're with. People who use condoms during sex dont trust the person that they are with. Las personas que usan condones durante el sexo, no confian en la persona con quien estan 24. People who share drug needles should clean the needles with bleach. People who share drug needles should clean them with bleach. Note: debern or is deben de better? Las personas que comparten agujas para drogas deberan limpiar stas con leja 25. HIV is something that teen-agers should think about when they date. AIDS is an issue that teen-agers should think about when they date someone. El SIDA es un asunto que los adolescents deben de pensar cuando salen con alguien. 26. I believe its OK for people my age to have sex with a steady boyfriend or girlfriend.I think its better for people my age to have sex with their boyfriend or girlfriend. Yo creo que est bien que las personas de mi edad tengan sexo con el novio o novia. 27. I believe people my age should wait until they are older before they have sex. I think people my age should wait a little while longer before having sex. Note: debern or is deben de better? Yo creo que las personas de mi edad deberan esperar un poco ms de tiempo para tener sexo. 28. I believe condoms should always be used if a person my age has sex. I think that condoms should always be used when someone my age is having sexYo creo que siempre se debe de usar condones cuando una persona de mi edad tiene sexo 218

PAGE 230

Appendix M. (continued) English Back translation, Spanish items for 1st version 29. I believe condoms should always be used if a person my age has sex, even if the girl uses birth control pills. I think that condoms should always be used when a person my age has sex, even if the young woman uses the birth control pill. Creo que los condones debern siempre ser usados si una persona de mi edad tiene sexo, aunque la joven use pastillas anticonceptivas. I am sure it's true. Im sure its true. Estoy seguro que es verdad I think it's true. I think its true. Creo que es verdad I don't know. I dont know Note: No s No se I think it's false. I think its false. Creo que es falso I am sure it's false Im sure its false. Estoy seguro que es falso Strongly Agree (SA) Completely Agree (CA) Completamente de Acuerdo (CA) Agree (A), Agree (A) De Acuerdo (DA) Not Sure (NS). Not Sure (NS) No Estoy Seguro (NES) Disagree (D) Disagree (D) No Estoy de Acuerdo (NEDA) Strongly Disagree (SD) Completely Disagree (CD Completamente en Desacuerdo (CED) Note: Form adapted from Beaton, D., Bombardier, C., Guillemin, F., & Ferraz, M.B. (2002). 219

PAGE 231

Appendix N. First Spanish Version Cuestionario de Conocimientos y Actitudes de HIV/SIDA INSTRUCCIONES: Lea cada pregunta. Con cuidado marque la respuesta que le parezca adecuada. Algunas preguntas usan la frase teniendo sexo. Esto significa relaciones sexuales. 1.Cul es tu gnero? a. Masculino____ b. Femenino____ 2.En que ao de bachillerato ests? a. 1____ b.2 ____ c. 3 ____ 3.Qu edad tienes? a. 13-14____ b. 15-16____ c.17-19____ 4.Has recibido educacin sobre El VIH/SIDA? a. Si____ b. No____ 5. En donde has recibido dicha informacin? a. Escuela_____ b. Charlas____ c. FUNDASIDA_____ d. Otras__________ Marque la repuesta que le parezca adecuada (x): Estoy seguro que es verdad Creo que es verdad No se Creo que es falso Estoy seguro que es falso 1. Tu no puedes contraer el SIDA si tu tienes sexo una o dos veces sin un condn. 2. Una persona puede pasar una p rueba de anticuerpos de VIH (resultados negativos) y aun estar infectado con VIH. 3. Los condones son 100% efectivos en prevenir el VIH. 4. Los hombres pueden pasar el VIH a travs de su semen. 5. T puedes adquirir VIH al sentarte en un inodoro que fue usado por una persona con VIH. 6. Abstinencia de sexo y de las drogas es la mejor forma para que los adolescentes eviten contraer el VIH. 7. Puedes adquirir VIH en los tomaderos de agua o tomando agua de un vaso usado por una persona con VIH. 8. El VIH puede estar presente en el semen, fluidos vaginales, y sangre. 9. Una persona puede adquirir VIH al compartir agujas para la droga. 10. VIH puede estar presente en la leche maternal. 220

PAGE 232

Appendix N. (continued) 11. Una vez que tu has sido infectado con VIH, tu ests infectado de por vida. 12. Las personas infectadas con VIH son bien delgadas y se notan enfermizas. 13. Algunas personas han contrado el VIH al nadar en la misma piscina que una persona que tiene SIDA. 14. T puedes adquirir VIH de una picada de mosquito. 15. Si tu quieres protegerte de contraer VIH, usando condones de piel de cordero es tan efectivo como usar condones de ltex. 221

PAGE 233

Appendix O. First Spanish Version INSTRUCCIONES: En sta encuesta se te pide que digas cuando estas de acuerdo y cuando no estas de acuerdo con lo que se dice aqu. Por favor lea cada pregunta, entonces indica si t ests: Completamente de Acuerdo, De Acuerdo, No Estoy Seguro, No Estoy de Acuerdo, Completamente en Desacuerdo Marque la respuesta que le parezca adecuada con una x. Completamente de Acuerdo De Acuerdo No Esto y Seguro No Esto y de Acuerdo Completa -mente en Desacuer do 1. Si tus amigos quieren que tu hagas algo y t crees que no es seguro, t deberas por lo menos tratar de hacerlo. 2. Est bien no tener sexo mientras sea un(a) adolescente. 3. Creo que los condones debern siempre ser usados si una persona de mi edad tiene sexo, aunque las dos personas se conozcan muy bien. 4. Un adolescente puede inyectarse drogas de vez en cuando sin riesgo de contraer VIH. 5. Los adolescentes tienen el riesgo de ser infectados con VIH si tienen relaciones sexuales sin usar un condn. 6. Para mantener tus amistades, t debes hacer la mayora de las cosas que tus amigos quieren que t hagas. 7. Las personas que no tienen sexo antes de casarse son extraos. 8. No es muy inteligente tener sexo sin usar un condn. 9. Usar agujas para inyectarse esteroides o drogas es una mala idea. 10. Esta bien tener sexo sin usar un condn porque es muy bajo el chance de ser infectado con VIH. 11. Los adolescentes tienen que aprender a resistir las presiones de sus amigos. 12. Es una buena idea para los adolescentes no tener sexo. 13. Las personas que comparten agujas para drogas no deben preocuparse porque ellos probablemente no se infectaran de VIH. 222

PAGE 234

Appendix O. (continued) 14. Los adolescentes tienen que saber que si no son cuidadosos, ellos pueden ser infectados con VIH. 15. Cuando tus amigos quieren que t hagas cosas que t sientes no deben hacerse, no hay ningn riesgo si t las haces. 16. Es menos el placer del sexo si se usan un condn. 17. Cualquiera que comparte agujas se est tomando el riesgo de ser infectado con VIH. 18. Si los adolescentes son cuidadosos al escoger su pareja para sexo, ellos no sern infectados con VIH. 19. Los adolescentes deben tener ms buena voluntad para resistir las presiones de sus amigos. 20. En estos tiempos es mas sensato esperarse a tener sexo hasta que uno se case 21. Si las personas piensan que podran tener sexo al salir con alguien, ellos debern traer un condn 22. Creo que est bien que las personas de mi edad tengan sexo con diferentes personas en el mismo mes. 23. Las personas que usan condones durante el sexo, no confan en la persona con quien estn. 24. Las personas que comparten agujas para drogas debern limpiarlas con leja. 25. El VIH es un tema en que los adolescentes deben de pensar cuando salen con alguien. 26. Creo que est bien que las personas de mi edad tengan sexo con el novio o novia. 27. Creo que las personas de mi edad debern esperar un poco ms de tiempo para tener sexo. 28. Ceo que siempre se debe de usar condones cuando una persona de mi edad tiene sexo. 29. Creo que los condones debern siempre ser usados si una persona de mi edad tiene sexo, aunque la joven use pastillas anticonceptivas. Muchas gracias por tu tiempo. 223

PAGE 235

Appendix P. Summary of Results from the Conceptual Equivalence. A Salvadorian panel composed of five members rated each item on a pass (1) or fail (0) scale. A mean score was calculated and a passing score of 3 or higher indicated conceptual equivalence. Knowledge Semantic Idiomatic Conceptual Comment or Suggested change 1 You can't get AIDS if you have sex only once or twice without a condom. Tu no puedes contraer el SIDA si tu tienes sexo una o dos veces sin un condn. Pass =4 Fail = Pass =4 Fail = Pass =5 Fail = 2 A person can "pass" an HIV antibody test (test negative) but still be infected with HIV. Una persona puede pasar una prueba de anticuerpos de VIH (resultados negativos) y aun estar infectado con VIH. Pass =4 Fail = Pass =4 Fail = Pass=5 Fail = 3 Condoms are 100% effective in preventing HIV. Los condones son 100% efectivos en prevenir el VIH. Pass = 4 Fail = Pass =4 Fail = Pass =5 Fail = 4 Males can pass HIV on to others through their semen. Los hombres pueden pasar VIH a travs de su semen. Pass =4 Fail = Pass =4 Fail = Pass =5 Fail = 5 You can get HIV by sitting on the seat of a toilet that a person with AIDS has used. T puedes adquirir VIH al sentarte en un inodoro que fue usado por una persona con VIH. Pass =4 Fail = Pass =4 Fail = Pass = 5 Fail = 6 Abstinence from sex and drugs is the best way for teen-agers to avoid getting HIV. Abstinencia de sexo y de las drogas es la mejor forma para que los adolescentes eviten contraer el VIH. Pass = 4 Fail = Pass =4 Fail = Pass =5 Fail = 224

PAGE 236

Appendix P. (continued) Knowledge Semantic Idiomatic Conceptual Comment or Suggested change 7 You can get HIV from drinking from the same glass or water fountain that a person with AIDS drank from. Puedes adquirir VIH en los tomaderos de agua o tomando agua de un vaso usado por una persona con SIDA. Pass =4 Fail = Pass =5 Fail = Pass = 5 Fail = Chorro o usando un vaso, instead of tomaderos. (Dr. Castro) 8 HIV can be found in semen, vaginal fluids, and blood. El VIH puede estar presente en el semen, fluidos vaginales, y sangre. Pass = 4 Fail = Pass =4 Fail = Pass = 5 Fail = 9 A person can get HIV by sharing drug needles. Una persona puede adquirir VIH al compartir agujas para la droga. Pass =4 Fail = Pass =4 Fail = Pass =5 Fail = 10 HIV can be found in breast milk. VIH puede estar presente en la leche maternal. Pass = 4 Fail = Pass =4 Fail = Pass =5 Fail = Leche maternal instead of leche maternal. (Dr. Castro) 11 Once you are infected with HIV, you are infected for life. Una vez que tu has sido infectado con VIH, tu ests infectado de por vida. Pass = 4 Fail = Pass =4 Fail = Pass =5 Fail = 12 People infected with HIV are usually very thin and sickly. Personas infectadas con VIH son bien delgadas y se notan enfermizas. Pass = 4 Fail = Pass =4 Fail = Pass = 5 Fail = Change to: Personas infectadas con VIH usualmente se ven bien delgadas y enfermizas. (Dr. Castro) 13 Some people have gotten HIV by swimming in the same pool as someone with AIDS. Algunas personas han contrado el VIH al nadar en la misma piscina que una persona que lo tiene. Pass = 4 Fail = Pass = 4 Fail = Pass = 5 Fail = Algunas personas han contrado el VIH al nadar en la piscina donde ha nadado una persona con SIDA. (Dr. Castro) 225

PAGE 237

Appendix P. (continued) Knowledge Semantic Idiomatic Conceptual Comment or Suggested change 14 You can get HIV from a mosquito bite. T puedes adquirir VIH de una picada de mosquito. Pass = 4 Fail = Pass = 5 Fail = Pass = 5 Fail = Zancudo instead of mosquito. (Dr. Castro) 15 If you want to keep from getting HIV, using "lambskin" condoms is just as good as using latex condoms. Si tu quieres protegerte de contraer VIH, usando condones de piel de cordero es tan efectivo como usar condones de latex. Pass = 4 Fail = Pass = 4 Fail = Pass = 5 Fail = Lambskin not used in El Salvador. Change to: condones de latex son los mejores. (Dr. Castro) 16 A new vaccine has been developed for the treatment of AIDS. Una nueva vacuna ha sido desarrollada para tratar el SIDA. Pass = 4 Fail = Pass = 4 Fail = Pass = 5 Fail = 17 If a pregnant woman has AIDS, there is a chance it may harm her unborn baby. Si una mujer embarazada tiene SIDA, existe la posibilidad que el feto sufra daos. Pass = 4 Fail = Pass = 4 Fail = Pass = 5 Fail = 1 If your friends want you to do something that you think might not be safe, you should at least try it. Si tus amigos quieren que tu hagas algo y t crees que no es seguro, t deberas por lo menos tratar de hacerlo. Pass = 4 Fail = Pass = 4 Fail = Pass = 4 Fail = Three persons suggested an alternative sentence structure. Add para complacerlos, tu deberias hacerlo (Rene Vasconcelos). 2 Its okay not to have sex while you are a teen-ager. Est bien no tener sexo mientras sea un(a) adolescente. Pass = 4 Fail = Pass = 4 Fail = Pass = 5 Fail = Add seas (Carlos S Zometa). 3 I believe condoms should always be used if a person my age has sex, even if the two people known each other very well. Creo que los condones debern siempre ser usados si una persona de mi edad tiene sexo, aunque las dos personas se conozcan muy bien. Pass = 4 Fail = Pass = 4 Fail = Pass = 5 Fail = 226

PAGE 238

Appendix P. (continued) 4 A teen-ager can inject drugs once in a while without a risk of getting infected with HIV. Un adolescente puede inyectarse drogas de vez en cuando sin riesgo de contraer VIH. Pass = 4 Fail = Pass = 4 Fail = Pass =5 Fail = 5 Teen-agers are at risk of getting infected with HIV if they engage in sex without a condom. Los adolescentes tienen el riesgo de ser infectados con VIH si tienen relaciones sexuales sin usar un condn. Pass = 4 Fail = Pass = 4 Fail = Pass = 5 Fail = 6 To keep your friends, you should go along with most things your friends want you to do. Para mantener tus amistades, t debes hacer la mayora de las cosas que tus amigos quieren que t hagas. Pass = 4 Fail = Pass = 4 Fail = Pass = 5 Fail = 7 People who don't have sex before they get married are strange. Las personas que no tienen sexo antes de casarse son extraos. Pass = 4 Fail = Pass= 4 Fail = Pass = 5 Fail = 8 It is not smart to have sex without using a condom. No es muy inteligente tener sexo sin usar un condn. Pass = 4 Fail = Pass = 4 Fail = Pass = 5 Fail = 9 Using needles to inject steroids or drugs is a bad idea. Usar agujas para inyectarse esteroides o drogas es una mala idea. Pass = 4 Fail = Pass =4 Fail = Pass = 4 Fail = Steroids not common, Tattoo or tatuajes is a relevant isuue. (Dr. Gavidia) 10 It's okay to have sex without a condom because your chance of getting infected with HIV is very low. Esta bien tener sexo sin usar un condn porque es muy bajo el chance de ser infectado con VIH. Pass =4 Fail = Pass =4 Fail = Pass = 5 Fail = 227

PAGE 239

Appendix P. (continued) Semantic Idiomatic Conceptual Suggested change or comments 11 Teen-agers should learn how to resist pressures from their friends. Los adolescentes tienen que aprender a resistir las presiones de sus amigos. Pass =4 Fail = Pass =4 Fail = Pass = 5 Fail = 12 It's a good idea for teen-agers not to have sex. Es una buena idea para los adolescentes no tener sexo. Pass =4 Fail = Pass =4 Fail = Pass =5 Fail = 13 People who share drug needles shouldn't worry because they probably won't get infected with HIV. Las personas que comparten agujas para drogas no deben preocuparse porque ellos probablemente no se infectaran de VIH. Pass = 4 Fail = Pass =4 Fail = Pass =5 Fail = 14 Teen-agers should realize that if they're not careful, they could get infected with HIV. Los adolescentes tienen que saber que si no son cuidadosos, ellos pueden ser infectados con VIH. Pass =4 Fail = Pass =4 Fail = Pass =5 Fail = 15 When friends want you to do things you don't feel like doing, there's no harm in going along. Cuando tus amigos quieren que t hagas cosas que t sientes no deben hacerse, no hay ningn riesgo si t las haces. Pass = 4 Fail = Pass =4 Fail = Pass = 4 Fail = 16 Using a condom doesn't make sex less pleasurable. Es menos el placer del sexo si se usan condones.Pass =3 Fail = Pass =3 Fail = Pass = 3 Fail = Re translate. Usar un condon no hace el sexo menos placentero.(Lic Armas and Dr. Gavidia) 17 Anyone who shares needles is taking a chance of getting infected with HIV. Cualquiera que comparte agujas se est tomando el riesgo de ser infectado con VIH. Pass =4 Fail = Pass = 4 Fail = Pass =5 Fail = 228

PAGE 240

Appendix P. (continued) Semantic Idiomatic Conceptual Suggested change or comments 18 If teen-agers are careful about choosing sexual partners, they won't get infected with HIV. Si los adolescentes son cuidadosos al escoger su pareja para sexo, ellos no sern infectados con VIH. Pass =4 Fail = Pass =4 Fail = Pass =5 Fail = 19 Teen-agers should be more willing to resist pressures from their friends. Los adolescentes deben tener ms buena voluntad para resistir las presiones de sus amigos. Pass = 4 Fail = Pass =4 Fail = Pass =5 Fail = 20 These days it makes a lot of sense to wait to have sex until you get married. En estos tiempos es mas sensato esperarse a tener sexo hasta que uno se case Pass = 4 Fail = Pass =4 Fail = Pass =5 Fail = 21 If people think they might have sex during a date, they should carry a condom. Si las personas piensan que podran tener sexo al salir con alguien, ellos debern traer un condn Pass = 4 Fail = Pass =4 Fail = Pass =5 Fail = Replace traer with llevar. (Dr. Castro) 22 I believe its OK for people my age to have sex with several different people in the same month. Creo que est bien que las personas de mi edad tengan sexo con diferentes personas en el mismo mes. Pass = 4 Fail = Pass =4 Fail = Pass =5 Fail = Add Yo creo (Rene Vasconcelos) 23 People who use condoms during sex don't trust the person they're with. Las personas que usan condones durante el sexo, no confan en la persona con quien estn. Pass =4 Fail = Pass =4 Fail = Pass =5 Fail = 24 People who share drug needles should clean the needles with bleach. Las personas que comparten agujas para drogas debern limpiarlas con leja. Pass = 4 Fail = Pass =4 Fail = Pass =5 Fail = 229

PAGE 241

Appendix P. (continued) Semantic Idiomatic Conceptual Suggested change or comments 25 HIV is something that teen-agers should think about when they date. El VIH es un tema que los adolescentes deben de pensar cuando salen con alguien. Pass = 4 Fail = Pass =4 Fail = Pass =5 Fail = 26 I believe its OK for people my age to have sex with a steady boyfriend or girlfriend. Creo que est bien que las personas de mi edad tengan sexo con el novio o novia. Pass =4 Fail = Pass =4 Fail = Pass =5 Fail = Add Yo creo (Rene Vasconcelos). Add con su novio o novia. (Dr. Castro). 27 I believe people my age should wait until they are older before they have sex. Creo que las personas de mi edad debern esperar un poco ms de tiempo para tener sexo. Pass =4 Fail = Pass =4 Fail = Pass = 5 Fail = 28 I believe condoms should always be used if a person my age has sex. Creo que siempre se debe de usar condones cuando una persona de mi edad tiene sexo. Pass = 4 Fail = Pass =4 Fail = Pass =5 Fail = 29 I believe condoms should always be used if a person my age has sex, even if the girl uses birth control pills. Creo que los condones debern siempre ser usados si una persona de mi edad tiene sexo, aunque la joven use pastillas anticonceptivas.Pass =4 Fail = Pass =4 Fail = Pass =5 Fail = 230

PAGE 242

Appendix Q. Individual Results for Cultural Acceptability and C ontent Validity of Knowledge and Attitudinal Items. Knowledge Items Panelist A Panelist B Panelist C Panelist D Panelist E Panelist F Panelist G Panelist H 1 Tu no puedes contraer el SIDA si tu tienes sexo una o dos veces sin un condn. Salvadorian Cultura 2 2 3 3 2 4 3 Knowledge Content 4 2 2 4 2 3 Comments Con o sin condon siempre hay un riesgo de contratar el VIH Eliminate double negatives. Add relaciones sexuales. Una persona p uede obten e un resultado negativo de l p rueba VIH, aun estar infectado? Tu puedes contraer el virus del VIH si tienessexo na o dos veces sin un condon Debe escribir como pregunta, es una afirmaci n. 2 Una persona puede pasar una prueba de anticuerpos de VIH (resultados negativos) y aun estar infectado con VIH. Salvadorian Cultura 3 2 4 4 3 3 4 Knowledge Content 4 4 4 3 4 3 Comments Se tendria que evaluar el conocimiento de anticuepos por los aolescentes Focus on results of the test. Prueba del VIH me resulta negativa 231

PAGE 243

Appendix Q. (continued) 3 Los condones son 100% efectivos en prevenir el VIH. Salvadorian Cultura 4 4 4 4 4 4 4 Knowledge Content 4 2 2 4 4 4 Comments Si se usan correcta y consistentemente Los condones usados enforma correcta y consistente reducen el riego de contraer el VIH 4 Los hombres pueden pasar VIH a travs de su semen. Salvadorian Cultura 4 4 4 4 4 4 4 Knowledge Content 4 4 4 3 4 4 Comments Add el in front of VIH. 5 T puedes adquirir VIH al sentarte en un inodoro que fue usado por una persona con VIH. Salvadorian Cultura 4 4 1 4 4 4 3 Knowledge Content 4 2 4 3 4 4 232

PAGE 244

Appendix Q. (continued) Comments Add el in front of VIH. 6 Abstinencia de sexo y de las drogas es la mejor forma para que los adolescentes eviten contraer el VIH. Salvadorian Cultura 4 2 4 4 4 4 4 Knowledge Content 4 4 4 4 4 4 Comments Add relaciones sexuales Abstinencia y fidelidad son las mejores formas para que los adolescentes eviten infectarse con el VIH. 7 Puedes adquirir VIH en el chorro o usando un vaso que ha sido usado por una persona con VIH/SIDA. Salvadorian Cultura 4 4 2 4 4 4 233

PAGE 245

Appendix Q. (Continued) Knowledge Content 4 1 4 4 4 Comments Add el VIH 8 El VIH puede estar presente en el semen, fluidos vaginales, y sangre. Salvadorian Cultura 4 4 4 4 4 4 3 Knowledge Content 4 4 4 4 4 3 Comments Add en la sangre. 9 Una persona puede adquirir VIH al compartir agujas para la droga. Salvadorian Cultura 3 4 4 4 4 4 Knowledge Content 4 4 4 3 4 3 Comments Una persona puede adquirir VIH al compartir agujas en el consumo de diversas drogas Add el VIH. 234

PAGE 246

Appendix Q. (Continued) 10 VIH puede estar presente en leche materna. Salvadorian Cultura 3 4 4 4 3 4 4 Knowledge Content 4 4 4 4 4 4 Comments Add el, and la Add el VIH. Add en la sangre. 11 Una vez que tu has sido infectado con VIH, tu ests infectado de por vida. Salvadorian Cultura 3 4 4 4 4 4 3 Knowledge Content 4 4 4 4 4 4 Comments Una vez. Tu eres portador del virus toda la vida(x estigma y descriminacion) Add el VIH. 12 Personas infectadas con VIH usualmente se ven bien delgadas y enfermizas. 235

PAGE 247

Appendix Q. (Continued) Salvadorian Cultura 4 4 4 4 2 4 3 Knowledge Content 4 2 4 2 4 3 Comments Las personas Add el VIH 13 Algunas personas han contrado el VIH al nadar en la misma piscina donde ha nadado una persona con SIDA. Salvadorian Cultura 4 4 4 4 3 4 4 Knowledge Content 4 1 4 4 4 3 Comments En la poblacin Salvadorea continua siendo un mito sobre VIH/SIDA Cambia nadado para evitar monotonia. 14 T puedes adquirir VIH de una picada de zancudo. Salvadorian Cultura 4 4 2 4 4 4 3 236

PAGE 248

Appendix Q. (continued) Knowledge Content 4 1 4 3 4 3 Comments Add el VIH 15 Si tu quieres protegerte de contraer VIH condones de latex son los mejores. Salvadorian Cultura 2 4 4 4 4 4 4 Knowledge Content 4 4 4 3 4 3 Comments Una forma de prevension es el uso de condones de latex Add el VIH. Add los condones. Los condones de latex evitan contraer el VIH durante una rlcion sexual. 16 Una nueva vacuna ha sido desarrollada para tratar el SIDA. Salvadorian Cultura 3 4 1 4 4 4 4 Knowledge Content 4 1 2 4 4 4 Comments Existe una vaccuna contra el VIH 237

PAGE 249

Appendix Q. (Continued) 17 Si una mujer embarazada tiene SIDA, existe la posibilidad que el feto sufra daos. Salvadorian Cultura 2 4 4 4 4 4 3 Knowledge Content 4 4 4 4 4 3 Comments Si una mujer(symbol) tiene vih/sida, existe la posibilidad de que su hijo nazca infectada Que sea transmitido a su hijo. Incluir no solo SIDA sino asi VIH/SIDA. No est apropiado = 1, Poco Apropiado = 2, Moderadamente apropiado = 3, Muy apropiado = 4. Attitudinal tems Panelist #1. #2 #3 #4 #5 #6 #7 #8. 1 Si tus amigos quieren que tu hagas algo y t crees que no es seguro, para complacerlos, t deberas de hacerlo. Salvadorian Cultura 4 1 4 2 4 2 Attitudinal Content 1 4 4 2 238

PAGE 250

Appendix Q. (Continued) Comments Confusing question. Se puede epresar mas directamente. More direct. Attitudinal tems Panelist #1. #2 #3 #4 #5 #6 #7 #8. 2 Est bien no tener sexo mientras seas un (a) adolescente. Salvadorian Cultura 4 4 2 4 2 4 4 Attitudinal content 4 4 3 4 3 4 4 Comments Change tener sexo for relaciones sexuales. 3 Creo que los condones debern siempre ser usados si una persona de mi edad tiene sexo, aunque las dos personas se conozcan muy bien. Salvadorian Cultura 4 4 4 4 4 4 3 Attitudinal content 4 4 4 2 4 3 239

PAGE 251

Appendix Q. (Continued) Attitudinal tems Panelist #1. #2 #3 #4 #5 #6 #7 #8. Comments Change tener sexo for relaciones sexuales Lo usan para evitar embarazo 4 Un adolescente puede inyectarse drogas de vez en cuando sin riesgo de contraer VIH. Salvadorian Cultura 4 4 1 3 4 4 Attitudinal content 1 2 3 4 4 Comments Un adolescente puede usar drogas de vez en cuando sin riesgo dcontraer VIH 240

PAGE 252

Appendix Q. (Continued) Attitudinal tems Panelist #1. #2 #3 #4 #5 #6 #7 #8. 5 Los adolescentes tienen el riesgo de ser infectados con VIH si tienen relaciones sexuales sin usar un condn. Salvadorian Cultura 2 4 4 4 4 4 4 Attitudinal content 4 4 4 3 4 3 Comments Hay riesgo con o sin condon 6 Para mantener tus amistades, t debes hacer la mayora de las cosas que tus amigos quieren que t hagas. Salvadorian Cultura 3 1 2 4 4 4 1 Attitudinal content 1 1 4 3 4 1 241

PAGE 253

Appendix Q. (Continued) Attitudinal tems Panelist #1. #2 #3 #4 #5 #6 #7 #8. Comments Replace: quieren que t hagas for digan. Creo que esta preguntase puedereformar, porque lo que quermos es que el o la adolescente fortalezcan y puedan decir no, como esta planteada es cierto. La redaccionpodria mejorarse.Rewrite. Suprim el 2 y 3 Tu. 7 Las personas que no tienen sexo antes de casarse son extraos. Salvadorian Cultura 4 2 1 4 4 3 3 Attitudinal content 2 4 3 4 3 242

PAGE 254

Appendix Q. (continued) Attitudinal tems Panelist #1. #2 #3 #4 #5 #6 #7 #8. Comments Add relaciones sexuales. Creo que las tienes pro no por eso. 8 No es muy inteligente tener sexo sin usar un condn. Salvadorian Cultura 4 1 4 4 4 4 4 Attitudinal content 2 4 4 3 4 3 Comments Add con una trabajadora sexual o prostituta. 9 Usar agujas para inyectarse esteroides o drogas es una mala idea. Salvadorian Cultura 4 3 4 2 3 4 3 Attitudinal content 4 2 3 4 4 243

PAGE 255

Appendix Q. (Continued) Attitudinal tems Panelist #1. #2 #3 #4 #5 #6 #7 #8. Comments Por que 10 Esta bien tener sexo sin usar un condn porque es muy bajo el chance de ser infectado con VIH. Salvadorian Cultura 4 4 1 4 4 4 3 Attitudinal content 4 2 3 3 3 Comments Add relaciones sexuales Cambia chance por p orcentaje u otro sinnimo. 11 Los adolescentes tienen que aprender a resistir las presiones de sus amigos. Salvadorian Cultura 4 4 4 4 3 4 4 Attitudinal content 4 4 4 4 4 Comments Los y las adolescentes 244

PAGE 256

Appendix Q. (continued) Attitudinal tems Panelist #1 #2 #3 #4 #5 #6 #7 #8 12 Es una buena idea para los adolescentes no tener sexo. Salvadorian Cultura 4 4 1 4 3 4 3 Attitudinal content 1 3 4 4 3 Comments Add posponer el inicio de las relaciones sexuales. 13 Las personas que comparten agujas para drogas no deben preocuparse porque ellos probablemente no se infectaran de VIH. Salvadorian Cultura 4 4 4 2 3 4 4 Attitudinal content 4 1 4 4 4 4 245

PAGE 257

Appendix Q. (Continued) Attitudinal tems Panelist #1 #2 #3 #4 #5 #6 #7 #8 Comments Con el uso de las drogas se pierden capacidad de auto cuidado y l otras [? ] se creen no vulnerables. Podria expresarse mas directamente. Be more direct. 14 Los adolescentes tienen que saber que si no son cuidadosos, ellos pueden ser infectados con VIH. Salvadorian Cultura 4 2 4 4 3 4 4 Attitudinal content 3 4 4 2 4 3 .0 Comments Cuidadoso es muy genrico. 15 Cuando tus amigos quieren que t hagas cosas que t sientes no deben hacerse, no hay ningn riesgo si t las haces. 246

PAGE 258

Appendix Q. (continued) Attitudinal tems Panelist #1 #2 #3 #4 #5 #6 #7 #8 Salvadorian Cultura 4 3 1 3 1 4 4 Attitudinal content 3 1 4 3 4 3 Comments No compensable. Suprima tu 2nd,3ro y 4. 16 Usar condones no hace el sexo meno placentero. Salvadorian Cultura 4 4 2 4 3 4 3 Attitudinal content 4 4 3 4 3 Comments Add s to menos, relaciones sexuales. Menos 17 Cualquiera que comparte agujas se est tomando el riesgo de ser infectado con VIH. Salvadorian Cultura 4 4 4 2 3 4 4 Attitudinal content 4 4 3 3 4 4 247

PAGE 259

Appendix Q. (continued) Attitudinal tems Panelist #1. #2 #3 #4 #5 #6 #7 #8. Comments Delete se. 18 Si los adolescentes son cuidadosos al escoger su pareja para sexo, ellos no sern infectados con VIH. Salvadorian Cultura 4 4 2 4 4 4 3 Attitudinal content 4 1 4 3 3 3 Comments Sexo for relaciones sexuales. 19 Los adolescentes deben tener ms buena voluntad para resistir las presiones de sus amigos. Salvadorian Cultura 4 4 4 4 3 3 4 Attitudinal content 4 4 4 4 4 3 248

PAGE 260

Appendix Q. (continued) Attitudinal tems Panelist #1. #2 #3 #4 #5 #6 #7 #8 Comments Change to Los adolescentes deben tener mas fuerza de voluntad. Mas buena voluntad por una gran voluntad. 20 En estos tiempos es mas sensato esperarse a tener sexo hasta que uno se case Salvadorian Cultura 4 4 3 4 3 4 4 Attitudinal content 4 4 4 3 4 3 Comments Relaciones sexuales. 21 Si las personas piensan que podran tener sexo al salir con alguien, ellos debern llevar un condn Salvadorian Cultura 4 4 3 4 3 4 2 Attitudinal content 4 4 4 3 4 2 249

PAGE 261

Appendix Q. (continued) Attitudinal tems Panelist #1. #2 #3 #4 #5 #6 #7 #8. Comments Relaciones sexuales 22 Creo que est bien que las personas de mi edad tengan sexo con diferentes personas en el mismo mes. Salvadorian Cultura 4 4 2 4 3 4 4 Attitudinal content 4 1 4 3 4 Comments Relaciones sexuales 23 Las personas que usan condones durante el sexo, no confan en la persona con quien estn. Salvadorian Cultura 4 4 4 4 4 4 3 Attitudinal content 4 3 4 4 3 Comments Relaciones sexuales 250

PAGE 262

Appendix Q. (continued) Attitudinal tems Panelist #1. #2 #3 #4 #5 #6 #7 #8. 24 Las personas que comparten agujas para drogas debern limpiarlas con leja. Salvadorian Cultura 4 2 2 2 2 4 2 Attitudinal content 3 4 3 4 2 Comments No creo que sea buena idea, ni aun mencionarlo. Sugiero reformular 25 El VIH es un tema que los adolescentes deben de pensar cuando salen con alguien. Expand. Si salen con intenciones de tener relaciones sexuales o se exponen a Salvadorian Cultura 4 3 4 4 4 3 Attitudinal content 4 4 4 3 251

PAGE 263

Appendix Q. (continued) Attitudinal tems Panelist #1. #2 #3 #4 #5 #6 #7 #8. Comments 26 Yo creo que est bien que las personas de mi edad tengan sexo con su novio o novia. Salvadorian Cultura 4 4 3 4 4 4 4 Attitudinal content 4 1 4 3 4 Comments Relaciones sexuales 27 Yo creo que las personas de mi edad debern esperar un poco ms de tiempo para tener sexo. Salvadorian Cultura 4 4 3 4 4 4 4 Attitudinal content 4 4 4 2 4 4 Comments Relaciones sexuales 252

PAGE 264

Appendix Q. (continued) Attitudinal tems Panelist #1. #2 #3 #4 #5 #6 #7 #8. 28 Yo creo que siempre se debe de usar condones cuando una persona de mi edad tiene sexo. Salvadorian Culture 1 4 3 4 3 4 3 Attitudinal content 4 4 4 3 4 4 Comments Confunde pregunta que el condon es 100% effectivo. Relaciones sexuales 29 Creo que los condones debern siempre ser usados si una persona de mi edad tiene sexo, aunque la joven use pastillas anticonceptivas. Salvadorian Culture 1 2 4 3 4 3 Attitudinal content 4 4 3 4 3 Comments Confunde pregunta que el condon es 100% effectivo Relaciones sexuales Aunque la joven use anticonceptivos. 253

PAGE 265

Appendix R. Final Spanish Instrument INSTRUCCIONES: No pongas tu nombre en la encuesta, es annima y nadie sabr que respondiste. Lea las afirmaciones y marque la respuesta que le parezca adecuada con una X. 1. Cul es tu gnero? a. Masculino____ b. Femenino____ 2. En que ao de bachillerato ests? a. 1____ b.2 ____ c. 3 ____ 3. Qu edad tienes? a. 13-14____ b. 15-16____ c.17-mas____ 4. Has recibido educacin sobre VIH/SIDA? a. Si____ b. No ____ 5. En que ao recibi esa educacin? _______________. 6. En donde has recibido dicha informacin? Si fue en ms de un lugar, indica con varias X s. a. Escuela_____ b. Charlas____ c. FUNDASIDA_____ d. Otras__________ Marque la repuesta que le parezca adecuada (x): Estoy seguro que es verdad Creo que es verdad No se Creo que es falso Esto y se g uro que es falso 1. Tu puedes contraer el VIH/SIDA si tu tienes relaciones sexuales una o dos veces sin un condn. 2. Una persona puede pasar una prueba de sangre para el VIH/SIDA (resultados negativos) y aun estar infectado con VIH. 3. Los condones son 100% efectivos en prevenir el VIH/SIDA. 4. Los hombres pueden pasar el VIH/SIDA a travs de su semen. 5. T puedes adquirir VIH/SIDA al sentarte en un inodoro que fue usado por una persona con VIH/SIDA. 6. Abstinencia de relaciones sexuales y de las drogas es la mejor forma para que los adolescentes eviten contraer el VIH/SIDA. 7. Puedes adquirir VIH/SIDA en el chorro o usando un vaso que ha sido usado por una persona con VIH/SIDA. 8. El VIH/SIDA puede estar presente en el semen, fluidos vaginales y sangre. 9. Una persona puede adquirir VIH/SIDA al compartir agujas en el consumo de diversas drogas. 10. El VIH/SIDA puede estar presente en la leche materna. 254

PAGE 266

Appendix R. (continued) Marque la repuesta que le parezca adecuada (x): Estoy seguro que es verdad Creo que es verdad No se Creo que es falso Esto y se g uro que es falso 11. Una vez que has sido infectado con VIH/SIDA, tu eres portador del virus toda tu vida. 12. Personas infectadas con el VIH/SIDA usualmente se ven bien delgadas y enfermizas. 13. Algunas personas han contrado el VIH/SIDA al nadar en la misma piscina donde han permanecido personas con SIDA. 14. T puedes adquirir VIH/SIDA de una picada de zancudo. 15. Una forma de prevencin del VIH/SIDA es el uso de condones de latex. INSTRUCCIONES: En sta encuesta se te pide que digas cuando estas de acuerdo y cuando no estas de acuerdo con lo que se dice aqu. Por favor lee cada pregunta, entonces indica con un X si t ests: Completamente de Acuerdo, De Acuerdo, No Estoy Seguro, No Estoy de Acuerdo, Completamente en Desacuerdo Marque la respuesta que le parezca adecuada con una x. Completamente de Acuerdo De Acuerdo No Esto y Seguro No Estoy de Acuerdo Completamente en Desacuerdo 1. Si tus amigos quieren que hagas algo incorrecto y t crees que no es seguro, para complacerlos, deberas por lo menos tratar de hacerlo. 2. Est bien no tener relaciones sexuales mientras seas un(a) adolescente. 3. Yo creo que los condones debern siempre ser usados si una persona de mi edad tiene relaciones sexuales, aunque las dos personas se conozcan muy bien. 4. Un adolescente puede inyectarse drogas de vez en cuando sin riesgo de contraer VIH/SIDA. 5. Los adolescentes corren el riesgo de ser infectados con VIH/SIDA si tienen relaciones sexuales sin usar un condn. 255

PAGE 267

Appendix R. (Continued) Marque la respuesta que le parezca adecuada con una x. Completamente de Acuerdo De Acuerdo No Estoy Seguro No Estoy de Acuerdo Completamente en Desacuerdo 6. Para mantener tus amistades, debes hacer la mayora de las cosas que tus amigos quieren que hagas. 7. Las personas que no tienen relaciones sexuales antes de casarse son extraos. 8. No es muy inteligente tener relaciones sexuales sin usar un condn. 9. Usar agujas para inyectarse drogas o hacerse tatuajes es una mala idea. 10. Esta bien tener relaciones sexuales sin usar un condn porque el riesgo de ser infectado con VIH/SIDA es muy bajo. 11. Los adolescentes tienen que aprender a resistir las presiones de sus amigos. 12. Es una buena idea para los adolescentes no tener relaciones sexuales. 13. Las personas que comparten agujas para drogas no deben preocuparse porque ellos probablemente no se infectaran con el VIH/SIDA. 14. Los adolescentes tienen que saber que si no son cuidadosos, ellos pueden ser infectados con VIH/SIDA. 15. Cuando tus amigos quieren que hagas cosas que sientes que no deben hacerse, no hay ningn riesgo si las haces. 16. Usar condones no hace la relacin sexual menos placentera. 17. Cualquiera que comparte agujas se est tomando el riesgo de ser infectado con el VIH/SIDA. 18. Si los adolescentes son cuidadosos al escoger su pareja para relaciones sexuales, ellos no sern infectados con el VIH/SIDA. 19. Los adolescentes deben tener ms fuerza de voluntad para resistir las presiones de sus amigos. 20. En estos tiempos es mas sensato esperarse a tener relaciones sexuales hasta que uno se case. 256

PAGE 268

Appendix R. (continued) 21. Si las personas piensan que podran tener relaciones sexuales al salir con su pareja, ellos debern llevar un condn. Marque la respuesta que le parezca adecuada con una x. Completamente de Acuerdo De Acuerdo No Estoy Seguro No Estoy de Acuerdo Completamente en Desacuerdo 22. Yo creo que est bien que las personas de mi edad tengan relaciones sexuales con diferentes personas en el mismo mes. 23. Las personas que usan condones durante las relaciones sexuales, no confan en la persona con quien estn. 24. Las personas que comparten agujas para drogas debern limpiarlas con leja. 25. El VIH/SIDA es un tema que los adolescentes deben de pensar cuando salen con alguien. 26. Yo creo que est bien que las personas de mi edad tengan relaciones sexuales con su novio o novia. 27. Yo creo que las personas de mi edad debern esperar un poco ms de tiempo para tener relaciones sexuales. 28. Yo creo que siempre se debe de usar condones cuando una persona de mi edad tiene relaciones sexuales. 29. Yo creo que los condones debern siempre ser usados si una persona de mi edad tiene relaciones sexuales, aunque la joven use pastillas anticonceptivas. Muchas gracias por tu tiempo. 257

PAGE 269

About the Author Carlos S Zometa attended Santa Fe Community College, received a Bachelors in Science in microbiology from the University of Florida and then obtained a Masters of Science in tropical public heal th and infectious diseases from the College of Public Health at the University of South Florida. His professional experience includes two years of laboratory based research in immunology/parasitology, six years as a science teacher at an inner city middle school in Tampa, and one year at a high school in Tampa, Florida.


xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001498168
003 fts
006 m||||e|||d||||||||
007 cr mnu|||uuuuu
008 041209s2004 flua sbm s000|0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0000507
035
(OCoLC)57714944
9
AJU6771
b SE
SFE0000507
040
FHM
c FHM
090
LB1555 (ONLINE)
1 100
Zometa, Carlos Salvador.
0 245
Translation, adaptation, and validation of an instrument to evaluate HIV/AIDS knowledge and attitudes for use with Salvadorian high school students
h [electronic resource] /
by Carlos Salvador Zometa, III.
260
[Tampa, Fla.] :
University of South Florida,
2004.
502
Thesis (Ph.D.)--University of South Florida, 2004.
504
Includes bibliographical references.
500
Includes vita.
516
Text (Electronic thesis) in PDF format.
538
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
Title from PDF of title page.
Document formatted into pages; contains 269 pages.
520
ABSTRACT: This study translated, cross-culturally adapted and validated an instrument's scores for use in public high schools in San Salvador, El Salvador. The original instrument consisted of items developed by the Centers for Disease Control and Prevention (CDC) to assess HIV/AIDS knowledge and five dimensions of attitudes (Abstinence, Peer-pressure, Condom use, Drug use, and Threat of HIV) in grades 7 to 12 in the United States. Items were translated into Spanish using the back-translation method. The instrument was cross-culturally adapted using guidelines proposed by Guillemin, Bombardier, and Beaton (1993). A cross-culturally equivalent version of the original instrument was obtained using three different Salvadorian review panels and two pretests with Salvadorian high school students. An expert panel of HIV Salvadorian professionals validated the content and established its cultural acceptability for public school use.A total of 483 students from 30 randomly selected public high schools in El Salvador participated in a series of validation studies. Confirmatory factor analysis of the translated instrument was used to evaluate the factorial validity of the five-factor attitudinal model. As part of the validation process, the translated Abstinence and Condom use subscales from the CDC were correlated with similar translated subscales from Basen-Engquist et al.'s (1999) published study as a measure of concurrent validity. Finally, internal consistency reliability (Cronbach's alpha) was determined with 483 students and test-retest reliability was obtained with a subsample of 39 students. Six major conclusions were: (1) The methodology used was successful in cross-culturally adapting the instrument. (2) HIV/AIDS content was rated as culturally acceptable and valid for use in public high schools of El Salvador.(3) The reliability of the scores from the knowledge section was moderate (test-retest reliability coefficient = .49 and coefficient alpha = .57). (4) Reliability (coefficient alpha) of the five attitudinal subscales was inconsistent: .55 (Peer-pressure), .58 (Abstinence), 0 (Condom use), .24 (Drugs), and .30 (Threat of HIV). (5) Confirmatory factor analysis provided support for a 4-factor attitudinal model (Peer-pressure, Abstinence, Drug use, and Threat of HIV). (6) Concurrent validity of the translated CDC Abstinence subscale was strong. The results provided support for the methodology to cross-culturally adapt an instrument. The psychometric properties from the knowledge section and the attitudinal component related to abstinence were acceptable but additional research is needed before the Spanish instrument can be used with confidence in El Salvador.
590
Adviser: Dedrick, Robert.
653
confirmatory factor analysis.
El Salvador.
validation.
cross-cultural adaptation.
HIV/AIDS knowledge and attitudes.
690
Dissertations, Academic
z USF
x Interdisciplinary Education
Doctoral.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.507