USF Libraries
USF Digital Collections

An explanatory examination of relationships between measures of school and student socioeconomic status and reading and ...

MISSING IMAGE

Material Information

Title:
An explanatory examination of relationships between measures of school and student socioeconomic status and reading and math achievement of Hispanic limited English proficient (LEP) high school students
Physical Description:
Book
Language:
English
Creator:
Piedra, Osvaldo
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla
Publication Date:

Subjects

Subjects / Keywords:
SES
Minority education
High stakes testing
Middleclass education concept
Academic achievement
Dissertations, Academic -- Educational Leadership -- Doctoral -- USF   ( lcsh )
Genre:
non-fiction   ( marcgt )

Notes

Summary:
ABSTRACT: Hispanics are the largest and fastest growing minority population in the U.S. requiring limited English proficiency services (Guglielmi, 2008). Many schools and school districts place heavy emphasis on high stakes testing. The No Child Left Behind (NCLB) Act of 2001 requires all children to have access to a fair, equal, and high-quality education. Many schools and school districts focus on individual students to show grade level performance for small populations such as limited English proficient (LEP) students. The purpose of this study was to examine the question, "What was the relationship between measures of school and student socioeconomic status, and reading and math achievement of Hispanic LEP high school students?" Guiding this study was the conceptual framework, the middleclass education concept.This concept suggested that underprivileged students attending an impoverished school had lower academic success than did underprivileged students attending a middleclass school with middleclass values. This study was developed from secondary data analysis of archived data from the Pinellas County School Board database. Pearson correlation coefficients were computed between the variables in this study. The correlation analysis was followed by a multiple regression analysis to estimate the capacity of the subsidized meal program, student absences, gender, grade point average (GPA), and the Norm Referenced Test (NRT) reading and math to explain reading and math achievement, as defined by the Florida Comprehensive Assessment Test in reading and math.The results of this study provided insight that the role played by indices of school and student SES may be different for Hispanic LEP students than for other minority students, such as African American and Hispanic or Latino high school students. The results showed that measures of SES were not as strong as expected based on the current research literature. GPA and NRT provided the strongest explanation of variability for FCAT reading and math scores. Other indicators such as student absences and gender were not statistically significant for explaining the FCAT reading and math scores. Thereby, indicating that the conceptual framework, the Middleclass Education Concept, may not apply to Hispanic LEP high school students and thus, further research for this population is needed.
Thesis:
Dissertation (Ed.D.)--University of South Florida, 2009.
Bibliography:
Includes bibliographical references.
System Details:
Mode of access: World Wide Web.
System Details:
System requirements: World Wide Web browser and PDF reader.
Statement of Responsibility:
by Osvaldo Piedra.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 140 pages.
General Note:
Includes vita.

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 - 002068373
oclc - 606875661
usfldc doi - E14-SFE0003138
usfldc handle - e14.3138
System ID:
SFS0027454:00001


This item is only available as the following downloads:


Full Text

PAGE 1

An Explanatory Examination of Relationships between Measures of School and Student Socioeconomic Status and Reading and Math Achievement of Hispanic Limited English Proficient (LEP) High School Students by Osvaldo Piedra A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Education Department of Educational Leadership and Policy Studies College of Education University of South Florida Major Professor: Darlene Bruner, Ed.D. William Blank, Ph.D. Bobbie Greenlee, Ed.D. Kristine Hogarty, Ph.D. Date of Approval: November 9 2009 Keywords: SES, Minority Education, High Stakes Testing, Middleclass Education Concept Academic Achievement Copyright 2009, Osvaldo Piedra

PAGE 2

Dedication This dissertation was dedicated to my dear friends, Bill Bowen and Charlie Coffee, who provided constant encouragement throughout my pursuit of this degree. To my parents, who in pursuit of a better life for themselves and hopes of a future for their children spent their first nights with their children in a refugee camp in this country in pursuit of an opportunity for a healthy, productive life for my brothers and me. To my colleague and dear frien d, Brenda Troy, this has been a sometimes funny, sometimes sad, and sometimes exciting adventure. May the adventures of Lewis and Clark, and Sacajawea continue To my seasonal suitcase friends, that like my favorite sweaters, I fold and place into a suitca se and tuck underneath my bed when spring brings in warmer weather, only to take out again in the fall. I dedicate this study to my dear friends that I have not seen during this season of my life. I look forward to starting a new season with you and new fr iends to come. To my four legged companions, Taffy and Littlebit, who provided around the clock support and unconditional companionship.

PAGE 3

Acknowledgements I owe a world of gratitude to my major professor and role model Dr. Darlene Bruner, for her patience, wisdom, and guidance in fulfilling my Thank you Dr. Bruner for helping me see through the clutter and for providing me with the clarity of vision that enabled me to accomplish my goal of helping students in need. I grateful ly acknowledge the fortitude, wisdom, and insistence on excellence of my dissertation committee. Thank you for your guidance throughout this process. To Dr. William Blank for walking into the middle of a study and providing the needed guidance without whi ch I could not complete my study. To Dr. Bobbie Greenlee for helping me stay focused and search for what was important for students to learn. Lastly, I owe special appreciation to Dr. Kristine Hogarty for helping me evolve and mature as a quantitative researcher. I also wish to acknowledge Dr. Nancy Zambito for providing leadership, wisdom, and encouragement in the pursuit of this study. To Dr. Coleen Sams for helping me stay focused as I made my way through the university jungle. To my parents, Alberto and Julia Piedra, I would like to say gracias por su amor, apoyo, y la oportunidad de tener no slo xito, sino tambin tener la oportunidad de una vida en este pas, g racias por todos sus sacrificios.

PAGE 4

i Table of Contents List of Tables iv ABSTRACT v Chapter 1 Introduction 1 Overview of Chapter 1 1 Literature Review 4 Research Problem and Significance 7 Nature of the Study 9 Research Question for Reading 9 Research Hypotheses for Reading 10 Research Question for Math 11 Research Hypotheses for Math 11 Methods 12 Purpose of the Study 12 Conceptual Framework 15 Data Collection 16 Definition of Key Terms 17 Delimitations of the Study 19 Summary 20 Overview of Upcoming Chapters 20 Chapter 2 Review of the Literature 22 Overview of Chapter 2 22 Introduction 23 Middleclass Education Concept 23 Strategies for Searching Literature 25 The LEP Population 26 Definition of Minority Status Used for the Purpose of this Dissertation 31 Socioeconomic Status (SES) 34 Defining SES 34 SES and Academic Achievement for Hispanic LEP S tudent 35 SES, High Stakes Testing, and NCLB for Hispanic LEP Students 41 SES Hispanic LEP Students, and the School Community 45 S ES, Hispanic Students, and LEP Programs 47 SES and Ethnicity as a Variable of Academic Achievement 49 SES, Hispanic LEP Students, and Educational Attainment 53 SES Characteristics of Disproportionately Hispanic LEP Schools 56 SES and Acculturation for Hispanic LEP Students 61 SES, Hispanic LEP Students, and Contradictions in the Research 65 Gender 66

PAGE 5

ii Conclusion 67 Chapter 3 Research Methods 70 Overview of Chapter 3 70 Introduction 70 Research Perspective 71 Research Context 74 I nstruments Employed in Data Collection 76 The Population 78 Dependent and Independent Variables 81 Research Question s and Hypotheses 85 Research Question for Reading 85 Research Hypotheses for Reading 85 Research Question for Math 86 Research Hypotheses for Mat h 86 Multiple Regression Formula for FCAT Reading Scores 88 Multiple Regression Formula for FCAT Math Scores 89 Correlations, Multiple Regressions, Regression Diagnostics 89 Leverage 92 Influential Data 93 Research Design 94 Data Analysis 95 Summary 96 Chapter 4 Research Results 97 P urpose of the Study 97 Design of the Study and Sample Criteria 98 Research Ques tion 98 Descriptive Statistics 99 Statistical Assumptions 100 Normality 101 Linearity 101 Variables are Measured Without Error 101 Homoscedasticity of Errors Homogeneity of Variance 102 Independence 102 The Independent Variables are Fixed 103 Influential Obser vations 103 Leverage 104 Summary of Analysis of Potential Outliers and Influence Diagnostics 104 Correlation between FCAT Reading and the Dependent Variables 106 Correlation between FCAT Math, the Dependent and Independent Variables 107 FCAT Reading Regression Analysis 108 Multiple Regression Formula for FCAT Reading Scores 109 FCAT Math Regression Analysis 109

PAGE 6

iii FCAT Reading and Math Q uestion 111 Reading Results 111 Math Results 111 Hypothesis 1 Reading and Math 111 Reading Results 111 Math Results 112 Hypothesis 2 Reading and Math 112 Reading Results 112 Math Results 112 Hypothesis 3 Reading and Math 112 Reading Results 113 Math Results 113 Hypothesis 4 Reading and Math 113 Reading Results 113 Math Results 113 Hypothesis 5 Reading and Math 113 Reading Results 113 Math Results 114 Summary 114 Chapter 5 Summary and Discussion 116 Overview 116 Discussion 118 Discussion and Findings 119 SES and Student Absences 120 GPA and NRT 121 Gender 122 Relevance of the Findings 123 Implications of the Findings 124 Limitations of the Study 125 Recommendations for Further Research 126 Conclusion 129 References 130 About the Author End Page

PAGE 7

iv List of Tables Table 1 High School Distribution by City 73 Table 2 Composite School SES = Black + Hispanic Enrollment in the subsidized meal program 76 Table 3 Percent of Hispanic LEP Sophomore Distribution 80 Table 4 Percent Hispanic LEP Sophomore Participation in the subsidized meal program 81 Table 5 Compatible FCAT Passing Scores for Reading and Math 8 4 Table 6 Table of Descriptive statistics 100 Table 7 Correlation Matrix: FCAT R eading and M ath and Measures of School and Student SES 10 6 Table 8 Summary Regression Analysis for Variables Explaining FCAT Reading Scores 10 8 Table 9 Summary Regression Analysis for Variables Explaining FCAT Math Scores 1 09

PAGE 8

v An Explanatory Examination of Relationships between Measures of School and Student Socioeconomic Status and Reading and Math Achievement of Hispanic Limited English Proficient (LEP) High School Students Osvaldo Piedra ABSTRA CT Hispanics are the largest and fastest growing minority population in the U.S. requiring limited English proficiency services (Guglielmi, 2008). Many schools and school districts place heavy emphasis on high stakes testing. The No Child Left Behind (NCLB) Act of 2001 re quires all children to have access to a fair, equal, and high quality education. Many schools and school districts focus on individual students to show grade level performance for small populations such as limited English proficient (LEP) students. The pur relationship between measures of school and student socioeconomic status, and reading Guiding this study was the conceptual framework the middleclass education concept. This concept suggested that underprivileged students attending an impoverished school had lower academic success than did underprivileged students attending a middleclass school with middleclass values.

PAGE 9

vi This study was developed from secondary data analysis of archived data from the Pinellas County School Board database. Pearson correlation coefficients were computed between the variables in this study. The correlation analysis was followed by a multiple regression analy sis to estimate the capacity of the subsidized meal program, student absences, gender, grade point average (GPA), and the Norm Referenced Test (NRT) reading and math to explain reading and math achievement, as defined by the Florida Comprehensive Assessmen t Test in reading and math. The results of this study provided insight that the role played by indices of school and student SES may be different for Hispanic LEP students t han for other minority students, such as African American and Hispanic or Latino high school students. The results showed that measures of SES were not as strong as expected based on the current research literature. GPA and NRT provided the strongest expla nation of variability for FCAT reading and math scores. Other indicators such as student absences and gender were not statistically significant for explaining the FCAT reading and math scores. Thereby, indicating that the conceptual framework, the Middlecl ass Education Concept, may not apply to Hispanic LEP high school students and thus, further research for this population is needed.

PAGE 10

1 Chapter 1 Introduction Overview of Chapter 1 This chapter introduced the research problem and the significance of the study. This was followed by a brief summary of the literature, the purpose of the study, the conceptual framework, the nature of the study, the research questions and hypotheses, the data collection method, def inition of key terms used, the limitations, and delimitations of the study, followed by a summary. Chapter 1 concluded with an overview of the upcoming chapters. As the U.S. population of foreign born residents from Spanish speaking countries increases, sc hools and school districts will need to develop educational strategies to meet the needs of the increasing Hispanic limited English proficient (LEP) population. The purpose of this study was to examine the relationship between school socioeconomic status ( SES ) participation in the subsidized meal program, student absences, gender, grade point average ( GPA ) Florida Comprehensive Assessment Test ( FCAT ) and Norm Reference Test ( NRT ) reading and math, and the achievement of Hispanic LEP students. This study sought to determine at what level of school SES did Hispanic LEP sophomores start to exhibit significantly poorer scores in reading and math achievement.

PAGE 11

2 The U.S. experienced an increase in foreign born residents of 1.4% between 2000 and 2006 (American Fa ctFinder, 2006). In 2006, the majority of the foreign residents, 12.5% of the U.S. population, spoke Spanish as their primary language (American FactFinder, 2006). Researchers such as Alexander and Salmon (2007), Andres, Adamuti Trache, Yoon, Pidgeon, and Thomsen (2007), and Arbona and Nora (2007) have noted that poverty disproportionately affects the minority population. Indeed, the 2003 American Community Survey (ACS), conducted by the U.S. Census Bureau, estimated the percentage of Hispanic families with children fewer than 18 years of age living in poverty was 46.3% for female householders with no spouse present. This was followed by African Americans at 42.6% and by European Americans at 32.2%, for families in the same circumstances (ACS, 2003). Accordi ng to research by Naglieri, Rojahn, and Matto (2006), Unger, Ritt Olson, Wagner, Soto, and Baezconde Garbanati (2007), and Weaver (2007), the fastest growing minority group in the U.S. was Hispanics with the majority being Spanish speaking immigrants from Mexico. The ACS (2007) confirmed the studies by Naglieri et al. and Weaver. The Hispanic population surpassed 45 million persons to comprise 15.1% of the total U.S. population, thus constituting the largest minority population. African Americans followed H ispanics at over 37.3 million persons, or 12.7% (ACS, 2007). The data were clear that Hispanics represented the fastest growing culturally and linguistically diverse population in the U.S. The ACS (2007) estimated that over 55 million persons, or 19.7% of the U.S. population, spoke a language other than English at home and of these, more than 24 million, 8.7% of the U.S. population, spoke English less

PAGE 12

3 than very well. Spanish, at over 34 million persons, or 8.7% of the U.S. population, was the most widely sp oken language at home other than English. Of the 8.7%, over 16 million, or 5.8%, spoke English less than very well. The remainder of the over 20 million persons, or 7.4% of speakers of languages other than English, included Indo European, Asian and Pacific Islander, and other language speakers. Of these, over 8 million or 2.9% spoke English less than very well. Thus, Spanish was the predominant language spoken at home after English, and Spanish speakers were the largest population to speak English less than very well (ACS, 2007). Using data from the U.S. Department of Education, Abedi, Hofstetter, and Lord (2004) estimated the majority (74%) of LEP students across the U.S. were Hispanic, followed by Vietnamese (2%). The need for an increase in the understand ing of the impact of SES on Hispanic LEP students was evident in the estimated 40% increase of the Hispanic K 12 LEP population by the year 2030 (Guglielmi, 2008). As the largest group, Hispanic LEP students were the most important, for practical purposes, to study. Hispanic LEP students differed from Hispanic students in that LEP students spoke a language other than English at home and non LEP students were either English speaking only or bilingual. The addition of Spanish spoken at home differentiated the Hispanic student from the Hispanic LEP student. The additional language spoken at home made the Hispanic LEP population a particularly interesting and unique population, worthy of further study. Therefore, the sample for this study was Hispanic LEP studen ts, a subgroup of the Hispanic population. It was impractical to study the entire Hispanic LEP population of the U.S. Although Florida had the fourth largest population of Hispanics, Pinellas County had a

PAGE 13

4 smaller proportion of Hispanics than Florida as a whole. Therefore, unlike many counties in the state, Hispanics were still a minority group. Although all Hispanic LEP students could have been selected, this study was limited to high school sophomores due to the testing requirements of the Florida Compreh ensive Assessment Test (FCAT) that required testing of all 10th grade students. Further, the literature indicated that Hispanic juniors and seniors had a higher dropout rate. Therefore, because of the mandatory age requirements that children must attend sc hool until the age of 16, at which time students ma y quit with parental permission juniors and seniors were excluded from this study, leaving sophomores as a logical population. The ACS (2007) estimated the Hispanic proportion of the population in vario us states. New Mexico at 44.1%, California at 35.7%, Texas at 35.5%, Nevada at 24.3%, and Florida at 20.1% were the states with the highest percentage of Hispanics of any race. According to the ACS (2007), there were 61,561 Hispanic persons residing in Pin ellas County, Florida. Of these residents, there were 2,232 Hispanic students enrolled in grades 9 12 in the 16 traditional high schools in Pinellas County. Of these students, 833 were enrolled as LEP students in grades 9 12. Literature Review The resear ch by Warren and Jenkins (2005) substantiated the need to examine the relationship between SES indices and academic achievement in states with high stakes testing such as Florida. In a related study, Warren and Edwards (2005) revealed that in states with h igh stakes testing, such as the Florida Comprehensive Assessment Test

PAGE 14

5 (FCAT), minority students were at a greater risk of dropping out of school than in states without a high stakes test. Both studies referenced SES as one of the factors affecting the acad emic attainment of Hispanic students. Guglielmi (2008) reported that minority and linguistically isolated populations were disproportionately represented in impoverished communities. Similarly, Attewell and Domina (2008) concluded in their research that minority children were disproportionately represented in the welfare system in California. The research by Attewell and Domina, and Guglielmi concluded that underprivileged children lived in communities that poorly served the needs of children. In the st udies by Arbona and Nora (2007), Burris, Heurbert, and Levin (2006), and Martin, Karabel, and Jaquez (2005), Hispanics were identified as the least proportionately represented population in educational attainment. Burris et al. (2006) concluded that the ed ucation gap between Hispanics and non Hispanics was evident by college admission rates. Burris and colleagues concluded low SES factors such as lower quality high school course offerings available to Hispanic students and the influence of peers in the clas sroom contributed to lower educational attainment. Their research supported the research by Arbona and Nora that cited Hispanics as having disproportionately the lowest college attainment level. Martin at al. (2005) attributed the affects of poverty such a s the quality of education to disproportionately low college application rates. The literature did not provide for a uniform definition of SES. Rather, SES was most often described as a construct definition using multiple variables. Specifically,

PAGE 15

6 Burris et al. (2006) and Martin at al. (2005) included participation in the subsidized meal program as part of the construct definition for student and school SES. Chen (2008) and Kennedy, Green, and Huerta (2007) included ethnicity as part of the composite defin ition for SES. Lastly, Lewis and Cheng (2006) included both ethnicity and participation in the subsidized meal program as part of their composite definition for SES. The research by Nichols (2003) and Reyes and Capper (1991) on education and poverty maint ained that economically disadvantaged students had negative attitudes about school, lower grades, and increased school absences. Nichols cited the low school attendance of underprivileged students as a possible reason for the increased dropout rate of mino rity children. He determined low SES students experienced lower school attendance than did students from a middleclass background. Thus, concluded that public education failed to teach minority children adequately. A review of the literature indicated t hat gender and parenthood played roles in academic achievement and the wellbeing of students Sus, Houston, and Suh (2007) educational attainment. Arbona and Nora (2007) found that the college aspiration rate for women showed an increase over men attaining a college degree. Further, Baer and language use in the home. In general, the lit erature showed a positive relationship among female headed households, poverty, and unemployment, and the educational attainment of their children.

PAGE 16

7 Research Problem and Significance A great number of researchers have described the relationship between SES and academic achievement for African Americans, Hispanics, and underprivileged children. The vast body of knowledge suggested that minority populations were disproportionally impover ished. However, the research literature was relatively silent regarding the relationship between SES and academic achievement of Hispanic LEP students. Educators may have assumed that the effects of poverty, such as literacy and educational attainment as e vident on African Americans and Hispanic students, also held true for the Hispanic LEP population, as indicated by the literature review. For many schools and school districts, there was an emphasis on high stakes testing (Abedi, 2004; Armor, 2006; Johns on, 2007; Kahlenberg, 2004). The No Child Left Behind Act of 2001 required that all children should meet Adequate Yearly Progress (AYP) by performing at grade level. Federal Law No Child Left Behind Act: requires each subgroup (all ethnic groups, studen ts with disabilities, students learning English, and economically disadvantaged students) in schools, districts, and the state as a whole to make adequate yearly progress (AYP) in reading, mathematics, writing, and graduation rate (No Child Left Behind Act of 2001, 2002, (Corporate Author, 2001). Many schools and school districts have focused on small groups of students to show a substantial percentage increase in AYP for subgroups such as minorities, students with disabilities, and LEP students (Meier, Ko hn, Darling Hammond, Sizer, & Wood,

PAGE 17

8 2004). An explanatory method for identifying struggling Hispanic LEP students based on SES would have been significant in increasing the understanding to help schools develop educational strategies to meet the requiremen ts of AYP for the Hispanic LEP student. The findings from th is study provided a meaningful understanding of the relationship between academic achievement and school SES, participation in the subsidized meal program, student absences, gender, GPA, FCAT, and NRT reading and math for Hispanic LEP high school sophomores. The results may assist school districts in identifying and implementing policies, such as pupil assignment based on the percentage of students participating in the subsidized meal program, whic h may improve student academic achievement for the fastest growing LEP population. Thus, the indices that this research investigated was the effect of school SES, participation in the subsidized meal program, student absences, gender, GPA, FCAT, and NRT r eading and math on the reading and math achievement of Hispanic LEP sophomores in an effort to help school districts develop educational strategies to meet the needs of Hispanic LEP students. Additionally, this study helped address the problem of the gap f ound in the research literature regarding the relationship between SES and reading and math achievement of the Hispanic LEP student. Lastly, this study expanded the knowledge base regarding the relationship between the SES of the school and the reading and math achievement of Hispanic LEP students.

PAGE 18

9 Nature of the Study In general, researchers in the field specializing in quantitative studies of minority issues in education have utilized correlational design multiple regression analyses to explore the relat ionships between a dependent variable and independent variables. Specifically, researchers such as Baer and Schmitz (2007), Carpenter, Ramires, and Severn (2006), Naglieri, Winsler, and Booth (2004) and others employed multiple regression analysis to study the relationship among independent continuous variables, such as absences from school, and categorical variables, such as gender, to a dependent continuous variable, such as Grade Point Average (GPA). The focus of this study included two dependent variabl es studied individually, FCAT reading, and FCAT math scores. Although many variables could have been used in a multiple regression analysis, Stevens (2007) cited compelling reasons to limit the number of predictor variables. Stevens presented the princip le of scientific parsimony as a reason to limit the number of predictor variables. Weisberg and Weisberg (1985) argued that the best selection of variables was based on prior knowledge based on research. Therefore, the predictor variables in this study, al so known as independent variables, were selected from the extensive literature review associated with this study Research Q uestion for R eading This study tested the following questions and hypotheses: i f there were a positive relationship between school SES and FCAT reading achievement, then, at what level of school SES did Hispanic LEP sophomores start to exhibit significantly poorer scores in

PAGE 19

10 reading achievement as defined by the FCAT after controlling for the effects of the subsidized meal program, student absences, gender GPA, and NRT reading? Research Hypotheses for R eading 1. There will be a negative relationship between enrollment in the subsidized meal program and reading achievement scores after co ntrolling for the effects of school SES, student absences, gender, GPA, and NRT reading. 2. There will be a relationship between the gender of Hispanic LEP students and academic reading achievement scores after controlling for the effects of school SES, the subsidized meal program, student absences, GPA, and NRT reading such that among Hispanic LEP students, females will have higher FCAT reading scores than males. 3. There will be a negative relationship between student absences and reading achievement scores af ter controlling for the effects of school SES, the subsidized meal program, GPA, and NRT reading. 4. There will be a positive relationship between GPA and reading achievement scores after controlling for the effects of school SES, the subsidized meal program gender, and NRT reading. 5. There will be a positive relationship between NRT reading and reading achievement scores after controlling for the effects of school SES, the subsidized meal program, student absences, gender, GPA.

PAGE 20

11 Research Q uestion for M ath I f there was a positive relationship between school SES and FCAT math achievement, then, at what level of school SES did Hispanic LEP sophomores start to exhibit significantly poorer scores in math achievement as defined by the FCAT after controlling for the effects of the subsidized meal program, student absences, gender GPA, and NRT math ? Research H ypotheses for M ath 1. There will be a negative relationship between enrollment in the subsidized meal program and math achievement scores after controlling for the effects of school SES, student absences, gender, GPA, and NRT math 2. There will be a relationship between gender o f Hispanic LEP students and academic math achievement scores after controlling for the effects of school SES, the subsidized meal program, student absences, GPA, and NRT math such that among Hispanic LEP students, females will have higher FCAT math scores than males. 3. There will be a negative relationship between student absences and math achievement scores after controlling for the effects of school SES, the subsidized meal program, GPA, and NRT math 4. There will be a positive relationship between GPA and m ath achievement scores after controlling for the effects of school SES, the subsidized meal program, gender, a nd NRT math.

PAGE 21

12 5. There will be a positive relationship between NRT math and math achievement scores after controlling for the effects of school SES, t he subsidized meal program, student absences, gender, and GPA. Methods Kromrey, Onwuegbuzie, and Hogarty, as cited in Permuth and Mawdsley (2006), relationship between minorit p 99). Therefore, a correlational design, also generally known as non experimental or non manipulative, was proposed to address the relationship between school SES, participation in the subsidized meal program, student absences, gender, GPA, FCAT, and NRT reading and math, and reading and math achievement of Hispanic LEP students. Although a correlational analysis was an appropriate method as it was designed to describe the relationship between variables, it was important to note that correlation was not c ausation. The intent of this study was not to prove causation, but rather, to determine if there was a relationship among the dependent variables, FCAT reading and math scores, and the independent variables, school SES, participation in the subsidized meal program, student absences, gender, GPA, FCAT, and NRT reading and math. Purpose of the Study The purpose of this study was to examine the relationship between school SES, participation in the subsidized meal program, student absences, gender, GPA, FCAT and

PAGE 22

13 NRT reading and math, and reading and math achievement for Hispanic LEP sophomores in the 16 traditional high schools in Pinellas County, Florida To address the central question, what was the relationship between these variables and reading and math achievement for Hispanic LEP sophomores? The Florida Comprehensive Assessment Test (FCAT) reading scores and FCAT math scores were employed as a measure of reading and math success. The FCAT test was a standardized test administered to all students in Fl orida, providing a score that was comparable from one participant to another and from one school to another in any public high school in the state. Since all public high school students were required to take the FCAT during their sophomore year, the FCAT s cores provided a measure of the reading and math success of Hispanic LEP sophomores. This correlational study used six independent variables selected from the literature review in Chapter 2 of this study: 1. The independent variable, school SES, was a composi te definition that included student percentage of participation in the subsidized meal program and the percentage of African American and Hispanic students per school. 2. The SES of the students was the second independent variable, defined by participation i n the subsidized meal program. 3. A third independent variable was the number of absences of the students as an explanatory variable based on literature that demonstrated a relationship between SES and school attendance of minority students.

PAGE 23

14 Absences was a c ontinuous variable representing the total number of days each student was absent for the 2007 2008 school year. 4. Gender was included as an explanatory variable due to the growing body of knowledge that cited the increase of reading and math ach ievement amon g female students. 5. Grade Point Average (GPA) was included as an explanatory variable based on the literature cited a relationship between SES, minority students, and GPA. 6. FCAT NRT reading and math scores were selected as an independent variable as a means of comparing one student to his or her peers ta king the same or equivalent test, such as the Stanford 10 Achievement Test from 2005 2008. Beginning in 2008, with the passing of Senate Bill 1908, school districts are no longer required to administer the nor m referenced component of the FCAT. Lastly, Hispanic students were selected taking part in one of the four LEP categories as the target sample for this study: (1) TT = at least one YES response on the Home Language Survey (HLS) test, (2) LY = ELL, eligib le for ESOL services, (3) LF = language fluent, exited, will be monitored for two years, (4) LZ = after a 2 year monitoring period, LF turns into LZ. A full description of the sample selection process was described in Chapter 3 of this study.

PAGE 24

15 Conceptual Framework The middleclass education concept holds that school structures built on White middleclass American culture differed from those of minority, impoverished culture (Kahlenberg, 2002). According to Kahlenberg White middleclass schools differed from minority, impoverished schools, in that White middleclass schools were able to provide an adequate financial base, smaller class size, increased technology, and an orderly learning environment as compared to minori ty impoverished schools that found it difficult to provide those same benefits. Indeed, minority children attending a middleclass school had higher reading and math achievement than did middleclass children attending a low SES school. SES was defined by pa rticipation in the subsidized meal program (U.S. Department of Education, 2002). The influences of peers, parents, and teachers were cited as some of the variables for academic success in middleclass schools. Hoxby (2002) found that minority students showe d increased academic performance when surrounded by academically motivated peers. Middleclass parents had greater resources to volunteer in schools and push for higher academic expectations than their lower middleclass counterparts. Teachers in low SES sch ools had higher attrition than did teachers in middleclass schools; further, lower SES schools had a higher percentage, 34%, of educators teaching out of field than did middleclass schools, which had a 19% rate of teachers out of field (Ingersoll, 2004). T hus, the literature suggested that underprivileged students, attending a poor school, had less academic success, lower grade point averages, and a greater number of absences than did underprivileged students attending a middleclass school with middleclass values

PAGE 25

16 Data Collection The data employed in this study were the 2006 2007 and 2007 2008 school years to attribute the school effect on the student population. The pre existing data were erred to as the SASI system. The archival nature of the SASI database made it possible for other researchers to replicate this study, thereby making the SASI database an effective means of gathering data to study the relationships between reading and math achievement and the SES of the school. Because SASI allowed for universal data collection for the population within its designated parameters, there was no need to conduct a random sampling to ensure group representation of the population; rather the data for all registered public high school students in the district was readily accessible. To increase the validity of this study, all Hispanic LEP sophomores in Pinellas County high schools were included in this study. This study limited the participants to 1 0th grade (sophomore) students who were enrolled in the same school during the 2006 2007 and 2007 2008 school years in order to attribute the relationship between school SES, participation in the subsidized meal program, student absences, gender, GPA, FCAT and NRT reading and math to the school. During the 2006 2007 and 2007 2008 school year, Pinellas County had 16 traditional high schools and three alternative secondary schools. All 16 high schools in the Pinellas County School District, Florida were inc luded in the research. Alternative school programs were excluded from this study due to their enrollment policies that permitted students to register in alternative education for a limited time. Due to school zoning, each school matriculated different perc entages of students of different SES,

PAGE 26

17 ethnicity, and race, and participation in educational and subsidized meal programs. Although studies have used students on the subsidized meal program as the exclusive indicator for SES, this study used the combined pe rcentage of students participating in the subsidized meal program and the percentage of African American and Hispanic students as the indicator for SES as cited in other studies such as Lewis and Cheng (2006). To help ensure the privacy of the participants and maintain an ethical study, the names of the schools were omitted to prevent any potential stigmatization (Stein, 2004). The schools were rated on a scale to indicate degree of poverty, as measured by the percentage of students registered in the subsid ized meal program and the percentage of African American and Hispanic students. See Chapter 3 for sample selection method. Definition of Key Terms Reading and math achievemen t: operationally defined as the FCAT reading and math scores with a higher score indicating a greater reading and math achievement level. Adequate Yearly Progress (AYP): a minimum of academic gains as defined by the No Child Left Behind Act (NCLB) 2001. AYP targeted participation and performance of various subgroups (White, Black, Hisp anic, Asian, American Indian, economically disadvantaged, limited English proficient, and students with disabilities). In Florida, the foundation of AYP was the FCAT. Dropout: a term used by the federal government to define students who had not completed the 12th year of high school and were not registered in school. European American: see White.

PAGE 27

18 Florida Comprehensive Assessment Test (FCAT): a state mandatory assessment d by the Florida Sunshine State Standards. The test was given only in public schools, though private school students who received opportunity scholarships must have also taken the FCAT. Home education students were tested only if their parents or guardians selected the FCAT as an evaluation option. FCAT Norm Referenced Test (NRT): score against the scores of a group of peers taking the same or equivalent test, such as the Stanford 10 Achievement Test from 2005 2 008. With Senate Bill 1908, Florida law removed the requirement to administer the norm referenced component of the FCAT. Grade Point Average (GPA): for the purpose of this study, GPA was a cumulative variable based on a 4 point scale where A = 4, B = 3, C = 2, D = 1, and F = 0. High school students start accumulating GPA with their first high school credit. Hispanic or Latin persons: persons of Cuban, Mexican, Puerto Rican, as well as South or Central American or other Spanish culture or origin regardless of race (U.S. Census Bureau, 2007). For the purpose of this study, the Federal term, Hispanic, was used, although terms such as Hispanic and Latino/Latina were interchangeable in the research literature. Limited English Proficiency (LEP): a term used by th e federal government to describe people who were deficient in the English language. Students who spoke a language other than English at home and had limited speaking, reading, and writing ability to understand English and scored below a designated level on the Language

PAGE 28

19 Assessment Battery Revised (LAB R) and New York State English as a Second Language Achievement Test (NYSESLAT), may have been eligible for LEP services. No Child Left Behind (NCLB) Act of 2001: a federal policy signed into law in 2002 design ed to increase the academic achievement of all students. The law required states receiving federal funds to evaluate all students in public schools to determine whether schools, school district, and states have met adequate yearly progress. The goal of No Child Left Behind was to have 100% of the students proficient in reading, writing, and math by 2013 2014. Minority: operationally defined as African American and/or Hispanic students for the purpose of this study. Socioeconomic status (SES): a composite te rm used to describe the economic status of individuals or a population. For the purpose of this study, SES referred to the percentage of minority students and students on the subsidized meal program. Delimitations of the Study The focus of this study was 10th grade Hispanic LEP students in Pinellas County, Florida. The reading and math achievement of Hispanic or other minority populations other than the 10th grade Hispanic LEP population was outside the scope of this study. Beca use of the limited access to private school educational and demographic data, this study was limited to the 16 traditional public high schools in Pinellas County, Florida. The purpose of this study was not to investigate the relationship between SES and th e

PAGE 29

20 likelihood of a Hispanic student participating in a LEP program, but rather this study investigated reading and math achievement of Hispanic LEP students. Summary Chapter 1 described the fastest growing minority population, the Hispanics, and explained that Hispanics were the largest population requiring LEP services. Hispanic LEP students were the focus of this study that examined the relationship between school SE S, participation in the subsidized meal program, student absences, gender, GPA, FCAT, and NRT reading and math, and reading and math achievement as defined by the FCAT reading and math scores. In order to attribute the effect of the school to the reading a nd math performance of the Hispanic LEP student, this study further limited the population to sophomores that were registered at the same school during their freshman year. Chapter 1 reported the research problem and significance, discussed the purpose of this study, conceptual framework, nature of this study, research questions and hypotheses, methods, data collection, and definitions of key terms. Chapter 1 concluded with the limitations and delimitation of this study. Overview of Upcoming Chapters Chap ter 2 review ed the literature associated with the academic achievement of minority and Hispanic LEP students. The literature introduced numerous variables to include school SES, participation in the subsidized meal program, student absences, gender, GPA, F CAT, and NRT reading and math as having a relationship with academic

PAGE 30

21 achievement for minority populations and Hispanic LEP students. The middleclass education concept was the guiding concept behind this study. Chapter 3 discussed the correlation approach for this quantitative study. Chapter 3 further elaborated on the research methods: research perspective, research context, sample, dependent, and independent variables, research questions and hypotheses, correlational assumptions, instruments employed in t he data collection, and research design employed in this study. Chapter 3 concluded with an explanation of the data analysis and a summary of the chapter. Chapter 4 analyzed the results of the correlation and multiple regression analyses for this study to include a review of the research questions, hypotheses, and the important variables under consideration. Chapter 4 examined the design of th e study and criteria, discussed the descriptive statistical analyses, presented the assumptions, and established th e screening and diagnostic procedures. T his chapter reintroduced each research question and hypothesis, and provided the results, including the descriptive statistics, the correlation matrix for FCAT reading and math, the regression analysis for FCAT read ing and math, and the multiple regression formula for both reading and math scores. Chapter 4 concluded with a summary of the substantive findings and a summary of the chapter. Chapter 5 provide d a summary and a context of the findings conducted in this st udy to include the research problem and purpose of this study the variables and sample, and the research design This was followed by a summary of the research questions, hypothesis, and their respective analysis implications and explanations of the findings the limitations recommendations for future researchers and a summary of this study.

PAGE 31

22 Chapter 2 Review of the Literature Overview of Chapter 2 Chapter 2 reviews the literature associ ated with this study used to define key terms (Limited English Proficient (LEP), Hispanic, minority status, and socioeconomic status (SES)) throughout this study. The following sections examined the effect of SES on academic achievement: (a) SES and Academ ic Achievement for Hispanic LEP Students, (b) High Stakes Testing and NCLB for Hispanic LEP Students, (c) SES, Hispanic LEP Students, and the School Community, (d) SES, Hispanic Students, and LEP Programs, (e) SES and Ethnicity as a Variable of Academic Ac hievement for Hispanic LEP students, (f) SES, Hispanic LEP Students, and Educational Attainment, (g) SES Characteristics of Disproportionately Hispanic LEP Schools, (h) SES and Acculturation for Hispanic LEP Students, and (i) SES, Hispanic LEP Students. L astly, the literature examined contradictions found in the research. Included throughout this literature review was a critical analysis of a sample of the literature in an effort to strengthen future research. A conclusion followed the literature review su mmarizing the information ascertained in this document.

PAGE 32

23 Introduction If there is to be meaningful change in the education of Hispanic LEP children in the U.S., then there must be an understanding of the challenges faced by students who have limited English proficiency in American schools. Thus, the literature review both ex tended the knowledge base and identified areas for future research with respect to SES and the academic achievement of Hispanic LEP students. Although many researchers (e.g., Alexander & Salmon, 2007; Battle & Pastrana, 2007; Flowers, 2008) have noted the relationship between socioeconomic status (SES) and academic achievement among African American and Hispanic students, few researchers have compared the academic achievement of students enrolled in Limited English Proficiency (LEP) programs based on the S ES of the school. Therefore, this literature review centers on the theme of SES and Hispanic LEP students. The literature review addressed the following question: What were the relationships among school SES, participation in the subsidized meal program, s tudent absences, gender, Grade Point Average (GPA), Florida Comprehensive Assessment Test (FCAT), and Norm Referenced Test (NRT) reading and math on the academic achievement of Hispanic LEP students? Middleclass Education Concept This study examined the relationships among school SES, participation in the subsidized meal program, student absences, gender, GPA, FCAT, and NRT reading and math to academic achievement of Hispanic LEP sophomores as seen through the middleclass education concept. The middleclas s education concept holds that school

PAGE 33

24 structures built on White middleclass American culture differ from those of minority, impoverished cultures (Kahlenberg, 2002). According to Kahlenberg (2004), White, middleclass schools differed from minority, impover ished schools, in that White, middleclass schools were able to provide an adequate financial base, smaller class size, increased technology, and an orderly learning environment as compared to minority, impoverished schools that found it difficult to provid e those same benefits. Indeed, minority children attending a middleclass school had higher academic achievement than did middleclass children attending a low SES school, where SES was defined by participation in the subsidized meal program (Burris et al., 2006; Lewis & Cheng, 2006; Martin et al., 2005). The influences of peers, parents, and teachers were cited as some of the variables related to academic success in middleclass schools. Hoxby (2002) reported that minority students demonstrate increased acade mic performance when surrounded by academically motivated peers. Because of resources, middleclass parents were able to volunteer in schools and push for higher academic expectations than their lower middleclass counterparts. Additionally, Ingersoll (2004) noted that teachers in low SES schools had a higher attrition rate than did teachers in middleclass schools; further, lower SES schools had a higher percentage, 34%, of educators teaching outside of their area of certification, known as out of field, than did middleclass schools, which had a 19% rate of teachers out of field. The research by Hoxby (2002), Ingersoll (2004), and Kahlenberg (2004) was similar in that these authors analyzed existing data to derive their findings. Ingersoll reviewed and analyz ed qualitative data from the School and Staffing Surveys and

PAGE 34

25 Teacher Follow Center for Educational Statistics (NCES). The focus of this literature review was to provide research outcomes based on the current literature within the last five years that described the academic achievement of Hispanic LEP students compared to SES indicators. Numerous terms throughout the literature have been used to refer to students who were not native English speakers. The U.S. government used the term limited English proficient in their official documents. Yet others, because of the negative connotation of the term limited English proficiency, used English language learners (ELL). Here, ELL and LEP were inter changeable throughout this study. Strategies for Searching Literature First, a broad search of the literature was conducted on Hispanic and Latino LEP using multiple databases such as Education full text, ERIC (Cambridge Scientific Abstracts), Academic Search Primer, and General OneFile. Key phrases used were of possible journal articles and other citations were narrowed by selecting articles related to academic achievement or educational attainment and SES to quantitative and peer reviewed research conducted in the United States within the last five years, from a pool

PAGE 35

26 of empirical research literature from 2004 findings, were inc luded for their historical and pertinent contribution to the role played by SES in academic achievement. Additionally, one international study by Andres, Adamuti Trache, Yoon, Pidgeon, and Thomsen, 2007, was included for its broad characteristics of low SE S schools. The search yielded 135 journal articles for the role SES played on the academic achievement and educational attainment of minority populations. The LEP Population LEP programs were designed to teach English to non English speaking or limited E nglish proficient students. School districts with more than 20 students in the same grade level and building speaking the same native language other than English were required to offer a bilingual program consisting of English language arts instructions an d content area instructions with the language arts instructions delivered in English and the content students in the same grade level and building speaking the same native language other than English were required to offer a freestanding English as a Second Language (ESL) program consisting of both the English language arts and content area instructions delivered through instructions in English and ESL methods. Students who may have been eligible for LEP services were those students who spoke a language other than English at home. These students also had limited speaking, reading, and writing ability in English due to their rudimentary English ability. They scored below a de signated level on the

PAGE 36

27 Language Assessment Battery Revised (LAB R) and the New York State English as a Second Language Achievement Test (NYSESLAT). According to research by Naglieri et al. (2006), Unger et al. (2007), and Weaver (2007), the fastest growing minority group in the U.S. was Hispanics, with the majority being Spanish speaking immigrants from Mexico. The 2007 American Community Survey (ACS) confirmed the studies by Naglieri, Unger, Weaver, and colleagues. The Hispanic population surpassed 45 mill ion persons to comprise 15.1% of the total U.S. population, thus constituting the largest minority population; African Americans followed Hispanics at over 37.3 million persons, or 12.7% (ACS, 2007). The ACS estimated that over 55 million persons, or 19.7% of the U.S. population, spoke a language other than English at home and of these, more than 24 million, 8.7% of the U.S. population, spoke English less than very well. Spanish, at over 34 million persons, or 8.7% of the U.S. population, was the most widel y spoken language at home other than English; of this 8.7%, over 16 million, or 5.8%, spoke English less than very well. The remainder of the over 20 million persons, or 7.4% of speakers other than English, included Indo European, Asian and Pacific Islande r, and other language speakers. Of these, over 8 million or 2.9% spoke English less than very well. Thus, Spanish was the predominant language spoken at home following English, and Spanish speakers were the largest population to speak English less than ver y well (ACS, 2007). Much of this minority population required LEP services. The data were clear that Hispanics represent a fast growing culturally and linguistically diverse population in the U.S. Using data from the U.S. Department of

PAGE 37

28 Education, Abedi et al. (2004) estimated that the great majority (74%) of LEP students were Hispanic, followed by Vietnamese (2%). According to Guglielmi (2008), who used data from the U.S. Department of Education, there were approximately 4.5 million pre K to 12th grade stud ents identified as in need of LEP services in the U.S. in 2000 2001. Guglielmi believed that the number of Hispanic children in need of LEP services would increase 40% by the year 2030. Guglielmi derived his findings using data from NELS: 88 2000, the Nati onal Education Longitudinal Study, in which 899 LEP participants self reported primary language (L1) multimethod (MTMM) design used two or more traits, such as parental reading ability and parental education, and two or more methods, such as logistic and ANOVA analyses, to decompose the variance and to assess the validity of the measures used for academic achievement. According to Guglielmi, Grade Point Average (GPA), English grades, Item Response Theory (IRT) scores, and other academic variables for a total of 97 variables helped to explain 75% of the variance of the academic achievement of Hispanic LEP students. Although a sophisticated design fficient information to allow future reanalysis of this research. Unfortunately, Guglielmi did not use the legislative definition of LEP (Title VII of P.L. 1000 297) and opted instead to use a composite definition that included numerous variables not found in legislation, thereby and Domina (2008), Battle and Pastrana (2007), C arpenter et al. (2006), Perna (2005),

PAGE 38

29 Schiller and Muller (2003), Warren and Edwards (2005), and Warren and Jenkins (2005), who used NELS to conduct their research. In Florida, the court case of the League of United Latin American Citizens (LULAC) v. Stat e Board of Education Consent Decree (1990) brought into focus students with limited English proficiency unidentified for LEP services. The purpose of the state decree was to ensure equal treatment of LEP students as mandated by the Civil Rights Act of 1964 Under this decree, Florida agreed to identify students eligible for LEP services and provide adequate placement and services when needed. Therefore, for this study, LEP referred to persons who spoke a language other than English as their primary language and who had limited ability to understand, write, speak, or read English, in agreement with Executive Order 13166, signed into law by President Clinton on August 11, 2000, entitled Improving Access to Services for Persons with Limited English Proficiency In his research, Weaver (2007) used the U.S. Census definition of Hispanic as defined by the country of origin. Weaver defined Hispanics as all races to include African Americans, European Americans, and Asians as well as ethnic groups from Mexico, Puer to Rico, Cuba, and countries from Central and South America. Indeed, according to Weaver, Hispanics were a diverse population that encompassed all races. Thus, for the purpose of this literature review, Hispanics were defined borr owed from the U.S. Census (2000). In his study, Weaver analyzed the statistical data based on a large scale national public opinion survey that consisted of approximately 2,817 participants 18 years of age and older, which was representative of the U.S. ad ult

PAGE 39

30 population, to study prejudice between Hispanics and European Americans. For the purpose of this survey, Weaver described his methodology adequately; however, the N = 1479, was sufficiently large to conduct his analysis; however, his minority representation dropped to N based a portion of his analysis. Further, Weaver selected the participants for certain geographic characteristics rather than by random selection. The small sample size limited generaliza bility of the results to other populations. The U.S. had experienced an increase of foreign born residents of 1.4% since the year 2000 (American FactFinder, 2006). The majority of the foreign residents, 12.5%, spoke Spanish as their primary language (Amer ican FactFinder, 2006). A review of the literature further suggested economics negatively affected the Hispanic population disproportionately. Indeed, the ACS (2003) estimated the percentage of Hispanic families with children younger than 18 living in pove rty was 46.3% for female householders with no spouse present. This was followed by African Americans at 42.6% and European Americans at 32.2%, for similar families (ACS, 2003). A review of the existing literature had suggested SES played a great role in th e academic achievement of minority students as reflected in the quality of education. In order to understand the relationship between Hispanic LEP students and academic achievement, there first must be an understanding of the role of SES on the quality of education in the academic achievement of Hispanic LEP students.

PAGE 40

31 In summary, the majority of the LEP population was Hispanic, at almost 75% of the total LEP population. Further, as indicated by numerous census data, Hispanic was the fastest growing minorit y population and the Spanish language was the most frequently spoken language in the home after English, followed by Asian languages at 7.4%. If the researchers were correct, the U.S. would continue to see an increase in the Hispanic LEP population. The li terature described the majority of the Hispanic population in the U.S. as Spanish speaking residents from Mexico, as indicated by the U.S. Census (American FactFinder, 2006). Definition of Minority Status Used for the Purpose of this Dissertation Gugliel mi (2008) referred to the Hispanic population as the minority population requiring the highest frequency of LEP services. Attewell and Domina (2008) reported that minority and linguistically isolated populations, because of their disproportional representa tion within impoverished communities, most strongly experienced the effects of poverty. Although considered a minority population in the U.S., Attewell and Domina excluded Asians due to a small sample size. In their research on the disproportionate number of minority children in the welfare system, 940 neighborhoods defined by the census tracts in three northern California counties, Freisthler, Bruce, and Needell (2007) also excluded from their study, people of Asian and American Indian descent due to low representation at the census tract level. Their research employed an ecological design considered by some researchers inferior to the cohort and case study designs (Graziano, Slavkovich, Liu, Factor Litvak,

PAGE 41

32 research was limited to three counties in California; their study limited the ability to generalize the results to other populations such as rural areas or neighborhoods with different racial or ethnic compositions. Freisthler and colleagues adequately re ported their findings; however, the authors never specified the total representation of the numerous racial/ethnic groups in their study. Simply stated, the researchers never quantified their sample size for the racial/ethnic participants in the study. Thu s, the external validity of the study was limited. Schiller (2003) noted people of Asian descent were often included in the European American comparison population. The restriction of this nature in the literature was not exclusive to the Asian population. In their research, Raver, Arber, and Gershoff (2007) restricted from their sample Alaskans, Native Americans, Pacific Islanders, Hawaiians a nd multi racial families due to insufficient sample size. Additionally, Raver et al. cited Alaskans, Native Americans, Pacific Islanders, Hawaiians and multi racial families as outside the scope of their research. Therefore, minority students were limited exclusively to African American Hispanic, and linguistically isolated populations for the purpose of this literature review. According to Planty from the National Center for Educational Statistics, Hispanics were disproportionately underrepresented in all levels of educational attainment (Livingston, 2008). Hispanics were least likely to have completed high school at 65% followed by African Americans at 87.7% and European Americans at 93.5%. degree at 11.6% followed by African Americans at 19.5%, and European Americans at 35.5%

PAGE 42

33 (Livingston, 2008). Burris et al (2006) demonstrated an academic achievement gap between Hispanic students and European American students and that this gap was evident in college admissions. The findings by Burris and colleagues were not surprising. In addition, Arbona and Nora (2007) and Stevens and colleagues (2006) indicated that Hispanics had a low high school graduation rate and the highest dropout rate, above that of African American and European American students. In their research, Arbona and 10%. Hispanics had lower rates than both African Americans, 18%, and European Americans, 34%. Martin et al. (2005) observed that Hispanics differed in college admission rates from African Americans and European Americans as a result of the that Hispanic African American and European American children. Indeed, Martin et al. investigated the relationship between college admissions at the University of California and discovered Hisp anics had fewer per capita admissions because fewer Hispanics applied as compared to their African American and European American counterparts. In summary, the researchers differed in their findings among minority groups. Often, low representation among ce rtain minority groups caused the researcher to omit underrepresented groups from the statistical analysis. Such was the case with Asian Americans and Native Americans. Instead of receiving minority representation, Asian Americans were often included with t he European American population, thus leaving the study population as predominately African Americans, Hispanics, and linguistically

PAGE 43

34 isolated populations. Of these, the Hispanic was the largest minority group and the fastest growing population. The literat ure further showed that the Hispanic populations had the largest high school dropout rates and the lowest educational attainment. Socio economic Status (SES) Defining SES In the literature review, there were numerous definitions given for SES. Burris (200 6), Lewis and Cheng (2006), and Martin et al. (2005) cited the subsidized school lunch program as an indication of both student and school SES. Yet other researchers such as Chen (2008), Kennedy et al. (2007), and Hoy, Tarter, and Hoy (2006), used a compos ite of variables that included housing value, family income, parental educational attainment, parental occupation, race/ethnicity, employment status, and residential stability as an indicator of SES. A further review of the literature showed that Attewell and Domina (2008), Battle and Pastrana (2007), Carpenter et al. (2006), Perna (2005), Schiller and Muller (2003), Warren and Edwards (2005), and Warren and Jenkins (2005) incorporated the most popular triad of possible variables, defining SES as family in their statistical analysis. Attewell and Domina (2008) based their findings on the NELS database from which the researchers selected 10,046 participants. Although the NELS: 88 1992 sample reflected national representation, Attewell and Domina did not randomly select the

PAGE 44

35 participants. Although the authors provided the results of the research without specifying the number of participants in each group studied, Attewell and Domina acknowledged that, although statistically significant, their findings were based on small effect sizes used to describe students required to take additional academic courses. y consisted of 4,156 people of Mexican ancestry and European Americans during a three year period exclusively used maternal education as an indicator of SES. In their research, Alexander and Salmon (2007) further defined absolute poverty to include the ina bility to provide the basic needs to maintain life. Their research also supported other research that concluded poverty existed disproportionately in schools with a high minority student population to include Hispanic students enrolled in LEP programs. Las tly, Rumberger and Palardy (2005) provided the final delineation of higher SES as one standard deviation above the mean. In summary, the researchers allowed for multiple definitions of SES. This pliable definition of SES allowed researchers to use the exis ting data to delineate their study population into subgroups for finding trends in the data. Although not explicitly defined, the majority of the researchers used parental education, family income, and parental occupation as the most popular variables to d efine SES. Further, the researchers pointed to a relationship between SES and the academic achievement of Hispanic and li nguistically isolated students. SES and A cademic A chievement for Hispanic LEP S tudent The current official definition of poverty was d eveloped in the 1960s in President

PAGE 45

36 Index (Duncan, 1994). In the year 2007, the average threshold for poverty for a household of four was $21,202 (American FactFinder, 200 7). Poverty clearly transcends all race, ethnicity, and gender (Love, 2007). Additionally, researchers have shown a positive relationship between academic achievement and SES (Arbona & Nora, 2007; Fry, 2005; Raver et al., 2007; Stevens, Olivarez, & Hammon, 2006). The researchers have also suggested that the longer the student lived in poverty, the more pronounced the decrease in academic achievement became (Duncan, 1994; Raver et al., 2007). Martin et al. (2005), in a data set from 1,069 public and private high schools in California in the fall of 1999, noted that the inequalities found between SES and high school course offerings. Duncan, Martin and colleagues, and Raver and others, discovered that the higher the SES of the student population, the greater t he course offerings of college preparatory classes. Additionally, they noted that a college preparatory curriculum related positively to an increase in college admission. Fry acknowledged that school characteristics were related to the education gap betwee n Hispanic and non minority students. However, Fry also noted that school characteristics did not explain the entire education gap. His study, based on data tabulated from the Pew Hispanic Center, used data from the U.S. Department of om the 2002 2003 school year. Arbona and Nora (2007) also substantiated the relationship between SES and high school course offerings. They used a national survey conducted by the U.S. Department hich included a national representative sample of 26,432 U.S. 8th grade students. Arbona and Nora found that students enrolled in rigorous academic courses were 46% more likely to have enrolled in

PAGE 46

37 college than were students enrolled in vocational or genera l tracks. Fry (2005) attributed the increase in college attendance of Hispanic students who have taken advanced academic courses to acculturation. That was to say, students placed in an advanced academic track had high educational expectations reinforced b y their peers. If correct, with a college degree. Arbona and Nora also compared college enrollment of Hispanic youth and determined that Hispanic students were the least likely to enroll full time and to remain enrolled. This further explained the disproportionate representation of Hispanics with a college degree because a strong predictor of college success was the successful completion of a rigorous course selection in high school. population in schools affected the academic achievement of the individual students. According to Kennedy et al. (2007), in a study of 303 public high schools i n New Jersey between 1999 and 2003, SES was the strongest predictor of college ambitions for high school students. The research by Kennedy and colleagues omitted private schools because the state neither tracked their resources nor held private schools acc ountable to the same standards as public schools. Likewise, Lewis and Cheng (2006) found the minority student population of the school, in addition to SES, was a possible indicator of educational aspiration. Lewis and Cheng deemed the educational experien ces of low SES and Hispanic students quite different from their European American middleclass counterparts. Indeed, Lewis and Cheng, in their study that consisted of data obtained from the National Center for

PAGE 47

38 Educational Statistics (NCES), utilized partic ipation in the school lunch program as an provided a smaller percentage, 33.3%, than what was recommended for a logistic regression by Babbie (1998). Lewis and Cheng used a five point Likert scale ranging from (1) very little to (5) very much to ask principals to rate the importance of curricular items such as perceived community level income, percent of students on the subsidized meal program, and the percent of minority students enrolled in the school. Their study cited several concerns that limited the ability to generalize the results to other characteristics such as poverty, which may have missed latent SES variables. In addition, the study cited low representation for a total of five urban schools, thus further limiting the ability to generalize to urban schoolchildren, as it was possible that urban schools may have differed from rural schools in student population. A possible explanation for the differences in curricular selection between Hispanic and non Hispanic students offered by Martin et al. (2005) was that these students were less able to afford college tuition which there fore reduced their motivation to register for advanced academic and college preparatory classes. However, Arbona and Nora (2007) contradicted the literature by Martin et al. that financial restraints kept Hispanic students from engaging in higher education Arbona and Nora noted that financial restraints were not predictors for Hispanic students that attended either a two year or a four year institution. Additionally, Andres et al. (2007) and Rumberger and Palardy (2005) linked a possible explanation for th e lower academic achievement of

PAGE 48

39 Hispanic students to the low SES of Hispanic parents. According to Andres et al., who Project of British Colombia, higher SES parents h ad the requisite skills and information necessary to make postsecondary opportunities available to their children. Andres et al. revealed the inability of low SES and Hispanic parents to communicate with the teaching staff. Rather than the views of lower S ES parents that viewed the educational professional staff as experts, higher SES parents communicated with the educational professionals as partners. This difference between the views between lower and higher SES parents accounted for the increased expecta tions of higher SES parents and their ability to communicate those expectations to the teaching staff, thereby providing higher SES students with a better quality education. In addition, Battle and Pastrana (2007) found increased speaking and writing ski lls in higher SES parents; higher SES parents had the ability to allocate necessary resources to educational endeavors that increased the academic performance of their children. Battle and Pastrana and Rumberger and Palardy (2005) further established a pos itive relationship between income and academic attainment. A review of their research showed Hispanic and linguistically isolated students were more likely to be poor and attend a lower SES school than were their European American counterparts (Rumberger & Palardy, 2005). Similarly, Hoy et al. (2006) referred to the landmark study, Equality of Educational Opportunity of 1966, commonly referred to as the Coleman Report, as overwhelming evidence of the strong association between SES and academic achievement. In their research on 96 non randomly selected high schools in an

PAGE 49

40 undisclosed Midwestern state, Hoy et al. hypothesized one of the reasons for increased academic achievement of higher SES students was the increase in what the researchers referred to as acad emic optimism. Even before the Coleman Report of 1966, Becker (1964) theorized an economic model that estimated academic attainment based on income. The research on education and poverty maintained that students who were economically disadvantaged tended to have negative attitudes about school, had lower grades and test scores, and were more likely to drop out (Reyes & Capper, 1991). The school attendance of low SES minority children was one possible explanation for the relationship between academic achiev ement and poverty. Nichols (2003) analyzed the attendance data of students from low SES and middleclass families and determined that low SES students experienced lower school attendance. He postulated the public educational system failed student population s, namely poor, Hispanic, and LEP students. A summary of the literature showed a relationship between SES and academic achievement, although SES did not explain 100% of the academic gap between Hispanic students and their middleclass counterparts. Indeed, the academic course offerings of high schools with a high SES student population partially explained the education gap between Hispanic students and their middleclass peers. Additionally, the post secondary aspirations of high school students reflected SE S as seen in the research by Kennedy et al. (2007). Lastly, the literature also tied SES to race and ethnicity. Certainly, the hypothesis that SES predicted academic achievement had existed since the 1960s, when Becker (1964) developed an economic model to estimate academic attainment.

PAGE 50

41 SES, High Stakes Testing, and NCLB for Hispanic LEP Students As the review of the literature indicated a relationship between SES and academic achievement, it also suggested a relationship between SES and enrollment in LEP for Hispanic students. Indeed, researchers such as Perna (2005), Chen (2008), Kennedy et al. (2007), Lewis and Cheng (2006), and Rumberger and Palardy (2005) used race/ethnicity was p 91). Warren and Edwards (2005) and Warren and Jenkins (2005) affirmed that the effects of race/ethnicity and low SES was most devastating in states with high poverty rates, increased minority rates, and states with high stake s testing such as Texas and Florida. Due to the No Child Left Behind Act of 2001 (NCLB) in states with high stakes testing, teachers tended to be more controlling, which resulted in a less desirable classroom for students who were extrinsically motivated (Stevens et al., 2006). These researchers concluded that Hispanic students tended to be more extrinsically motivated and the effect of high stakes testing disenfranchised Hispanic students. Stevens et al. surmised that because Hispanic students appeared to be extrinsically motivated, they stood to lose in states with high stakes testing because of the effect these tests had on the classroom teacher. Stevens et al. derived their data from convenience sampled schools in west and south Texas. Stevens et al. ra ndomly selected the 666 4th to 10th grade participants based on their representation of their school populations in age, race, ethnicity, and gender. The authors adequately described the statistical analysis used in the study to replicate for future studie s. Likewise, the authors adequately described the

PAGE 51

42 results to compare to future studies. However, there were limitations to this study because of its limited participant sample. The sample was not representative of any school or school district because the sample was not randomly selected. Rather, the sample was selected based on the location of the schools. Further, the analysis relied on subjective self evaluation from students, parents, and teachers, thereby providing many opportunities for participants t o skew the results. The study provided for a theoretical model to evaluate math performance based on emotion, motivation, and condition; however, the researchers acknowledged that these variables might not be appropriate to measure mathematics performance. The majority of the participants were in either 4th grade (133), 8th grade (393) or 10th grade (71), which left 69 students divided by four additional grade levels, or approximately 17 students from the 5th, 6th, 7th, and 9th grades, thus providing too sm all of a sample to conduct meaningful statistical analysis for these grade levels. Lastly, the authors reported that a cross validation yielded a poor model to fit for the Hispanic population, further reducing the ability to generalize these results to oth er populations. States that required passing a high stakes test in order to graduate placed Hispanic and linguistically isolated low SES students at a disadvantage, as was evident in Texas and Florida. According to Warren and Jenkins (2005), from data obt ained from the Current Population Survey conducted by the Bureau of Labor, there was a high relationship between school retention at grade level and the dropout rate. A review of literature found high stakes testing most negatively affected Hispanic and lo w SES students (Schiller & Muller, 2003; Warren & Edwards, 2005; Warren & Jenkins, 2005).

PAGE 52

43 Supporting the assertion that high stakes testing contributed to the dropout rate for linguistically isolated low SES and Hispanic students, Schiller and Muller analy zed the Carnegie units, high school credits, of low SES Hispanic students in states with extensive testing and determined that low SES Hispanic students had, on average, almost one European American counterparts. Similarl y, Battle and Pastrana (2007) observed that as the SES of students increased, so did the standard test scores for high school seniors. Adding to the literature that illustrated the effects of high stakes testing, Warren and Edwards found that students from low SES families who failed the 8th grade high stakes test to enter high school were increasingly more likely to drop out of high school than were middleclass students. In spite of the empirical data that cited the relationship between high stakes testing and the dropout rate, there seemed to be a public perception that raising the passing scores for high stakes testing would promote academic gains for students (Solochek, 2008). Prolific writers and researchers on the subject of poverty and education such as Meier et al. (2004) addressed issues of poverty between Hispanic LEP students and NCLB. One of the broad goals of NCLB was to address the achievement gap of underachieving stu dents and their racial/ethnic and social counterparts. Meier and colleagues argued that NCLB punished schools with a high proportion of Hispanic LEP students and therefore created a financial incentive for schools with a disproportionate minority populatio n to drop Hispanic LEP students from their school rosters to raise the test scores of schools and make Adequate Yearly Progress (AYP). Indeed, a single Hispanic LEP student could cause a school not to meet AYP. This was especially true for

PAGE 53

44 schools with a d isproportionate representation of Hispanic and LEP students. Meier et al. found that the more diverse the population, the more the school was at risk of not meeting AYP because each subgroup, such as Hispanics and LEP students, must meet the same standard as schools that were more affluent with a more homogeneous population. Meier and fellow researchers also addressed the financial inequities of NCLB with respect to funding schools. According to these authors, federal statistics showed disproportionate numb ers of Hispanic LEP and low SES students in more crowded classrooms with fewer certified teachers, counselors, and school supplies such as books and computers. According to Meier et al., NCLB had not addressed societal issues such as health insurance, nutr ition, housing, and other variables, widening the education gap between Hispanic LEP and low SES students and their middleclass counterparts. In summary, the literature further tied low SES to Hispanic students, specifically Hispanic LEP students. The comp osite construct for SES included race and ethnicity. Further, the greater the minority population, the greater was the effect of SES in states with high stakes testing, such as the Florida high stakes test, the FCAT (Florida Comprehensive Assessment Test). Schools with a disproportionate Hispanic LEP population also had more difficulties meeting the requirements of NCLB, known as AYP. Lastly, although logic would dictate that the neediest would receive the most help, researchers such as Meier et al. (2004) concluded that the poorest students, namely Hispanic LEP students, received the least in terms of certified teachers and classroom supplies, and were placed in more crowded classrooms than were their more affluent counterparts.

PAGE 54

45 SES Hispanic LEP S tudents, a nd the S chool C ommunity Martin et al. (2005) and Rumberger and Palardy (2005) concluded that, because housing patterns were segregated into minority and SES clusters, so, too, schools were segregated by housing patterns. Thus, segregated schools experien ced higher concentrations of academically disadvantaged students (Rumberger & Palardy). In their empirical research, Rumberger and Palardy found that in the year 2000, approximately 70% of Hispanic LEP students attended predominately minority majority scho ols. Similarly, Fry (2005) noted that Hispanic children were more likely to attend overcrowded and lower SES high schools with a higher student teacher ratio than were African American or European American children. An additional pattern prevalent in scho ols with a disproportionately low SES and high student Hispanic LEP rates was course offerings. Lewis and Cheng (2006) found schools located in less affluent communities tended to offer more vocational education and fewer advanced placement and college pre paratory classes than schools located in SES status affected the academic expectations of the educational staff. Rumberger and Palardy (2005) asserted that teachers in high poverty schools responded to Hispanic LEP and low SES students with a less demanding curriculum and teachers in Hispanic LEP and low SES schools had lower expectations than those in predominantly European American middleclass schools. Simply stated, s tudents attending low SES schools had lower academic achievement than did students with comparable backgrounds attending middleclass schools, according to Rumberger and Palardy. These researchers also

PAGE 55

46 postulated that students from advantaged backgrounds at tending a lower SES school would be successful, regardless of the characteristics of the school. Rumberger and Palardy further noted that less qualified teachers, those teaching outside of their area of certification, taught low SES minority students as co mpared to qualified teachers determined both the educational track and the economic future of the graduate. Lewis and trust in parents, as the three main school organizational factors that influenced student that the academic focus of the school played a great role in determining student achievement. The study by Sus et al (2007) used data obtained from a national sample of approximately 9000 youth ranging from 12 to 16 years of age. Sus, Houston, and Suh cited poor academic achievement, school suspension rates, low SES, and Hispanic status as strong predictors of the dropout rate. These researchers additionally cited the educational attainment of single mothers as being a strong predictor of the high school dropout rate for low SES and Hispanic students. However, a closer review of the study revealed that using a partial correlation explained less than 3% of the variance for attainments. Indeed, the greatest variance for dropout rates, 3%, was explained by the independent variable, first sexual experience occurred at age 15 or below, with the remaining 19 variables decreasing in their ability to explain the variance in the dropout

PAGE 56

47 rate. In all, the authors researched 20 d low GPA, out of school suspensions and low SES. A summary of the findings indicated the SES of the community reflected in the SES of the school. Further, as communities were segregated, so too did SES an d ethnic housing patterns segregate schools. An additional pattern witnessed in high Hispanic LEP schools in less affluent communities was the increase in vocational and less academically demanding course offerings along with the decrease in less challengi ng courses, followed by lower teacher expectations. SES, Hispanic S tudents, and LEP P rograms Researchers such as Arbona and Nora (2007) and Naglieri et al. (2006) have surmised that the low achievement of Hispanic children was further exacerbated by the level of parental education as shown by the vocabulary scores of Hispanic children. Naglieri et al. suggested that tied to the academic achievement of ethnic children was parental education. The research by Naglieri et al. illustrated that their sample of Hispanic LEP students came from predominantly working class families with limited English speaking skills and little educational training. These resea rchers compared Hispanic children to non Hispanic children and concluded that both Hispanic and non Hispanic children whose parents had limited education had a lower vocabulary score as compared to children with a higher vocabulary score for both Hispanic and non Hispanic research by Thornton, Collins, and Daugherty (2006) that linked SES to the education

PAGE 57

48 gap, low birth weight, emotional stress, and unwed childbirth in school aged children. Barlow suggested that insufficient exposure to books, intellectually stimulating toys, and activities found in middleclass homes was one possible explanat ion for the lower verbal and cognitive skills in minority adolescents. Furthermore, Barlow cited increased stress and lower self esteem of teenaged children in lower SES families. Carpenter et al. (2006) suggested that SES and inclusion in an LEP program y ielded similar outcomes for mathematics achievement. These authors observed that as SES increased, so too increased math achievement and educational attainment. Similarly, an increase in LEP participation yielded a decrease in mathematics achievement and l earning. The issue of mathematics achievement was of particular concern for researchers such as Burris et al. (2006), whose quasi experimental research on six middle school cohorts from 1995 through 2000 observed that mathematics attainment for high school students was a stronger indicator for college completion than grade point average (GPA) or SES. Burris and colleagues further extrapolated that Hispanics and students on the subsidized lunch program were underrepresented in upper level math classes. Buttr essing the research by Burris and fellow researchers, Martin et al. (2005) provided evidence to suggest that the lack of college information and college preparatory classes offered in the more affluent schools hampered Hispanic enrollment in college. Anoth er factor for the decreased academic achievement of the Hispanic population explored by social science researchers, such as Battle and Pastrana (2007), was the second language acquisition processes of many Hispanic students.

PAGE 58

49 To summarize, the literature s howed relationships between SES and inclusion in LEP programs and mathematics achievement scores. In other words, as SES increased, so did mathematic achievement scores. Likewise, as participation in LEP programs increased, math achievement scores decrease d. The math achievement scores for high school students were of interest for those concerned with college enrollment, as high school math attainment was a more reliable predictor of college completion than GPA. SES and E thnicity as a V ariable of A cademic A chievement Certainly, there was much literature that illustrated the relationship between race/ethnicity and educational achievement. According to Naglieri et al. (2006), 27% of Hispanics above 25 years of age had less than a ninth grade education, compa red to 4% for European Americans. As demonstrated in their research, Hispanics were less likely than were their European American counterparts to graduate from high school (Naglieri et al., 2006). Weaver (2005) estimated Hispanic parents were not only yo unger than their African American and European American counterparts were, but also less educated and thus, less likely to hold a white derived from the U.S. Census report, the Hispanic dropout rate exceeded the dropout rate of African American and European American children. In his critical review of Title I, Title IV, No Child Left Behind, and Supreme Court Decisions, Johnson noted that African Americans had the highest graduation rate in 1954 and predicted to replace the African American graduation rate with the highest terminal dropout rate were Hispanics (2007).

PAGE 59

50 The role ethnicity played in academic achievement in the Hispanic LEP population was not clear. The literature noted a disproportionate percentage of Hispanic LEP students in special education courses (Naglieri et al., 2004; Naglieri et al., 2006). In their research, Raver et al. (2007) argued that the gap in education for Hispanic children began in elementary school. Raver et al. advance d the hypothesis that the education gap paralleled the racial/ethnic and economic gap among Hispanic children and their counterparts. These researchers further concluded the gap faced by Hispanic children went beyond the school grounds and extended into ne ighborhood resources. Raver and colleagues cited one of the difficulties with interpreting their results was the complexity of their model. Based on a nationally representative sample of 21,260 children, they concluded that investment of resources and time was a stronger indicator of educational achievement for Hispanic children than for non Hispanic children. These authors used a complex model of association among school readiness variables and family factors to include income, material hardship, and paren ting to test for in equivalences. The authors acknowledged the likelihood of errors when testing for measures of parenting behaviors equivalently across participant groups. An additional concern with this study was that it was not possible to discern the s ample size. Rather, Raver and colleagues quantified the participants by percentages, but did not provide a clear quantification of N The authors cited that on average, they sampled 23 kindergarteners from each school sampled; however, the research did not provide the total number of schools used in the study, making replication of the study unfeasible without additional clarification of the participants. Raver and colleagues derived the original 21,260 pool of potential

PAGE 60

51 participant from a nationally repres entative sample. Although the lack of sample data line with other research that addressed the relationship between economic struggles and academic achievement. Alth ough the Hispanic population had experienced an increase in student enrollment, Hispanic students had not experienced an increase in the high school graduation rate (Stevens et al., 2006). Thus, the education gap further reflected the high school dropout r ate for Hispanic students. The research by Weaver (2005) characterized the social capital and educational attainment of Hispanics as different from European Americans. Indeed, in the area of educational attainment, researchers such as Coatsworth, Duncan, P antin, and Szapocznik (2006) used ethnicity as an indicator to explain academic achievement. Although Coatsworth et al. (2006) used ethnicity as an indicator for retention, Willson and Hughes (2006), using a small sample of 283 Hispanic first graders from three school districts in Texas and following school retention, found Hispanic children fared worse than did African American children and that both African Americans and Hispanics fared worse than European Americans in post academic retention. Social demo graphics were stronger predictors for academic retention for Hispanic children than for African American and European American children (Coatsworth et al., 2006). Academic distracters such as bullying were most closely associated with grade level retentio n in the minority student population. In their findings, Peskin, Tortolero, and Markham (2006) analyzed the participation rates of bullying among African Americans,

PAGE 61

52 European Americans, and Hispanics, and concluded that in their sample minority students rep orted greater participation in bullying than their European American counterparts. In their research, Perkin et al. (2006) concluded that of the three levels of bullying (victims, bullies, bully victims), Hispanic students self reported more participation in all forms of bullying than did their European American counterparts. The data by Perkin et al. were derived from five middle schools and three high schools in a large disproportionately minority school district whose student populations were predominant ly minority and Hispanic LEP students, and eligible for the subsidized lunch program. The U.S. Department of Justice, Education, and Health and Human Services provided the funding for this research. Minority students were disproportionately represented in special education programs as well (Naglieri et al., 2006; Naglieri et al., 2004; Raver et al., 2007). For these researchers, part of the discrepancy in the disproportionate representation of minority students in special education programs was due to the t raditional I.Q. test that focused on verbal ability and therefore negatively affected Hispanic LEP students, and students who did not speak standardized English (Naglieri et al., 2006). One alternative suggested by Naglieri et al. (2006), based on their st udy of 2,200 students tested by trained professionals during the Cognitive Assessment System process in 1997, was to use non verbal tests of intelligence. Naglieri et al. (2004) also made a similar recommendation to use alternative intelligence testing in their research based on the achievement scores of 148 Hispanic children receiving LEP services compared to 148 Hispanic children who did not receive LEP services. In this research, the authors

PAGE 62

53 concluded that Hispanic children who have received LEP services did not score significantly different from Hispanic children who had not received LEP services, as both groups scored in the lower half of the achievement test. In such tests of intelligence as the Planning, Attention, Simultaneous, and Successive (PASS) theory of intelligence, Naglieri et al. (2006) concluded that there were no mean statistical differences in intelligence among African American Hispanic, and European American students. Using alternative testing, Raver et al. concluded that measures of ma th and general knowledge scores adhered in similar fashion for African American Hispanic, and European American students. In summary, academic achievement and educational attainment for Hispanic students was below that of both African American and Europe an American students. In addition, Hispanics had the highest terminal dropout rate. Researchers also found that special education classes disproportionately represented Hispanic LEP students. Many researchers used the Hispanic student population as an indi cator of school retention (Coatsworth et al., 2006). Furthermore, Hispanic students fared worse than did African American students and both fared worse than did European American students in post school retention. The research also cited parental education al level as a strong predictor of school retention and demonstrated that higher levels of parental education led to lower levels of school retention. SES, Hispanic LEP S tudents, and E ducational A ttainment Reyes and Capper (1991) investigated the student dropout rate within the Hispanic community. The study cited numerous factors for the student dropout rate

PAGE 63

54 lack of effort and motivation. However, as the study reveale d, this was insufficient to explain the entire dropout rate within the Hispanic LEP population. Social scientists suggested that there were other factors to take into consideration regarding the dropout rate. The literature suggested that schools contribut ed to the dropout phenomenon through curriculum characteristics, safe school environment, and teacher student relationships among other factors (Hoxby, 2002; Ingersoll, 2004; Kahlenberg, 2000). Hispanic LEP and poverty issues were enormous hurdles for teac hers and administrators. Kahlenberg (2000) validated that low SES and LEP students were at a greater disadvantage than were students in middleclass schools. His findings suggested that LEP students fared better in schools with a higher SES student populati on. Wilkins (2006) acknowledged that reading and math were among several key factors that kept Hispanic students from moving forward in the educational system. His research found that by the 4th grade, 88% of Hispanic students were reading below grade lev el as compared to 60% of European American students; by 8th grade, 90% of Hispanic education gap of minority students by comparing the educational level of 17 year old Hisp anic students to middle school age European American students. In his research, Wilkins studied the matriculation of minority students in advanced math classes, such as trigonometry and calculus, and determined that the potential for earning a university d egree increased by 50% for minority students enrolled in advanced math classes.

PAGE 64

55 course matriculation was explanatory of college enrollment. Wilkins found that the combinat ion of reading, math, and high stakes testing explained up to 50% of the dropout rate for minority students. These findings illustrated the impact of poverty and educational quality on the dropout rate. The education gap did not stop at the high school. H ispanics were less likely to graduate from college than were their African American and European American counterparts (Arbona & Nora, 2007). Arbona and Nora estimated only 6% of Hispanics had completed a college degree six years after attending a junior c ollege, as compared to 44% of Hispanics attending a four year institution who graduated within the same six year period. Thus, Arbona and Nora concluded that the junior college was not a steppingstone for completing a four year degree. Like many researcher s, Arbona and Nora used NELS: 88 2000. They selected from a nationally representative sample of 26, 432 8th grade students. For their survey, Arbona and Nora selected 517 Hispanics from the 8th grade cohort who enrolled in a two year community college and 408 Hispanics from the 8th grade cohort who enrolled in a four year college or university from 925 Hispanic participants who received a follow up survey. Arbona and Nora used 17 independent variables to conduct their logistic regression analysis used to ex plain the significance of the dependent variables used in the study. Compared to other researchers who used NELS to conduct their statistical analysis, the research by Arbona and Nora was among the least complicated. Their summary of the results was easy t o interpret and presented in an unbiased tabular form. Further, the authors did not speculate as to the

PAGE 65

56 significance of the results, but rather simply reported the information as represented by the regression analysis. SES C haracteristics of D isproportiona tely Hispanic LEP S chools What was a disproportionately minority school? How did disproportionately minority schools differ from schools with a low minority and Hispanic LEP student population? Generally, both disproportionately minority and non disproportionately minority schools had a relatively high poverty rate as marked by free and reduced cost lunch data; certainly, poverty was a feature of both rural districts as well as disproportionately minority districts, as was low student achievement (Jacob, 2007; Reyes & Capper, 1991). Disproportionately Hispanic schools, however, also had a high proportion of Hispanic students who were in LEP programs (Meier, et al., 2004; Reyes & Capper, 1991; Viadero, 2000). These schools were located in large central cities rather than in rural set tings, small towns, or suburban areas (Russo, 2004). A single race student population, which was usually Hispanic LEP and low SES students, defined disproportionately minority schools (Tillman, 2005). Disproportionately Hispanic schools lacked certified t eachers, especially Hispanic teachers. The combination of an uncertified and non diverse teaching staff presented a concern unique to disproportionately Hispanic LEP schools. Due to the challenges faced by the community, disproportionately minority school districts had lower teacher retention, higher percentages of alternatively certified teachers, lack of educational resources and support, higher dropout rates, higher percentages of Hispanic LEP students,

PAGE 66

57 and a higher percentage of low SES students (Meier, et al., 2004; Reyes & Capper, 1991; Tillman, 2005; Viadero, 2000). Teachers in disproportionately minority schools taught of students classified as LEP was twice as high in central cities as it was in suburbs. The literature showed that teacher shortages in disproportionately minority districts made it more challenging to employ qualified teachers; thus, the hired teachers were often less qualified and less experienced tha n teachers in school districts with a low Hispanic population (Barth, 1990; Brennan & Bliss, 1998; Ingersoll, 2004; Jacob, 2007). Both disproportionately minority and non disproportionately minority schools had difficulties staffing teaching vacancies, but disproportionately minority schools were more likely to have vacancies in critical areas such as math and science. Moreover, disproportionately minority schools with a disproportionate LEP student population were substantially more likely to fill these va cancies by hiring substitutes or hiring less qualified teachers (Ingersoll, 2004). An additional reason for the teacher shortages was the lack of teachers willing to teach in the schools with the neediest of children (Zeichner, 2006). Certainly, numerous researchers had established a link between low SES and academic performance (Attewell & Domina, 2008; Battle & Pastrana, 2007; Burris, 2006; Carpenter et al., 2006; Chen, 2008; Lewis & Cheng, 2006; Martin et al., 2005; Meier et al., 2004; Perna, 2005; Schi ller, 2003; Warren & Edwards, 2005; Warren & Jenkins, 2005). The effects of poverty on the IQ scores of Hispanic and low SES students were well documented (Duncan, 1994; Naglieri et al., 2004; Naglieri et al., 2006). Duncan

PAGE 67

58 asserted that poverty was especi ally hard on the Hispanic population that lived in lower SES neighborhoods. The research on education and poverty maintained that students who were economically disadvantaged tended to have negative attitudes about school, had lower GPAs and test scores, a nd were more likely to drop out of school (Reyes & Capper, 1991). The school attendance of lower SES children was one possible explanation for the data, surveyed from lower SES students, revealed that lower SES Hispanic students had (2002) findings, based on the 1988 NELS survey, supported other researchers who cited that middleclass paren ts participated in the educational process more than did their lower SES counterparts. Fletcher found that students whose parents participated in educational indings that showed that middleclass parents were more likely than were lower SES parents to participate in school functions such as volunteering in the classroom and joining the PTA. With respect to Hispanic parents, in a great many cases, Hispanic famili es lack certain essential resources for effective parenting (National Council of Juvenile and Family Court Judges [NCJFC], 1994). In their research, Drummond and Stipek (2004) r, and were limited in meeting the basic needs of their children, such as emotional needs,

PAGE 68

59 clothing, and socialization. It was worthwhile to note more than just family members influenced academic achievement. School peer interactions also influenced academic achievement (Hanushek, Kain, & Rivkin, 2004). Armor (2006) argued that family characteristics influence children long before they begin school. Family characteristi cs racial and ethnic groups. Researchers, such as Kahlenberg (2000), found that impoverished children were more costly to educate, on average, than were children from mid dleclass families. Kahlenberg also noted students in high poverty schools received less resources such as books, computers, and advanced curriculum than did their middleclass counterparts. Alexander and Salmon (2007) used a comparison of international test scores to determine the quality of schools. In their analysis, they determined poverty negatively affected both disproportionately minority and non disproportionately minority schools by the per capita expenditures on Hispanic students was the presence of middleclass culture and values. The school community, students, parents, and teachers propelled school quality (Kahlenberg, 2000). Again, according to Kahlenberg, a key determinant in sc hool success was parental involvement. It was important to note that in a review of the literature the researchers did not blame the family for poor student performance. Instead, Wilkins (2006) cited an inadequate educational system that poorly prepared Hi spanic and low SES students for poor student performance.

PAGE 69

60 In summary, Russo (2004) found that schools with a disproportionately Hispanic LEP student population were often located in larger, less affluent central cities rather than rural settings. There wa s a contrast between disproportionately minority schools and middleclass schools. Local and national research have surmised that impoverished students and middleclass students did better in middleclass majority schools than did their counterparts in low SE S schools (Kahlenberg, 2000). According to Kahlenberg, certified teachers, advanced course offerings, well behaved and more motivated peers, and more active and influential parents marked middleclass schools. On average, teachers in high poverty, dispropor tionately Hispanic schools were more likely to teach out of field and be less qualified than teachers in middleclass schools. In addition, Kahlenberg postulated that teachers in poorer schools had lower expectations. He cited a national study of Title I sc hools that showed grade inflation in high poverty schools. The combination of uncertified teachers and teachers of color presented a concern unique to disproportionately Hispanic schools. The research suggested that a shortage of certified teachers in disp roportionately Hispanic schools and schools that served large numbers of minority and low income students (Meier, et al., 2004). Even for the certified teachers working in a high poverty school taking a stance on the achievement gap was challenging. The re search indicated that the quality of a teacher was the paramount factor in the academic achievement of students, even more so than the out of school factors such as the earning potential and educational level of the parents (Brennan & Bliss, 1998; Jacob, 2 007). The challenges in hiring certified teachers, especially Hispanic teachers, to

PAGE 70

61 in attracting, hiring, and retaining teachers (Brennan & Bliss; Jacob). SES and A cculturation for Hispanic LEP S tudents The trend for Hispanic families was a process of alienation whereby Hispanics felt a sense of social, political, economic, and linguistic isolation, which lead to a sense of powerlessness in the educational process f or the linguistically isolated Hispanic population (Johnson, 2007; NCJFJ, 1994). Indeed, minorities were over represented in the group with the least economic, social capital and political resources. Wilkins (2006) asserted that the educational system fai led its responsibility to educate the poor and minority youth in disproportionately minority schools. Wilkins acknowledged that the longer Hispanic children were in school, the greater the achievement gap between them and their European American counterpar ts. Additionally, Battle and Pastrana (2007) cited acculturation issues as a negative factor affecting student academic achievement. In their April 2005 pilot survey, conducted in a large high school with a disproportionate number of Hispanic students fro m southern California, comparing median household income and school communities, Unger et al. (2007) associated acculturation with health and social capital outcomes for Hispanic high school students to include drug and alcohol abuse, violence and suicide. Unger et al. defined acculturation as an interchange of behavior and cultural attitude when persons from different communities and backgrounds interacted with each other. The stress associated with acculturation among the immigrant community promoted feel ings of isolation among the Hispanic population as evident in the research by

PAGE 71

62 Coatsworth et al. (2006). Their research focused on 143 African American and Hispanic families assigned to an intervention program, randomly selected from three secondary inner c ity schools in a southeastern city. In his research, Johnson (2007) found that Hispanic immigrants had a disproportionate mobility rate as compared to their nonimmigrant counterparts, which may have lead to the perceptions of isolation among Hispanic immig rants. Although many of the researchers wrote about the benefits of high levels of acculturation, Bear and Schmitz (2007) noted that low levels of acculturation among Hispanic families were associated with families that were more cohesive. However, other r esearchers, such as Arbona and Nora (2007), noted positive associations between acculturation and academic achievement. For example, they noted that when Hispanic students associated and socialized with college bound students, the academic achievement of H ispanic students increased. These researchers further espoused that peer association with a friend with similar college aspirations was a motivating factor in college plans for Hispanic students. Thus, Arbona and Nora recommended a cohort approach for Hisp anic students as a tool for educational attainment. This assertion supported the hypothesis that academic tracking facilitated peer association (Arbona & Nora). Indeed, Arbona and Nora believed that by the tenth grade, Hispanic students with college aspira tions identified others in their peer group with similar academic goals and together, planned for their academic future. Likewise, Baer and Schmitz (2007) reported that more acculturated, educated Hispanic families promoted their children entering the Eur opean American middleclass than did their non acculturated counterparts. Based on their research, Baer and Schmitz

PAGE 72

63 believed that linked to SES was language usage and a marginalized social capital Likewise, they also believed that the more highly educated Hispanics were more likely to participate in the middleclass culture. Indeed, Baer and Schmitz surmised that the more acculturated, more educated, more affluent first generation Hispanic Americans were less likely to speak Spanish fluently. Additionally, A rbona and Nora (2007) noted Hispanic students were more likely to attain a college degree if their parents expected them to attend college. However, Arbona and Nora noted parental education and expectation had little play on whether the student selected a four year versus a two year institution. Battle and Pastrana (2007) found that low SES Hispanic parents expected their children to translate both linguistically and culturally for non English speaking family members. The role of family translator added add itional social, legal, and financial concerns normally not faced by children, thereby distracting Hispanic children from their academic endeavors. The findings from Battle and Pastrana also revealed that even when the SES of the Hispanic population rose, t heir academic achievement levels remained lower than that of their European American counterparts. Battle and Pastrana believed that the increase in SES by the Hispanic population was due to an increase in exertion in the work force such that their efforts drew the Hispanic family away from educational activities that promoted higher academic achievement. Stevens et al. (2006) found that Hispanics were a vital source of low skilled, low wage, agricultural, and other forms of labor, further buttressing the r esearch by Battle and Pastrana. In addition to lower status and professional opportunities, Hispanics also experienced almost twice the unemployment rates of their European American counterparts (Stevens et al.).

PAGE 73

64 In a similar finding, Johannessen (2003) o bserved the plight of linguistically isolated children as a key factor for their diminished level of academic achievement. Johannessen noted that linguistically isolated families drew disproportionately from low SES families. Indeed, in his research, Weave r (2007) cited Hispanics as having a lower social capital less education, lower income, less professional positions, and higher rates of poverty than their European American counterparts. Battle and Pastrana (2007) and Raver et al. (2007) acknowledged tha t the Hispanic population was a greatly varied population, but categorized them into one ethnic classification for their research. For example, Raver and colleagues considered Cubans, Mexicans, and Puerto Ricans as Hispanics. The documented variables of th e Hispanic population in the research by Wilson and Hughes (2006) supported the considerable variation of the Hispanic population. Coatsworth et al. (2006) also noted that the variables affecting the academic achievement, retention, and intervention of His panic students varied greatly across different Hispanic subgroups. In summary, Hispanics were not just isolated linguistically, but also socially, politically, and economically. Acculturation issues helped explain some of the reasons for the lower academic achievement of Hispanic LEP students. Economic stress for linguistically isolated families was an additional factor cited for lower academic performance of Hispanic children. However, the literature also cited positive aspects associated with acculturatio n associated with academic achievement. The researchers noted that more educated and acculturated families promoted entrance into the American middleclass by encouraging their children to pursue higher educational attainment.

PAGE 74

65 SES, Hispanic LEP S tudents, an d C ontradictions in the R esearc h Although a review of the literature indicated that linguistically isolated and low SES students did not fare as well academically as did their higher SES and European American counterparts, some researchers seemed to contradict these findings. Battle and Pastrana (2007) noted they were surprised to discover that their data showed that when controlling for SES, Hispanics outperformed European Americans 2 years post high school o n socioeconomic indicators. Indeed, their research indicated no statistical differences between European American and Hispanic high school seniors when controlling for SES. Battle and Pastrana concluded that SES, not race or ethnicity, was a better predict or of high school completion. Nonetheless, they also found that European Americans, more than minorities, benefited from an increase in SES. Carpenter et al. (2006) concluded that Hispanics fared as well as their European American counterparts if Hispanic s had not been in LEP or similar programs, had an intact home, and had attended a private school. Carpenter et al. further concluded that the two variables, private school attendance and intact homes, were not as strong a predictor of academic success for Hispanics as they were for European Americans. Although Kennedy et al. (2007) acknowledged the role of SES on college aspirations of isolated students; they noted their surprise at discovering that their data yielded no noticeable difference in the college aspirations of upper middle SES minority students. The research by Andres et al. (2007) found that although both low and middle SES parents desired their children to graduate from high school, low SES families focused on completing high

PAGE 75

66 school whereas mid dleclass families focused on graduating from high school as a pathway for higher education. Gender A review of the literature by Arbona and Nora (2007), Baer and Schmitz (2007), Freisthler et al. (2007), and Perkin et al. (2006) demonstrated that gender and parenthood played roles in the academic achievement and wellbeing of students. Students from college educated mothers had higher reading scores than did students whose parents did not have college educations. Sus, Houston, and Suh (2007) noted a relati onship between Arbona and Nora cited that the college aspiration rate for women showed an increase with 33% more likelihood than men to hold a college degree. Further, Baer and Schmitz also added that gender played a role in bilingualism and language use in the home among Hispanic children. Their study indicated that children fr om more highly educated, bilingual parents had higher educational attainment than did children from less educated parents who were not bilingual. The literature showed a relationship between female headed households and poverty, unemployment, and the educa tional attainment of their children. As the income from female headed households increased, so did the educational attainment of their children.

PAGE 76

67 Conclusion The purpose of this literature review was to explore the relationship between SES of the school and the academic achievement of Hispanic LEP students as seen through a middleclass education concept. I was primarily interested in peer reviewed empirical res earch conducted within the last five years in the United States that addressed the relationship between SES and Hispanic LEP students; as a result, the literature yielded 73 references to help explain this relationship. Critical analyses of eight of these references were selected based on their empirical data as cited throughout the literature. It was important to consider some of the limitations of this analysis before discussing the conclusions. First, although the literature was rich in the number of pe er reviewed journal articles on the subject of SES, academic achievement, and minority status, few peer reviewed journal articles existed discussing the relationships among academic achievement, SES, and Hispanic LEP students. Yet another limitation was th e conceptual framework used to filter the research question and interpret the literature involved. The middleclass education concept was but one framework; other models might lead to different perspectives. Despite the limitations cited, the literature ana lyses on these studies revealed several pertinent conclusions. In addition to being the largest and fastest growing minority demographic at over 15% of the U.S. population, Hispanics also represented the largest population speaking a language other than E nglish in the home. Further, the U.S. Census (2000) cited Spanish speakers as speaking English less than very well, more so than speakers of other languages. In addition, Hispanics overwhelmingly represented the majority, or 74%, of

PAGE 77

68 the LEP population. The researchers used race/ethnicity as a conditional variable for researching the relationship between SES and academic success. However, eliminated from the literature due to insufficient sample size were people of Asian and Native American descent and other minority groups. Yet, in other studies, people of Asian descent were included in the European American population. Nonetheless, researchers clearly defined the Hispanic populations, to include all races from Mexico, Puerto Rico, Cuba, and people from Span ish speaking Central and South America, as Hispanics, thus making Hispanics a diverse population. Therefore, minority status was limited to, specifically, African American Hispanic, and linguistically isolated persons. A review of the literature showed n o clear definition for SES. Rather, SES was a composite definition using numerous variables. The most common definition of SES used literature also showed there was a rel ationship between SES and academic achievement. In addition to affecting students directly, SES also influenced academic achievement indirectly by providing low SES students with less qualified, less experienced teachers in schools with less educational ma terials. The literature reviewed in this paper illustrated how SES played an implicit role in the educational achievement of all children. Specifically, the literature showed a strong relationship of the role SES played in the academic achievement of Hispa nic LEP students. Because Hispanics were the largest and fastest growing minority population, it made sense that Hispanics played an increasing role in the U.S. educational system.

PAGE 78

69 Numerous variables influenced academic success. Acculturation had both a po sitive and negative impact on academic achievement. The researchers cited positive peer relationships as a motivating factor for Hispanic students when making college plans. However, the researchers cited acculturation for increased substance abuse among m inority populations. In addition to the influence of acculturation on academic success, the literature cited similar educational effects of low SES such as lower verbal, math, and literacy skills for all students. The research cited differences in low SES educational attainment as witnessed by the college admissions rate of Hispanic high school students. Researchers such as Martin et al. (2005) concluded that ethnicity and linguistic isolation for Hispanics resulted in lower college application ra tes. Further, the literature also cited the impact of race/ethnicity reflected at the local community college. Arbona and Nora (2007) noted that community colleges enrolled high proportions of minority students, and that Hispanic students had lower transfe r rates to the university system than did their African American and European American counterparts. Lastly, this literature review helped highlight the lack of empirical research available on the role played by SES on the academic achievement of Hispanic LEP students. This literature, as reviewed, helped fill the missing gaps in the knowledge base and stimulated needed research in the area of SES and the academic achievement of Hispanic LEP students for providing positive change in the education of Hispani c LEP children in U.S. schools.

PAGE 79

70 Chapter 3 Research Methods Overview of Chapter 3 Included in this chapter were sections describing th is purpose, research perspective, research context, resear ch data sample collection instruments and data collection procedures, variables, research questions and hypotheses, data analysis, and a summary. The time frame and location of the study were described in detail in the research perspective section. The re search variables were described in the research participant section. Introduction The purpose of this chapter was to describe the research method employed in this correlational study of the relationship between the FCAT reading and math scores and school SES, participation in the subsidized meal program, student absences, gender, GPA, FCAT, and NRT reading and math of Hispanic LEP high school sophomores in Pinellas County. School SES was defined as a composite variable that included the proportion of students enrolled in the subsidized meal program and the percentage of African American and Hispanic students per school. In 2007, all Florida public schools participated in the FCAT. The FCAT, aligned to the Sunshine States Standards, was the

PAGE 80

71 students were expected to take the FCAT during their sophomore year and then again in subsequent semesters if they did not pass the FCAT. The sample for this correlational study was Hispanic sophomores in Limited English Proficiency (LEP) programs drawn from the 16 traditional high schools in the Pinellas County School District. Research Perspective The research employed a correlation research design. This method helped generalize the findings beyond the population in the Pinellas County School district. There were six predictor variables: (1) school SES; (2) participation in the subsidized meal program; (3) student absences; (4) gender; (5) GPA ; and (6) FCAT NRT reading and math scores. The first predictor variable, school SES, was the combined enrollment in the subsidized meal program and the percentage of the minority population, as defined by African American and Hispanic students. The combin ed percentage of participation in the subsidized meal program and minority enrollment was rank ordered from the largest number, representing a more affluent school, to the smallest number, representing the least affluent school. The American Community Survey (ACS, 2007) estimated the percentage of Hispanics in New Mexico at 44.1%, California at 35.7%, Texas at 35.5%, Nevada at 24.3%, and Florida at 20.1%. These were the states with the highest percentage of Hispanics. According to the ACS (2007), there were 61,561, 6.7%, Hispanic persons residing in Pinellas County. Of these residents, there were 2,232, 7.3%, Hispanic students

PAGE 81

72 enrolled in grades 9 12 in the 16 traditional high schools in Pinellas County. Of the 2,232 high school Hi spanic students, 833, 37%, were categorized as LEP students. The traditional 16 high schools were located within eight municipalities: Clearwater, Dunedin, Gulfport, Largo, Palm Harbor, Seminole, St. Petersburg, and Tarpon Springs (see Table 1 for distri bution of high schools and magnet programs by cities within Pinellas County, Florida ). Additionally, eight high schools also house d magnet programs: Boca Ciega High School, Gibbs High School, Largo High School, Lakewood High School, Pinellas Park High Scho ol, Palm Harbor University High School, Tarpon Springs High School, and St. Petersburg High School. Of these magnet programs, Boca Ciega High School, Gibbs High School, Largo High School, Lakewood High School, Pinellas Park High School, and St. Petersburg High School were created as desegregation magnet programs. Because each high school dr ew primarily from its local municipality and community in Pinellas County, minority representation and SES were not evenly distributed throughout the school district. Rat her, concentrated pockets of minority representation and low SES, as represented by participation in the enrollment in the subsidized meal program, were geographically clustered throughout the county. Lastly, Pinellas County had three alternative schools. These include Bayside High School, a credit recovery program, and GED center; Pinellas Secondary School for students who were reassigned for zero tolerance student code infractions, such as drugs or alcohol; and a telecommunication program, Tele School, in which expelled students may receive limited educational services. Alternative schools provide a different curriculum and had different requirements for completion than did traditional schools. For example,

PAGE 82

73 GED programs did not require a passing score on t he FCAT. The alternative schools provide d educational services on a temporary period and were outside the scope of this study. Therefore, this study focused on the traditional high school programs in which 191 Hispanic LEP sophomores, 8.5% of the Hispanic population who were enrolled in the same school during the 2006 2007 and 2007 2008 academic years. Table 1 High School Distribution by City High School Magnet Program City Boca Ciega High School Center for Wellness and Medical Professions Gulfport Clearwater High School Clearwater Countryside High School Clearwater Dixie Hollins High School St. Petersburg Dunedin High School Dunedin East Lake High School Tarpon Springs Gibbs High School Business Economics Technology Academy Pinellas County Center for the Arts (PCCA) St. Petersburg Lakewood High School Center for Advanced Technologies (CAT) St. Petersburg Largo High School Exploring Careers & Education in Leadership (ExCEL) Largo Northeast High School St. Petersburg Osceola High School Seminole Palm Harbor University High School Center for Wellness and Medical Professions International Baccalaureate (IB) Palm Harbor Pinellas Park High School Criminal Justice Academy Largo Seminole High School Seminole St. Petersburg High School International Baccalaureate (IB) St. Petersburg Tarpon Springs High School Leadership Conservatory for the Arts Tarpon Springs

PAGE 83

74 Research Context This study was a secondary data analysis of achievement and school SES, participation in the subsidized meal program, student absences, gender, GPA, FCAT, and NRT reading and math employing archival data collected by the Pinellas County School District on Hispanic LEP students. The Pinellas County School District enrolled over 105,000 students in grades K 12 for the 2007 2008 school year. Of this number, 30,211 students were registered in grades 9 12 in the 16 traditional high schools. The ethnic demographi cs of the high school students were 1,175 or 3.88% Asian, 5,251 or 17.38% were African American, 2,332 or 7.30% were Hispanic, 91 or 0.30% were Native American Indian, 860 or 2.85% were multicultural, and 20,602 or 68.19% were White non Hispanic. Overall, 8,635 or 28.58% of the student population participated in the enrollment in the subsidized meal program, including 3,236 or 10.71% African American 1,151 or 3.81% Hispanic, and 4,248 or 14.06% students other than African American or Hispanic participating in the subsidized meal program. This study analyzed archival data from the Pinellas County School Board database. Specifically, the archived data investigated w ere derived from the 2007 2008 school year from the 16 traditional high schools. To protect c onfidentiality and to insulate the sample from potential stigmatization, the schools were referred to as Schools 1 through 16. School 1 represented the school with the largest percentage of combined students on the enrollment in the subsidized meal program and the percentage of minority students, as defined by African American and Hispanic. For example, school 1 had a minority population of African American (50.9%) and Hispanic (6.3%) of 57.2% plus a

PAGE 84

75 subsidized meal program of 48.42% for a total composite S ES score of 105.62. Whereas, school 16 had a minority population of African American (4.1%) and Hispanic (5.6%) of 9.7% plus a subsidized meal program of 9.3% for a total composite SES score of 19. This study did not use any identifiers that would link to any individual student in this study, thereby insuring anonymity for all students. The combination of minority enrollment and the percentage of students participating in the subsidized meal program provided a composite measure of SES for the purpose of thi s study. A preliminary investigation of the school ranking suggested that there was a relationship between minority enrollment and participation in the enrollment in the subsidized meal program, as expected from the literature presented in this study (see Table 2 for school ordered by the composite SES, as defined by the percentage of Black and Hispanic students and the percentage of students participating in the enrollment in the subsidized meal program). The combination of minority status and the subsidiz ed meal program gave a better measure of school SES than only using one variable such as minority representation or the subsidized meal program used individually. McLaughlin (2009) cited that ineligible students were participating in the subsidized meal pr ogram. According to his study, the federal government required school districts receiving federal assistance for the subsidized meal program to audit at least 3% of the families participating in the program for eligibility. McLaughlin reported that 52% of the families participating in the subsidized meal program were ineligible, and therefore, dropped from the subsidized meal program. Although this was not a scientific study, McLaughlin did provide insight on why

PAGE 85

76 exclusive dependence on the subsidized meal program may have been a poor indicator of school SES. Table 2 Composite School SES = African American + Hispanic + Enrollment in the S ubsidized M eal P rogram School African American H ispanic Sub Meal Prg Composite SES Score 1 50.90 6.30 48.42 105.62 2 44.23 4.90 45.01 94.14 3 48.99 4.56 33.89 87.44 4 17.52 8.92 40.96 67.40 5 11.77 12.51 40.94 65.22 6 14.85 8.30 38.21 61.36 7 14.78 9.87 32.40 57.05 8 11.85 12.64 32.15 56.64 9 23.61 3.81 26.34 53.76 10 16.47 7.09 26.58 50.14 11 8.69 12.03 23.55 44.27 12 6.63 6.52 21.50 34.65 13 4.74 5.63 21.41 31.78 14 4.11 3.37 14.17 21.65 15 4.98 5.21 10.79 20.98 16 4.12 5.69 9.30 19.11 Note. Sub Meal Prg = Enrollment in the subsidized meal program Instruments Employed in Data Collection Multiple techniques and data collection processes were employed in this study. The primary data source was the 2007 2008 SASI system, a data management system designed and published by Pearson, Inc. for school use. The data encompassed in SASI included sch ool identification number, student race and ethnic identification, student language spoken at home, parent language spoken at home, participation in the enrollment in the subsidized meal program, participation in LEP program, student

PAGE 86

77 absences, and students County Schools website that provided public access to data on special programs such as magnet programs and academies found that in the traditional high schools, school racial and ethnic demo graphics, and the percentage of students participating in the enrollment in the subsidized meal program. Additionally, the Pinellas County School website this link, the pu blic can access data such as mean FCAT reading scores, FCAT math scores, racial and ethnic demographics, percentage of students participating in the enrollment in the subsidized meal program, school grade, adequate yearly progress (AYP), and average yearly gains. The primary source of data for this study was the SASI database followed by the Pinellas County School District website and the FLDOE from which racial and ethnic demographic, percentage participation in the subsidized meal program, FCAT reading and FCAT math scores data were obtained. Additional sample data were ascertained, LEP program participation, FCAT reading scores, and student absences for individual stu dents. Several procedures were followed prior to accessing information for this study. First, the requirements for Institutional Review Board (IRB) were met in order to grant approval for this study. Because this study used archival data exclusively, and therefore, no participant was interviewed or questioned, an IRB Exemption was requested and approved Further, no identifying data such as student number, residential address or any

PAGE 87

78 kind of data that can link this study to any individual were used. Additio nally, the Pinellas County School District provided a procedure similar to that of the IRB. Once the IRB and district approvals were obtained, the data requested were downloaded into an electronic spreadsheet, Excel 2007, for further manipulation and analy sis by SAS 9.1.3., a statistical analysis program. The Population The sample selected was Hispanic LEP sophomore students in the 2007 2008 2006 2007, in the 16 tradi tional high schools in Pinellas County. Thus students who changed schools or were repeat 10th graders were excluded from the analyses, to reflect students must ha ve participated in at least one of the four LEP categories: ( 1 student was classified as limited English proficient and was enrolled in a program or receiving services that were specifically designed to meet the instructional needs of ELL student 2 followed up for a two 3 ) LP 12th grade, tested fully English proficient on an Aural/Oral test and was Limited English Proficient pending the r eading and w riting assessment, or the student in K assessme 4 year follow up period had

PAGE 88

79 2007 2008). The Hispanic LEP sophomore student data in the 16 traditional high schools (without na mes, addresses, etc. included) were exported to a computerized spreadsheet, Excel 2007. To aid in the query, Hispanic, LEP, and 9th grade filters for the 2007 2006 school year were utilized to reduce the overall population to Hispanic LEP sophomores. In the 2007 2008 school year, there were 1,960 (6.5% of the total high school student population in the 16 traditional high schools) LEP students in grades 9 12 enrolled in the traditional high schools. Of these, 833 were Hispanic LEP students in grades 9 12. Of these, 191 were enrolled as sophomores and had been enrolled as freshman in the same school the prior year in 2006 2007. See Table 3 for Hispanic LEP sophomore distribution.

PAGE 89

80 Table 3 Percent Hispanic LEP Sophomore Distribution School School LF LY LZ TT Total 1 2.10 0.00 1.60 0.00 3.70 2 0.00 0.00 2.60 0.00 2.60 3 0.50 0.00 0.00 0.00 0.50 4 1.00 1.00 4.20 0.00 6.30 5 2.10 5.20 10.50 0.00 17.80 6 3.70 2.60 2.10 0.00 8.40 7 2.10 2.60 7.30 0.00 12.00 8 4.20 6.80 7.30 0.00 18.30 9 0.00 0.00 2.10 0.00 2.10 10 0.00 0.50 2.60 0.00 3.10 11 0.50 6.80 8.40 0.00 15.70 12 0.00 0.00 3.10 0.00 3.10 13 0.00 0.00 1.60 0.00 1.60 14 0.00 0.00 1.60 0.00 1.60 15 0.00 0.00 0.50 0.00 0.50 16 0.00 0.00 2.60 0.00 2.60 Total 16.20 25.70 58.10 0.00 100.00 Note LF = language fluent, exited, will be monitored for two years, LY = ELL, eligible for ESOL services, LZ = after two years monitoring period, LF turns into LZ, TT = at least one YES response on the Home Language Survey (HLS) test. Of the Hispanic LEP sop homore population enrolled in the traditional high schools, over half, 116 or 60.7%, were participating in the subsidized meal program. Of the 116 students, 91 or 78.4% were receiving a free meal whereas 25 or 21.6% were receiving a reduced cost meal. See Table 4 for Hispanic LEP sophomore participation in the enrollment in the subsidized meal program.

PAGE 90

81 Table 4 Percent Hispanic LEP Sophomore Participation in the Subsidized Meal Program School % Subsidized % Unsubsidized Total 1 3.74 0.00 7 2 2.14 0.53 5 3 0.00 0.53 1 4 3.21 3.21 12 5 12.30 5.88 34 6 4.81 3.74 16 7 8.02 3.74 22 8 12.30 5.35 33 9 1.60 0.53 4 10 1.60 1.60 6 11 8.56 6.95 29 12 0.53 2.67 6 13 1.07 0.53 3 14 0.53 1.07 3 15 0.00 0.53 1 16 0.00 2.67 5 Total 187 Note. Sub Meal Prg = Enrollment in the subsidized meal program, Free = participant receives a free meal, Reduced = participant receives a meal at a reduced priced. Dependent and Independent Variables For this study, six independent variables were selected. O ne independent variable, school SES was based on the school level. The remaining five independent variables, (participation in the subsidized meal program, student absences, gender, GPA, NRT reading and math) were student level indices. The selection of t hese variables was based on the extensive literature review that cited these variables as strongly related to academic achievement for minority students.

PAGE 91

82 1. School SES was a composite definition that included the percentage of students participating in the s ubsidized meal program added to the minority representation in each school (i.g., the percentage of African American and Hispanic students). 2. Student SES was one of two levels, participating or not participating in the subsidized meal program. 3. Student abs ences were defined as an absence for four of more periods in a school day. Student absences were limited to excused, unexcused, and out of school suspensions. 4. Gender was included as an explanatory variable due to the growing body of knowledge that cited th e increase of reading achievement among female students. 5. Grade Point Average (GPA) was included as an explanatory variable. For the purpose of this study, GPA was a cumulative variable based on a 4 point scale where A = 4, B = 3, C = 2, D = 1, and F = 0. High school students start accumulating GPA with their first high school credit. 6. FCAT Norm Referenced Test (NRT) reading and math scores were selected as independent continuous variable s The NRT allowed for res of a group of peers taking the same or equivalent test, such as the Stanford 10 Achievement Test from 2005 2008. In July 17, 2008, Florida Senate Bill 1908 removed

PAGE 92

83 the requirement for school districts to administer the norm referenced component of the FCAT beginning in the 2008 2009 school year The dependent variables for this study were the FCAT reading scores and FCAT math scores employed as a measure of reading and math success, although other measures of academic success, such as grade point average (GPA), Scholastic Aptitude Test (SAT), and Academic Competency Test (ACT), could have been selected. However, GPA, SAT, and ACT have limiting characteristics that would have reduced validity. GPA would have reduced the strength to generalize across the state, school districts, schools within a dist rict and even the same classes within a school, as individual teachers had great autonomy over the grading processes. Additionally, PCSB policy, 4.30 of the Policy Manual for the 2006 through the 2007 school year, stipulated that LEP students shall be grad ed on effort and application during an adjustment period of up to two years. The grading policy for LEP students further reduced the ability to generalize GPA to other LEP populations. However, the general research literature cited relationships between SE S and GPA. Therefore, GPA was included as an independent variable in this study. Other standard measures such as SAT or ACT may have been considered as indicators of academic achievement, this sub set would not have been representative of the population. T he FCAT test is a standardized, criterion referenced assessment administered to all participant to another and from one school to another in any school in the state. FCAT reading and FCAT math scores were selected because they were valid, reliable, and standardized. FCAT reading and math scores can also be generalized across all schools in

PAGE 93

84 the state of Florida for high school students within the same academic year, thereby increasing the generalizability of the research results. Florida law required all public high school students to pass the 10th grade reading and math portion of the FCAT in order for students to graduate with a standard high school diploma. Students may re take the FCAT as often as they want until they pass the test. Rather than receiving a standard diploma, students who did not pass the FCAT were issued a Certificate of Completion. The Certificate of Completion was not equivalent to a standard high school d iploma. A scale score of 300 for the FCAT reading and math was considered passing. As an alternative to the FCAT, students may earn a compatible score on the SAT or ACT after retaking the FCAT a minimum of three times. See Table 5 for FCAT compatible score s. Therefore, because the FCAT test was the most widely administered test in Florida, the FCAT reading and math variables were selected as the best standardized measure for reading and math achievement for this study. Table 5 Compatible FCAT Passing Scores for Reading and Math Compatible Scores Reading Math FCAT Scale 300 FCAT Scale 370 SAT 410 SAT 370 ACT 15 ACT 15 Note. FCAT = Florida Comprehensive Test, SAT = Scholastic Aptitude Test, ACT = Academic Competency Test.

PAGE 94

85 Research Questions and Hypotheses Research question for reading If there was a positive relationship between school SES and FCAT reading achievement, then, at what level of school SES did Hispanic LEP sophomores start to exhibit significantly higher scores in reading achievement as defined by the FCAT after controllin g for the effects of the subsidized meal program, student absences, gender GPA and NRT reading? A correlation and multiple regression analysis were employed to address this question. Research hypotheses for reading 1. There will be a negative relationship between enrollment in the subsidized meal program and reading achievement scores after controlling for the effects of school SES, student absences, gender, GPA and NRT reading. A correlation and multiple r egression analysis were employed to test this hypothesis. 2. There will be a relationship between gender of Hispanic LEP students and academic reading achievement scores after controlling for the effects of school SES, the subsidized meal program, student ab sences, GPA and NRT reading such that among Hispanic LEP students, females will have higher FCAT reading scores than males. A correlation and multiple regression analysis were employed to test this hypothesis. 3. There will be a negative relationship between student absences and reading achievement scores after controlling for the effects of school SES,

PAGE 95

86 the subsidized meal program, GPA and NRT reading. A correlation and multiple regression analysis were employed to test this hypothesis. 4. There will be a positi ve relationship between GPA and reading achievement scores after controlling for the effects of school SES, the subsidized meal program, gender and NRT reading. A correlation and multiple regression analysis were employed to test this hypothesis. 5. There wil l be a positive relationship between NRT reading and reading achievement scores after controlling for the effects of school SES, the subsidized meal program, student absences, gender and GPA. A correlation and multiple regression analysis were employed to test this hypothesis. Research question for math If there was a positive relationship between school SES and FCAT math achievement, then, at what level of school SES did Hispanic LEP sophomores start to exhi bit significantly higher scores in math achievement as defined by the FCAT after controlling for the effects of the subsidized meal program, student absences, gender GPA and NRT math? A correlation and multiple regression analysis were employed to address this question. Research hypotheses for math 1. There will be a negative relationship between enrollment in the subsidized meal program and math achievement scores after controlling for the effects of school S ES, student absences, gender, GPA and NRT math. A

PAGE 96

87 correlation and multiple regression analysis were employed to test this hypothesis. 2. There will be a relationship between gender of Hispanic LEP students and academic math achievement scores after controllin g for the effects of school SES, the subsidized meal program, student absences, GPA and NRT math such that among Hispanic LEP students, females will have higher FCAT math scores than males. A correlation and multiple regression analysis were employed to te st this hypothesis. 3. There will be a negative relationship between student absences and math achievement scores after controlling for the effects of school SES, the subsidized meal program, GPA and NRT math. A correlation and multiple regression analysis we re employed to test this hypothesis. 4. There will be a positive relationship between GPA and math achievement scores after controlling for the effects of school SES, the subsidized meal program, gender and NRT math. A correlation and multiple regression anal ysis tested this hypothesis. A correlation and multiple regression analysis were employed to test this hypothesis. 5. There will be a positive relationship between NRT math and math achievement scores after controlling for the effects of school SES, the subsi dized meal program, student absences, gender and GPA. A correlation and multiple regression analysis were employed to test this hypothesis.

PAGE 97

88 A Pearson Product Moment Correlation Coefficient referred to as r was used to examine the direction and strength of the relationships among the variables under investigation. r may have been either positive or negative, indicating the direction of the relationship. For example, if r were negative, as one variable increased, the other variable decreased. In a multiple re gression analysis, R squared was useful for understanding the relationship between the FCAT reading and FCAT math scores and the independent variables; student participation in the subsidized meal program and percentage of African American and Hispanic stu dents, participation in the subsidized meal program, student absences, gender, GPA, and FCAT NRT reading and math. Additionally, the correlation matrix in SAS also produced a p value associated with the correlation. The p value described the probability of obtaining a relationship of that strength purely by chance. This information was generated in a table form illustrating the intersections between each variable. Multiple regression formula for FCAT reading scores FCAT reading = b1 X1 school SES + b2 X2 meal SES + b3 X3 student absences + b4 X4 gender + b5 X5 GPA + b6 X6 NRT reading + a Where: = reading score b k = nonstandardized multiple regression coefficient for the predictor variable X k = predictor variable a = intercept constant

PAGE 98

89 Multiple regression formula for FCAT math scores FCAT math = b1 X1 school SES + b2 X2 meal SES + b3 X3 student absences + b4 X4 gender + b5 X5 GPA + b6 X6 NRT math + a Where: = b k = nonstandardized multiple regression coefficient for the predictor variable X k = predictor variable a = intercept constant Correlations, Multiple Regressions, Regression Diagnostics The statistical software package, SAS 9.1.3, was used to analyze the data and provide descriptive and inferential statistics. To screen the data for possible violations of assumptions and as a diagnostic instrument for indicating possible outliers and non linearity, scatter plots with regression lines and 95% regression intervals were employed. It was wort 2007, p 57). That was to say, some assumptions (such as the assumption of normality, that the independent variables were fixed, dependent variable errors were independent of each o ther and independent of the independent variables, and observations were both independent and representative of the population) were robust to violations, whereas other assumptions (linearity, independence, reliability of measurement, and homoscedasticity) were not robust to violations (Osborne & Waters, 2003). First, it was assumed that the variables were normally distributed. This was verified by inspection of the skewness and kurtosis of the variables. An inspection of studentized residuals (a type

PAGE 99

90 of st andardized residual) was employed using a stem and leaf plot as a means of identifying possible outliers providing they appeared to be erroneous data. Special consideration for possible deletion of data from the analyses was given to studentized residuals exceeding 2. This was done to reduce the probability of Type I and Type II errors and to improve estimates. Type I error referred to the probability of concluding there was a relationship between two variables when a relationship did not exist. Type II e rror referred to the probability of concluding there was no relationship when a relationship did exist. Secondly, because multiple regression analyses can reliably estimate the relationship between FCAT reading or FCAT math scores and the independent varia bles, if a linear relationship between the variables existed, the researcher created a scatter plot of the studentized residuals to analysis for nonlinearity. This scatter plot additionally indicated the extent of both curvilinear and linear relationships, if they existed. Failure to meet the linearity assumption may result in an increase in Type I and Type II errors by underestimating the true relationship between the dependent variable and independent variables. The assumption of reliability of measurem ent, as implied by the name, related to the accuracy of the predictor variables. That was to say, the independent variables were presumed to be measured without error. If the predictor variables were erroneous, or lacking in reliability, then each addition al predictor variable introduced reflected part of the error variance. Thus, the greater the number of predictor variables introduced into the study, the greater the likelihood that the variance was inappropriately related, causing more questionable result s that lead to Type I and Type II errors.

PAGE 100

91 Lastly, homoscedasticity was defined as consistent equal variance of errors across all independent variables. When the variance in errors was unequal, heteroscedasticity was said to occur. A marked increase in he teroscedasticity can lead to an increase in Type I errors by distorting the findings of the regression. This study employed the White s test to check if the variance of the residuals was homogenous in order to test for was a test of the null hypothesis that there was no difference in the variance of the residuals. Therefore, if the p value was statistically significant, then reject the null for the alternative hypothesis that the variance was not homogenous. Some assum ptions were less critical if they were not met. The first of these were that the independent variables were fixed. The same independent variables would be used if the study were repeated. In addition, if any errors existed, they were independent of each ot her. However, because this study used the population of Hispanic LEP sophomores in the Pinellas County School System that were enrolled in the same school at the end of the 2006 2007 freshman year, there should have been no sampling errors. In order to as sure validity of the research, a sufficient sample size was needed. Too small of a sample (e.g., 20 or smaller) may have been problematic (Stevens, 2007). Generally, researchers recommend samples of at least 15 25 per independent variable. Because this stu dy involved six independent variables, this placed the required sample size between 60 and 100. Power was defined as the probability of finding a statistical p 3). Power was dependent on alpha ( ) level, effect size, and the

PAGE 101

92 size of the sample ( N ), the design of the research, and the statistical analysis employed in the research. In general, as N increased, so did power. Cohen (1988) described Effect Size (ES) a s the ability to detect an effect of practical significance in the population. To this end, Cohen presented estimates of power for small, medium, and large effect size 2 /(1 R 2 ) for multiple regression analyses and correlations, small E S sample size of N sample in this study was N = 191, thereby exceeding the minimum recommended sampl e size for a medium effect. However, it was important to note that a statistical significance did not mean that the difference was important, practical, or helpful in decision making. Of concern, were the undue influences on the coefficients by unusual an d influential observations. To test for unusual and influential observations, this study tested for outliers, leverage, and influence of the observations. In linear regressions, outliers were defined as observations with large residuals, which may indicate data error or some sampling peculiarity. To test for outliers and unusual observations, in addition to the scatter plots and stem and leaf of the FCAT reading and math scores against each of the independent variables, the study considered studentized resi duals that exceeded two to highlight potential outliers. Leverage The measure of the extent an independent variable deviated from the mean was referred to as leverage. These points were of concern due to the effect on the est imate regression coefficients. This study employed a test for leverage. Points that

PAGE 102

93 exceed (2k+2)/ n where k was the number of independent variables and n was the number of observations, may have been considered as having had a high leverage. Influentia l d ata If removing an observation substantially changed the estimate coefficient then the observation was considered influential. To test for influential 88) a cutoff of 4/( n k 1), where k was the number of independent variables and n was the number of observations, was considered for further investigation. It was worthy to note that outliers, leverage and influential observations serve to identify suspici ous points. The removal of these data points may improve the fit of the regression; they may also destroy important information. Thus, regression diagnostics serve to point out possible errors in the observation or other substantive information that would warrant removal of these observations. Therefore, if this study excluded any variables, the results would have been reported both with and without outliers, leverage, and influential observations in Chapter 4. This study started with descriptive statistic s to produce a table of simple statistics, N mean, standard deviation, sum, minimum and maximum, for each variable. Following the descriptive statistics, correlation analyses were employed to compute the Pearson product moment correlation between each pai r of variable s This was useful for understanding the bivariate relationships among the variables. The correlation analysis was followed by a multiple regression analysis to estimate the explanatory attributes of school SES, participation in the subsidized meal program, student absences, gender, GPA, FCAT NRT reading and math to reading and math achievement, as defined by the

PAGE 103

94 FCAT. Both FCAT reading and FCAT math scores were used independently of each other in separate multiple regression analyses. A multi ple regression analysis was employed to explain variance in an interval dependent variable that could have been attributed to a set of independent variables. A multiple regression analysis can determine if there was a significant relationship between a dep endent and independent variable when analyzed as a group, or if a given independent variable was statistically significant as measured by a test of the regression coefficient for a given independent variables. In addition, a multiple regression analysis ca n address whether a specific independent variable explains a specific amount of the variance in a dependent variable. A multiple regression analysis produced what was referred to as a set of multiple regression coefficients. Each coefficient represented t he mean change in the predicted dependent variable given an unstandardized one unit change in a given independent variable, while holding the other independent variables constant. The unstandardized multiple regression coefficient was symbolized by B while the standardized multiple relationship between the dependent variable and each independent variable. Research Design This quantitative study used a correlational analysis to examine the relationship between school SES, participation in the subsidized meal program, student absences, gender, GPA, FCAT, NRT reading and math and reading and math achievement for

PAGE 104

95 Hispanic LEP sophomores in the 16 traditional high schools in Pinellas County, Florida The measures of achievement for the sample were both the FCAT reading and math scores. To analyze the data, multiple strategies were employed and sequentially executed. First, desc riptive analyses were employed to provide an understanding of the demographic data to include gender, GPA, FCAT NRT reading, participation in the subsidized meal program (both free or reduced meals), student absences as defined by an absence for four of mo re periods in a school day and limited to excused, unexcused, and out of school suspensions, and mean FCAT scores for the student population. The second phase was to reduce the student population to the study sample, Hispanic sophomore LEP students in the 16 traditional high schools in Pinellas County, Florida E xamined in this study was the relationship between school SES, participation in the subsidized meal program, student absences, gender, GPA, and FCAT NRT reading and math to reading and math achieve ments for Hispanic LEP sophomores in the 16 traditional high schools in Pinellas County. The following SAS procedures were employed to address the following research hypotheses and questions. Data Analysis A correlational study was conducted using a regr ession model to analyze the data provided by the Pinellas County School District, the Pinellas County Website, and the FLDOE website. The data were transferred into a comma separated variable (.csv) format that allowed the data to be opened and saved into an Excel spreadsheet used in Microsoft

PAGE 105

96 Office 2007. The data were then imported into SAS for analyses of the data. As was the convention in social science, an alpha level of .05 was used as a level of significance for the data in this study. Summary In summary, the analysis in this study investigated the relationship between reading and math achievement and measures of SES for Hispanic LEP students in the 16 traditional high schools in the Pinellas County School District. These analyses provided a found ation for school district personnel when deciding upon student assignment for Hispanic LEP students. The statistical data presented in this study was a cross sectional analysis of Hispanic LEP students in the 16 traditional high schools in Pinellas County for the 2006 2007 and 2007 2008 school years. The results of the aforementioned analysis were presented in Chapter 4

PAGE 106

97 Chapter 4 Research Findings The purpose of this chapter was to provide the results of t his study. The first part of the chapter reiterated the purpose of the study, and stated the broad research questions and hypotheses in order to remind the reader of the important variables under consideration The characteristics of the sample were also e lucidated. The second part of the chapter examined the design of th e study and criteria, discussed the descriptive statistical analyses, presented the assumptions, and established the screening and diagnostic procedures. The third part of the chapter reint roduced each research question and hypothesis, and provided the attendant results, including the descriptive statistics, the correlation matrix for FCAT reading and math, the regression analysis for FCAT reading and math, and the multiple regression equat ions for both FCAT reading and math scores. Lastly, a brief summary of the substantive conclusions were provided. Purpose of the Study The purpose of this study was to determine whether the relationship between SES and academic achievement of minority students, as indicated by the literature, held true for a sub population of students with limited English proficiency. This study examined the relationship among measures of school and student SES, student absences, and reading and math achievement on the FCAT for Hispanic LEP high school students. This

PAGE 107

98 study added to the body of knowledge regarding the relationships between measures of schoo l and student SES and the a cademic achievement of Hispanic LEP high school students. Design of the Study and Sample Criteria This study employed a correlational research design, utilizing a secondary data analysis. The target population for this correlati onal study was Hispanic sophomores in LEP programs enrolled in the 16 traditional high schools in the Pinellas County School District. The eligibility criteria to be included in the study were: (1) the Hispanic LEP students were sophomores in 2007 2008; an d (2) the Hispanic LEP students had been enrolled in the same high school as freshmen ninth graders. This selection criterion was established to attribute school characteristics to student performance. Students who changed schools or were repeating 10th gr ade were excluded from these analyses. Research Question If there was a relationship between school SES and FCAT reading and math achievement, then at what level of school SES did Hispanic LEP sophomores start to exhibit significantly higher scores in re ading and math achievement, as defined by the FCAT, after controlling for the effects of the subsidized meal program, student absences, gender, GPA, and NRT reading and math.

PAGE 108

99 Descriptive Statistics After examining the data for date entry errors, descri ptive statistics were executed in order to examine the important characteristics of the variables under investigation. Descriptive statistical analyses were conducted for FCAT r eading and FCAT m ath, respectively. Table 6 provides measures of central tenden cy and measures of dispersion for the appropriate variables in the study The mean, median, minimum, maximum, range, standard deviation, skewness and kurtos is are provided for each of continuous variables under investigation. The d escriptive statistics s how ed that the average FCAT reading achievement was, 273.47, ( SD = 56.90) from a possible FCAT reading score range from 100 500 for the whole population. The average FC AT m ath achievement was, 308.55, ( SD = 47.57) from a possible FCAT math score range from 100 500 for the whole population The average school SES was, 8.29, ( SD = 3.55). The average GPA was estimated to be 2.50 ( SD = .86) For NRT reading, the mean was estimated to be 685.90, ( SD = 43.60 ) For NRT math the average score was 697.05, ( SD = 29. 55). Approximately 60% of the students in these schools were participating in the subsidized meal program The mean of student absences was 11.62 ( SD = 12.82) The sample was 56% male and 44% female

PAGE 109

100 Table 6 Descriptive Statistics for FCAT Reading and M ath, N = 172 Variable M ean Median Min Max Range SD Skew Kur FCATRead 273.47 279.50 109.00 465.00 356 56.90 0.08 0.49 FCAT Math 308.55 317.00 100.00 398.00 298 47.57 1.64 5.15 School SES 8.29 9.00 1.00 16.00 15 3.55 0.14 0.63 Meal SES 0.40 0.00 0.00 1.00 1 0.49 0.43 1.84 Absences 11.62 8.00 0.00 97.00 97 12.82 2.74 12.13 Gender .56 1.00 0.00 1.00 1 0.50 0.26 1.96 GPA 2.50 2.53 0.42 4.00 3.58 0.86 0.21 0.93 NRT Read 685.90 691.00 555.00 803.00 248 43.60 0.04 0.20 NRT Math 697.05 692.00 555.00 796.00 241 29.55 0.03 3.11 Note N = S ample S ize, Min = Minimum, Max = Maximum, SD = Standard Deviation Skew = skewness, Kur = kurtosis. Statistical Assumptions There are certain statistical assumptions underlying multiple regression analyses that must be examined prior to the conduct of these analyses. The statistical assumptions underlying the multiple regression procedure are illustrated below: 1. X is fixed 2. X is measured without error 3. Linearity 4. Collinearity 5. Homoscedasticity of errors (or homogeneity of variance) 6. Errors are normally distributed 7. Errors are independent of each other 8. Errors are independent of X

PAGE 110

101 Data collected for this study was from the P inellas County School District and the Florida Department of Education and were assumed accurate. Normality The assumption of norma lity of the individual variable distributions was met using a visual inspection of the stem a nd leaf plot, and an inspection of skewness and kurtosis for all variables. The results of the stem and leaf plot reflected normal distribution s The values of the skewness and kurtosis fell within a near normal range of 2, as suggested by Chen and col leagues (n.d.). Stem and leaf plots were created using studentized residuals ( a type of standardized residual created by predicting FCAT scores from school SES, student SES, student absences, gender, GPA, and NRT ) Examination of these plots revealed relatively normal distributions. Linearity The assumption of linearity was tested by creating individual scatter plots of each outcome variable FCAT reading and FCAT math with each indep endent variable This analysis revealed the linear nature of these bivariate relationships Collinearity A term used to describe a perfect linear relationship between t w o explanatory variables is c ollinearity An increase in c ollinearity may cause an infl ation of the standard errors for the coefficients for the regression analysis. A scatter plot of the residuals was employed to check for c ollinearity among the explanatory variables yielding no indication of a linear or near linear relationship among the e xplanatory variables. Variables are measured without error The Florida Department of Education Alpha, ranging from .88 to .92 for the subtests of the FCAT reading and math achievements tests (FLDOE, 2008). The NRT reliability

PAGE 111

102 was evaluated using KR 20, a process to measure reliability for tests that contain dichotomous scored items (correct/incorre ct) The KR 20 for the NRT reading and math subtests ran ged from .88 .93. Because this wa s a secondary data analysis, it was not possible to conduct additional tests of the reliability of the score obtained for the measures of achievement The data retriev ed from the Pinellas County School District and the Florida Departme nt of Education were assumed to be accurate and recorded without error. Homoscedasticity of errors Homogeneity of v ariance The assumption o f homogeneity of variance / h omoscedasticity of the error terms was confirmed using Test This test employs a test of th e null hypothesis that there were no statistically significant difference s in the variance of the residuals of the dependent variable when regressed against each of the independent variables The results of the Square 21.79, p > .65. The results for FCAT math resulted in a Chi Squa re 32.52, p >.1436. According to Chen and colleagues (n.d.), these results were not statistically significant, resulting in a failure to reject the null and concluding that there was no violation of this statistical assumption. Independence The errors associated with one observation were not correlated with the errors of other observations. The data was not collected over time but rather were cross sectional in nature, and t herefore, it was concluded that this assumption w as not violated.

PAGE 112

103 The independent variables are fixed If repeated, a researcher would be expected to obtain the same values for the variable under investigation. That is, the values were not sampled or selected at random by the investigator. Influential Observations Several tests were conducted to examine potential outliers and influential observations to identify suspicious data points. In linear regression, outliers can be defined as observations w ith large residuals, which may indicate data error or some sampling peculiarity. The removal of these data points may improve the fit of the regression; they may also mask important information. Thus, regression diagnostics serve to point out possible erro rs in the observation or other substantive information that would Studentized Residuals, Leverage, and DFITS to test for outliers and influential observations. To test for ou Belsley (1988) a cutoff of 4/( n k 1) was utilized where k was the number of independent variables and n was the number of observations Applying the criteria suggested by Belsley nine observations ranging from .026 to .177 exceed the suggested value of .024. Consequently these observations were considered for removal from the study. To further test for outliers and unusual observations, the researcher considered studentized resi duals that exceeded two. This yielded eight observations (derived from the set of individual independent variables) with studentized residuals ranging from 2.02 to 3.36 with higher residuals indicating more cause for consideration as potentially unusual or erroneous observations ( Chen, Ender, Mitchell, Eyben, & W ells, n.d.) These

PAGE 113

104 observations were considered for removal from the study as potential erroneous observations. Leverage Leverage is another way to think about outlie rs and potentially influential observations These points were of possible concern due to large deviations from the mean and the effect on the estimated regression model. Points that exceed (2k+2)/ n where k was the number of independent variables and n wa s the number of observations, were considered to have considerable leverage (Chen et al., n.d .). For this study, six observations were observed to evidence values ranging from .09 to .31 It was important to note that the test for outliers using the studen tized residuals and the test for leverage did not share common observations with each other. The last procedure, DFITS, provided a combined estimate of the leverage and the Studentized residual, and in this study, used the conventional cutoff of 2 The DFITS analysis yielded 14 observations that met the cutoff threshold. These 14 observations had a DFITS range from .404 to 1.200 Summary of analysis of potential outliers and influence diagnostics The leverage, .31, suggesting this observation may have been influential. Thus an additional test for influenti al observations, DFITS, was conducted using the conventional cutoff of 2 The DFITS analysis yielded 14 observations that met the cutoff threshold. These 1 4 observations had a DFITS ranging from .404 to 1.200. Because the observation and largest leverage, this observation was given further consideration as a potentially

PAGE 114

105 unusual or erroneous observation. Upon further inspection, the observation with t he was a female student who had missed approximately half the school year with a high FCAT reading score, but low GPA, possibly due to 97 reported absences. The correlation and multiple regression analysis for FCAT reading and math were conducted both with and without this observation with no substantial change in the results. There were no indications that the observation in question was erroneous and therefore, she was not removed from the study. As a result, there were 191 students that met the criteria for the study, not all the students participated in the FCAT reading and math, or the NRT. Four stude nts failed to take the FCAT and 16 s tudents failed to take the NRT. Excluding students failing to take the FCAT and NRT left a total sample of 172 eligible students. This study met the minimum sample size required to ensure adequate statistical power, according to Stevens (2007), who suggested that a sample size less than 20 Primer suggested a sample size of N = 102 for a medium effect size with power = .80 and six independent variables. Additionally, a power a nalysis for a n = 172, and k = 7 yielded power = .90. Hence the sample in this study exceeded the minimum sample size recommendations for a correlational study.

PAGE 115

106 Correlation between FCAT Reading and the Dependent Variables In a Pearson Product Moment Correlation Coefficient matrix, the strength and direction of the relationship between pairs of variables is represented. As evidenced in Table 7, the strongest relationship was observed between FCAT reading and NRT r eading ( r = .77, p .0 0 01). A statistically significant relationship was observed between FCAT reading GPA ( r = .56, p .0 0 01), student a bsences ( .26, p .001), meal SES ( r = .18, p .02) and school SES ( r = .14, p .06). The only relationships that were not statistically significant were the relationships between FCAT reading and s chool SES ( r = .14, p .06) and g ender ( r = .07, p .33) therefore, these variables were excluded from the multiple regression analysis for FCAT reading

PAGE 116

107 Table 7 C orrelations Matrix: FCAT Reading and Math and Measure s of School and Student SES 1 2 3 4 5 6 7 8 9 Students ( n = 172) 1. FCATRead 2. FCATMath 0.67 .0001 3. SchoolSES 0.14 0.06 0.15 0.05 4. MealSES 0.18 0.02 0.02 0.84 0.20 0.01 5. Absences 0.26 0.001 0.29 .0001 0.08 0.29 0.04 0.64 6. Gender 0.07 0.33 0.12 0.12 0.03 0.69 0.03 0.67 0.10 0.20 7. GPA 0.56 .0001 0.58 .0001 0.14 0.07 0.07 0.40 0.47 .0001 0.08 0.38 8. NRTRead 0.77 .0001 0.64 .0001 0.06 0.47 0.10 0.21 0.23 0.00 0.02 0.81 0.51 .0001 9. NRTMath 0.60 .0001 0.68 .0001 0.09 0.23 0.01 0.87 0.15 0.05 0.16 0.04 0.42 .0001 0.67 .0001 Note. FCATRead = FCAT reading score, FCATMath = FCAT math score, SchoolSES = a composite definition that included student percentage of participation in the subsidized meal program and the percentage of African American and Hispanic students per school, MealSES = student participation in the subsidized meal program. Absences = student absences, GPA = Grade Point Average, NRTRead = FCAT NRT reading, NRTMath = FCAT NRT math. Correlation between FCAT Math, the Dependent and Independent Variables In a Pearson Product Moment Correlation Coefficient matrix, the strength and direction of the relationship between pairs of variables is represented. As evidenced in Table 7, the strongest relationship was observed between FCAT math and NRT math ( r = .68, p .0001). A statistically significant relationship was observed between GPA ( r =

PAGE 117

108 .58, p .0001), Absences ( .29, p .0001) and school SES ( r = .15, p .05) and FCAT math The only independent variables that were non significant were meal SES ( r = .02, p .84 and gender ( r = .12, p .12). These two were eliminated from the Multiple Regression analysis for FCAT math Summary : T he correlation coefficients indicated a statistically significant relationship between the FCAT r eading and four of the independent variables NRT r eading, GPA, student a bsences, and m eal SES. There was no statistically significant relationship between FCAT r eading and s chool SES or g ender ; therefore, these two variables were excluded from the multiple regression analyses for FCAT reading T he correlation coefficients indicated a statistically significant relationship between the FCAT math and four of the independent variables, NRT reading GPA, student a bsences, and school SES There was no statistically significant relationship between FCAT math and meal SES or gender ; therefore, these two variables were excluded from the multiple regression analyses f or FCAT math FCAT Reading Regression Analysis Regressing the FCAT reading scores on the linear combination of meal SES, absences, GPA, and NRT reading accounted for approximately 64% of the variance (R 2 =.64) in FCAT reading F (167) 73.34, ( p < .0001). Not surprisingly, the standardized regression coefficien NRT reading (.64) followed by GPA (.22), and meal SES (.12). This regression weight ( ) described the relative importance of the independent variable. For example, we would

PAGE 118

109 expect to see more than half a standard deviation increase in FCAT reading for every one standard d eviation increase in NRT reading Hence a one standard deviation increase in NRT reading led to a .64 standard deviation increase in FCAT reading scores holding the other independent variables constant. A one standard deviation increase in GPA led to a .2 2 standard deviation increase in FCAT reading scores holding the other independent variables constant. A one standard deviation increase in meal SES led to a .11 standard deviation increase in FCAT reading scores holding the other independent variables con stant Student absences was not a statistically significant predictor of the achievement for FCAT reading. A summary of the regression analysis for FCAT reading scores is presented in Table 8 Multiple Regression Formula for FCAT Reading Scores FCAT rea ding = 358 + .64(NRT reading ) + .22(GPA) + .11( meal SES ) Table 8 Summary Regression Analysis for Variables Explaining FCAT Reading Scores ( N = 172) Variable B SE B t Value Pr>/ t / Intercept 358 NRT Reading .84 .07 11.82 <.0001 .64 GPA 14.57 3.94 3.70 .0003 .22 Meal SES 12.56 5.43 2.31 .0220 .11 Absences .02 .24 .09 .9288 .01 Note. R 2 = .64, F Value = 73.34 ( p <.0001). FCAT Math Regression Analysis Regressing the FCAT math scores on the linear combination of school SES absences, GPA, and NRT math accounted for over 57% of the variance ( R 2 =.57) in FCAT math F (167) 54.84, ( p < .0001). Not surprisingly, the standardized regression coefficien t,

PAGE 119

110 math (.53 ) followed by GPA (.32 ). This regression weight, described the relative importance of each of the independent variable s. For examples, we would expect to see more than half a standard deviation increase in FCAT math for every one standard deviation increase in NRT math Hence a one standard deviation increase in NRT math led to a .53 standard deviation increase in FCAT math scores holding the other independent variables constant. A one standard deviation increase in GPA led to a .32 standard deviation increase in FCAT math score s holding the other independent variables constant. Student absences and school SES were not statistically significant predictor s of the achievement for FCAT math A summary of the multiple regression analysis for FCAT reading scores is presented in Table 9 Multiple Regression Formula for FCAT Math Scores FCAT math = 330 + .53(NRT math) + .32(GPA) Table 9 Summary Regression Analysis for Variables Explaining FCAT Math Scores ( N = 17 2 ) Variable B SE B t Value Pr>/ t / Intercept 330 NRT Math .86 .09 9.50 <.0001 .53 GPA 17.88 3.48 5.13 <.0001 .32 School SES .73 .69 1.05 .2945 .06 Absences .20 .21 .96 .3388 .06 Note. R 2 = .57, F Value = 55.65 ( p <.00001).

PAGE 120

111 Analysis of Research Question and Hypotheses for FCAT Reading and Math FCAT Reading and Math Question If there was a positive relationship between school SES and FCAT reading and math achievement, then, at what level of school SES did Hispanic LEP sophomores start to exhibit s ignificantly higher scores in reading and math achievement as defined by the FCAT reading and math achievement, after controlling for the effects of the meal SES, school SES student absences, GPA, and NRT reading and math? Reading r esults The relationship between school SES and FCAT reading achievement was not statistically significant ( r =.14, p = .06). Consequently school SES was excluded from the multiple regression analysis. Math r esults The multipl school SES was not statistically significant in exp laining FCAT math achievement t ( 1 ) 1.05, p > .2945 after controlling for student absences, GPA and NRT math. Hypothesis 1 Reading and Math There will be a negative relationship between enrollment in the subsidized meal program and reading and math achievement scores after controlling for the effects of school SES, student absences, GPA, and NRT reading and math. Reading r esults The multiple regression analysis indicated that a controlling for student absences, GPA, and NRT reading and math, the subsidized meal program was statistically significant t ( 1 ) 2.3 1 p >.0220.

PAGE 121

112 Math r esults The relationship between the subsidized meal program and FCAT math was not statistically significant ( r =.02, p =.84). Consequently the subsidized meal program variable was excluded from the multiple regression analysis. Hypothesis 2 Read ing and Math There will be a relationship between gender of Hispanic LEP students and academic reading and math achievement scores after controlling for the effects of school SES, the subsidized meal program, student absences, GPA, and NRT reading and mat h such that among Hispanic LEP students, females will have higher FCAT reading and math scores than males. Reading r esults The relationship between gender and FCAT reading was not statistically significant ( r =.07, p = .33). Consequently gender was excluded from the multiple regression analysis. Math r esults The relationship between gender and FCAT math achievement was not statistically significant ( r =.12, p = .12). Consequently gender was excluded from the multiple regression analysis as statistically insignificant at explaining FCAT math achievement. Hypothesis 3 Reading and Math There will be a negative relationship between student absences and reading and math achievement scores after controlling for the effects of school SES, the subsidized meal program, GPA, and NRT reading and math.

PAGE 122

113 Reading r esults controlling for the s ubsidized meal program, GPA, and NRT reading, student absences was not statistically significant t ( 1 ) .09, p >.93 Math r esults =.05, and controlling for school SES, GPA, and NRT math, student absences was not statistically significant t ( 3 ) 0.96, p >.34. Hypothesis 4 Reading and Math There will be a positive relationship between GPA and reading and math achievement scores after controlling for the effects of school SES, the subsidized meal program, student absences, and NRT reading and math. Reading r esults controlling for the eff ects of the subsidized meal program, NRT reading, GPA was statistically significant at explaining FCAT reading achievement t ( 1 ) 1.05, p >.0003. Math r esults co ntrolling for the effects of school SES, student absences, and NRT math, GPA was statistically significant at explaining FCAT math achievement t ( 1 ) 5.13, p > 0001. Hypothesis 5 Reading and Math There will be a positive relationship between NRT reading and math scores after controlling for the effects of school SES, the subsidized meal program, student absences and GPA. Reading r esults The multiple regression controlling for the effects of the subsidized meal program, student absences, and GPA,

PAGE 123

114 NRT reading was statistically significant at explaining FCAT reading achievement t ( 1 ) 11.82, p >.0001). Math r esults controlling for the effects of school SES, student absences and GPA, NRT math was statistically significant at explaining FCAT reading achievement t ( 1 ) 9.5, p >.0001. Summ ary Chapter 4 examined the relationship between measures of school and student socioeconomic status and FCAT reading and math scores for Hispanic LEP high school sophomores in Pinellas County, Florida for the 2007 2008 school year. S chool SES meal SES, s tudent absences, gender, GPA, and NRT reading were analyzed to determine their relationship to FCAT reading using a correlational study employing the correlational and multiple regression analyse s. A Pearson Product Moment Correlation Coefficient was used to examine the strength and direction among the measures of achievement and indicators of school and student socioeconomic status. To examine these relationships, one research question and five research hypotheses were analyzed for each achievement variable. The results indicated that there were statistical ly significan t relationships between FCAT reading NRT reading, GPA and meal SES after controlling for student absences. This means that mos t of the variance in FCAT reading scores (64%) is accounted for by the scores, GPA and meal SES in that order. For FCAT math, the multiple regression results indicated the variance (57%) was accounted for by NRT math and GPA

PAGE 124

115 after controlling for the effects of s chool SES, and student absences. Collectively, the analyses indicated t hat there was no meaningful or statistically significant relationship between student absences for either reading or math Chapter 5 of this study addressed the fin dings and explored related issues. It also addressed the limitations associated with this study and made recommendation for future studies.

PAGE 125

116 Chapter 5 Summary and Discussion Overview The final chapter in this quantitative research study provide d a summary of the study and the findings The introduction discusse d the problem and purpose of this study to include the variables and sample, and research design employed in this study. This is followed by a summary of the research questions, hypothes e s, and their respective analysis. Finally implications and explanations of the findings are discussed including the limitations associated with this study and recommendations for future researchers interes ted in this research topic. Introduction The purpose of this study was to examine the relationships between measures of student academic achievement and school SES as defined by the variable s of school SES, meal SES, student absences, gender, GPA, and NRT reading and math to FCAT reading and math scores for Hispanic limited English proficient (LEP) high school students. These variables were selected from the extensive literature regarding the academic achievement of minority students and indices of SES. Th e population in this study was Hispanic LEP sophomores in the 16 traditional high schools in Pinellas County, Florida.

PAGE 126

117 Hispanic LEP students are unique from other Hispanic students as LEP students speak a language other than English at home. In order to be eligible for LEP services, students must score at a designated level on the Language Assessment Battery Revised (LAB R) and New York State English as a Second Language Achievement Test (NYSESLAT) test measuring language fluency. Sophomore students were se lected because of Florida requirement that all sophomores take the FCAT. To assess the influence of school SES factors on Hispanic LEP students, students included in the study must have been registered as freshman in the previous year, 2006 2007, at the same school. Due to the emphasis placed on the FCAT and the No Child Left Behind Act of 2001 by school districts, many schools have focused their efforts on the reading and math achievement of small groups of students such as minority, economically disadv antaged, and LEP students. However, there remains an academic achievement gap between advantaged and disadvantaged students. An explanatory examination of the relationship between measures of SES and reading and math achievement of the FCAT may be benefici al for school districts to address policies such as student assignments and allocation of resources to the largest LEP population, Hispanics. Additionally, this study addressed the characteristics associated with SES and minority populations and their fit to the Hispanic LEP population. Lastly, this study addressed the lack of substantive research literature regarding the relationship between SES and academic achievement of Hispanic LEP students

PAGE 127

118 Discussion This study employed a correlational design mult iple regression analys i s to explore the relationship between school SES, meal SES, gender, student absences, NRT reading and math, and GPA to FCAT reading and math scores. Researchers in the field, such as Baer and Schmitz (2007), Carpenter, Ramires, and S evern (2006), Naglieri, Winsler, and Booth (2004), specializing in quantitative studies of minority issues in education have utilized correlational design multiple regression analyses to explore the relationships between academic achievement and measures o f SES. In this study, the Pearson Product Moment Correlation showed a statistically significant bivariate relationship s between FCAT reading scores and NRT reading, GPA, meal SES, and student absences only. Hence these variables were included in the FCAT reading regression analysis. The Pearson Product Moment Correlation also showed a statistically significant bivariate relationship s between FCAT math scores and NRT math, GPA, school SES, and student absences only. Consequently, these variables were includ ed in the FCAT math regression analysis. The linear combination of NRT reading scores, GPA, and meal SES were statistically significant in the FCAT reading regression model, FCAT reading = b 1 X 1 NRT reading + b 2 X 2 GPA + b 3 X 3 meal SES + b 4 X 4 student absences + a. Whereas, student absences were not statistically significant The linear combination of NRT math scores and GPA were statistically significant in the FCAT reading regression model, FCAT math = b 1 X 1 NRTmath + b 2 X 2

PAGE 128

119 GPA + b 3 X 3 school SES + b 4 X 4 student absences + a. Whereas, school SES and student absences were not statistically significant Discussion and Findings This study supported the hypothes e s that meal SES, GPA, and NRT reading significantly explained the variance for FCAT readin g. This study also supported the hypothes e s that GPA and NRT math significantly explained the variance for FCAT math. However, this study did not support the hypothes e s that student absences, gender, and school SES display ed a statistically significant rel ationship for both FCAT reading and math scores. Further, this study did not support the hypothesis that meal SES display ed a statistically significant relationship with FCAT math. The role of the individual predictor variables, school SES, meal SES, stude nt absences, and gender was unexpected based on the general literature on the relationship between SES and academic achievement of minority students. A possible explanation for the unexpected relationship s among student absences, gender, and school SES, a nd FCAT reading and math scores could be attributed to the familial culture of the Hispanic LEP student. Hoxby (2002), Ingersoll (2004), and Kahlenberg (2004) acknowledged that European American, middleclass culture moves education forward ; further, minori ty students show higher academic performance and attainment when attending middleclass schools. The literature describing the middleclass education concept suggests that middleclass W hite students have access to a higher quality education, more highly educ ated teachers, and more advanced curriculum than their counterparts and therefore perform higher on standardized measures (Hoxby, 2002;

PAGE 129

120 Ingersoll, 2004; Kahlenberg, 2000; Kahlenberg, 2002; Kahlenberg, 2004 ). However, there is little indication that the se same qualities, such as work ethic s and value of education found in the culture of middleclass education are not also found in the native culture of the Hispanic LEP student SES and student absences Accordi ng to the literature reviewed for this study, impoverished students attended disproportionately high minority, highly impoverished schools (Meier, et al., 2004; Reyes & Capper, 1991; Tillman, 2005; Viadero, 2000). Generally, u nderprivileged students exhibi ted lower GPA, scored lower on standardized tests, and experienced higher incidents of school related absences. Participation in the subsidized meal program is a common index for SES used by numerous educational researchers such as Burris, Heubert, and Lev in, 2006, Lewis and Cheng, 2006, and Martin, Karabel, and Jaquez, 2005. The No Child Left Behind Act of 2001 uses the criteria of the subsidized meal program as the indices for economically disadvantaged. Thus it was surprising that the findings in this s tudy show that school SES and student absences were not statistically significant in explaining FCAT reading and math scores. Also unusual was the relationship between meal SES and the academic performance of Hispanic LEP students. Although meal SES showed a statistical difference for explaining FCAT reading, the same was not true for FCAT math. For FCAT reading, the coefficients for meal SES were lower than suggested in the research literature (Burris, Heurbert, & Levin, 2006; Lewis & Cheng, 2006; Martin Karabel, & Jaquez, 2005). The literature placed great emphasis on the relationship between SES and academic achievement. Researchers such as Burris et al., (2006), Lewis and Cheng

PAGE 130

121 (2006), and Martin and colleagues (2005), used student participat ion in the subsidized meal program as an index for poverty. The findings in this study regarding SES and the academic achievement of Hispanic LEP students could be attributed to the operational definition of school SES (the percentage of African American + Hi spanic + students on the subsidized meal program) used in this study. There are reports that students eligible for the subsidized meal program do not apply for this benefit whereas, ineligible students apply for the subsidized meal program and receive the se benefits ( McLaughlin, 2009) Further, with the inability to function fully in the English language, it is possible that Hispanic LEP students, and their families, are not aware of the application processes to obtain a free or reduced price meal at their school. Hence it stands to reason that the subsidized meal program may not be the most accurate measure for SES for the Hispanic LEP population. Other levels of SES, such as parental educational level or parental profession, may prove better indices of S ES than the subsidized meal program. Additionally, if the school had a substantial percentage of situational impoverished Hispanic LEP students participating in the subsidized meal program, then participation in the subsidized meal program may not be the a ppropriate indicator of generational poverty (Payne, 1999). Thus, doubt was cast on the explanatory value of the subsidized meal program as an indicator of SES for Hispanic LEP students GPA and NRT For both FCAT reading an d math, GPA and NRT were the strongest variables in explaining FCAT scores. Both GPA and NRT indices were statistically coefficient showed that both variables were the most influential variable for explaining FCAT reading and math score of all the variables

PAGE 131

122 used in this study. The findings that GPA and standardized tests, such as the NRT, were strong predictors w ere suppor ted by the numerous references in the research literature such as Reyes and Capper ( 1991 ) and Sus, Houston, and Suh ( 2007 ) that confirmed statistically significant relationships among GPA, standard ized tests, such as the NRT, and minority students Gen der According to the literature, there should have been a difference in the FCAT scores between males and females (Andres, Adamuti Trache, Yoon, Pidgeon, & Thomsen, 2007; Sadker, 1999). Traditionally, female students have scored higher than have their male counterparts on standard ized reading tests, and males have scored higher on standardized math tests, such as the ACT. In 2008, female students scored an average of 21.5 on ACT reading whereas males scored an average of 21.2 on ACT reading. In math, males have a slight increase over females with an average score of 21.6 for males and 20.4 for females (Digest of Educational Statistics, 2008). The ACT is a standard test with scores ranging from 1 to 36. Likewise, according to the National Center for Educational Statistics (2009), female high school students outperformed male high school students on reading scale scores with females scoring an average of 292 and males scoring an average of 279 on a scale ranging from 0 to 500; in ma th, males scored 308, and females scored an average of 305. However, contrary to the literature, this study found that gender was not statistical ly significant at predicting FCAT reading and math. Perhaps this is explained by the efforts of test writers to reduce gender and cultural bias in standardized tests such as the FCAT, NRT, and ACT. Other factors, such as teacher training in second language acquisition or cultural ly bias free classroom s may have

PAGE 132

123 reduced gender bias in the FCAT test results. Finally the results of this study establish a strong relationship between the indices used for measure s of school and student SES and FCAT reading and math scores Relevance of the Findings The finding s of this study were not as expected taking into considerati on the research literature on the relationship between SES and academic achievement of minority students. These findings suggest that SES indices such as the subsidized meal program and school SES were not as strong explanatory indices for Hispanic LEP hig h school students as they might have been for minority students such as African Americans and non LEP Hispanics. Variables such as student absences failed to explain FCAT scores indicating that the relationship between measure s of school and student SES an d Hispanic LEP high school students appeared to be different from that of other minority students A possible explanation for these findings is that Hispanic LEP students may come from middleclass background s in their country of origin. The inference that Hispanic LEP students living in poverty have lower academic performance does not appear to hold true for the Hispanic LEP students in this study First impressions are not necessarily reality (Shaw, 2007). M iddleclass culture and work ethic s are found in c ountries and cultures beyond the U.S. The fact that a student is in an impoverished school or is a Hispanic LEP student does not necessarily predict that the student will have poor academic performance

PAGE 133

124 Implications of the Findings Since the time of this study, t he Pinellas County School district has eliminated NRT testing. With the strong explanatory power of the NRT it would benefit the school district to reemploy the NRT to help identify students who may require additional educational services to demonstrate FCAT competency in reading and math. This, in combination with GPA, could provide a strong predictor of FCAT success in reading and math. Therefore, school districts may want to give serious consider ation of reinstating the NRT. It was expected that GPA and standardized tests, such as the NRT, helped explain FCAT reading and math scores. This study confirmed that GPA and standard tests such as the NRT are strong indicators of academic success and help to identify students in need of remediation. Elimination of the NRT may cause school districts to look to other indicators such as GPA, which is only calculated at the end of the semester, or twice a year. Other alternatives such as math attainment as indicated by math rigor may be a better predicto r for performance on high stakes tests F uture researchers should consider including math attainment as a possible variable for academic achievement for minority populations, such as Hispanic LEP students School personnel should be cognizant of all explan atory indices for high stakes test s to indentify struggling students to provide appropriate instructions If districts have made the decision to drop NRT due to financial reasons as they work to balance budgets in a struggling economy, any savings should be wisely

PAGE 134

125 appropriated for assisting Hispanic LEP students. Zuniga, Olson and Wilson suggested : (1) recruiting and retaining of qualified ESOL teachers; (2) procuring of translators and other support personnel; (3) integrating Hispanic LEP students into r egular classrooms; and (4) tutoring for after school academic assistance would benefit struggling students (2003) Educational leaders should be cautious when using SES labels such as poor students, low SES students, or minority students, to describe the H ispanic LEP student, and indeed, all students. Rather, all educators should work to help Hispanic LEP students, and all students in need, to become productive and contributing members of the society. Outside of the educational community, agricultural indu stries dependent upon Hispanic migrant workers would also be interested in the reading and math skills of their employees An educated workforce can increas e the productivity decision making skills and knowledge of workers in general. This benefits both t he employer and the employee. An increase in reading and math skills could potentially increas e the productivity of Hispanic worker s making them a more valued employee and a more productive citizen. Limitations of the Study Several limitations to this study existed. Information such as race/ethnicity, gender, enrollment in the subsidized meal program, GPA, FCAT scores, NRT scores, and student absences were accepted without verification for accuracy from the school tabase. Therefore, generalization of this study was limited to the accuracy of The population and sample

PAGE 135

126 was limited by the term used to describe their ethnicity, Hispanic, as there is great dive rsity in the Hispanic population. This study was unable to distinguish between nationalities among the Hispanic population. Hispanics such as Argentineans, Cubans, and Mexicans were categorized as one ethic population. Further, the indicator of the subsidi zed meal program limited the ability to generalize the results regarding SES The report by McLaughlin (2009) suggested that ineligible students participated in the subsidized meal program and eligible students especially in high school s were underreport ed McLaughlin revealed that over half of the families participating in the subsidized meal program were ineligible to receive the benefits of the subsidized meal program Therefore, the ability to generalize SES was limited to students participating in th e subsidized meal program. Lastly, a multiple regression analysis examines relationships between an independent variable and multiple independent variables. The results of a multiple regression cannot be interpreted as causation Multiple regression analy sis are corelational in nature and should not be interpreted to represent casual associations among the variables. Recommendations for Further Research As expected, research yields more questions than answers. A similar study with similar populations in d ifferent school districts may help confirm the findings that measures of SES, student absences, and gender may not be valid predictors of academic success for Hispanic LEP high school students as they are for other minority populations.

PAGE 136

127 Thereby, strengthen ing the conclusions derived from this study that measures of school and student SES need to be more closely scrutinized as predictors of academic success on standardized tests, such as the FCAT, for Hispanic LEP high school students. Educators, counselors and administrators i n school districts with LEP populations, other than Hispanics, may be interested in replicating this study on other sub LEP populations, such as Asians. Often, Asians are excluded from research populations due to too small a sample si ze (Attewell & Domina 2008). Additionally, Asians are also included in the European population as noted by Schiller and Muller (2003). The study of other LEP populations may provide school districts with possible explanatory variables and proactive indice s for high stakes testing. A qualitative study may prove fruitful in unveiling the cultural context that differentiates the Hispanic LEP students from other minority groups. Each ethnic population brings forward its own unique cultural context and richness to the school and community thereby establishing its own unique consideration of academic instruction and other educational services. Additionally, similar studies on different Hispanic nationalities may also prove enlightening in helping the educational community find strategies an d set policies to help foreign born and LEP students succeed academically. The literature suggested that impoverished and minority students disproportionately were taught by less qualified teachers (Alexander & Salmon, 2007; Brennan & Bliss, 1998; Ingersoll, 2004; Jacob, 2007; Meier et al., 2004). The No Child Left Behind Act, 2001, addressed the need to have highly qualified teachers. The literature indicated that middleclass schools had a higher percentage of teachers with

PAGE 137

128 advanced d egrees than did less affluent schools. Teachers with advanced degrees was one of the key indicators alluded to in the literature as one of the possible factors for the increased academic gains of students in middleclass schools (Ingersoll, 2004; Jacob, 200 7; McElroy, 2007). The literature was clear that impoverished, high minority schools had less qualified and fewer teachers with advanced degrees (Brennan & Bliss, 1998; Ingersoll, 2004; Jacob, 2007; Johnson, 2007). This study was not able to assess the per centage of Hispanic LEP students that had teachers with advance d degrees. Further studies should investigate the relationship between teacher level characteristics such as training and certification, and Hispanic LEP reading and math achievement at the stu dent level. The role of teacher training and certification has been addressed by NCLB in the call for highly qualified teachers. Future studies could look at a comparison of teachers with and without advanced degrees teaching LEP students. Similar studies might include teacher experience, teachers teaching out of field, ethnicity and minority status of educators teaching LEP students to see if these variables contribute to the academic performance of Hispanic LEP students as suggested by Ingersoll (2004) and Kahlenberg (2000) Such future studies could help explain the role between teaching credentials and reading and math achievement for Hispanic LEP high school students. Lastly, future studies should include an expanded definition of SES that increases the ability to generalize the results of future studies to other populations. Future studies should consider other indices for SES such as housing value, family income, parental educational attainment, parental occupation, and parental employment status. E xpanding

PAGE 138

129 the definition of SES to include other indices will increase the ability to generalize the results of future studies to other populations Conclusion It can be argued that the subject of education has long been in the forefront of the American consciousness. Education has also been the talk on the political landscape as cited in and directed by the No Child Left Behind Act of 2001 to improve the education for all students. This study is but one of a relatively small groups of studies that attemp ted to address the relationship between measures of SES and academic achievement for a specific subgroup of minority students, the Hispanic LEP student. The results from this study suggest that when viewed through the middleclass education concept, Hispani c LEP students could be considered a unique population from other minority students. If past performance is an indicator of future performance, then school districts must find appropriate variables to help predict academic success to assist all struggling students. The middleclass concept that poor school + poor student = poor academic performance may not necessarily hold true for Hispanic LEP students. The message is not to judge a book, or in the case of this study a student, by its cover. The broad brus h of poverty does not necessarily accurately paint the current portrait of the Hispanic LEP student.

PAGE 139

130 References Abedi, J. (2004). The No Child Left Behind Act and English language learners: Assessment and accountability issues. Educational Researcher, 33 4 15. Abedi, J., Hofstetter, C. H., & Lord, C. (2004). Assessment accommodations for English language learners: Implications for policy based empirical research. Review of Educational Research, 74 (1), 1 28. Alexander, K., & Salmon, R. (2007). Warranting failure: The "system" that breeds poverty and starves public schools. Journal of Education Finance, 33 (2), 203 230. American Community Survey (2003). Retrieved Spetember 15, 2008, from U.S. Census Bureau: http://americancpm munitysurvey.census.gov/home/saff/main.html?_lang= en American FactFinder (2006). Retrieved September 15, 2008, from U.S. Census Bureau: http://factfinder.census.gov/home/saff/main.html?_lang=en Andres, L., Adamuti Trache, M., Yoon, E., Pidgeon, M., & Thom sen, J. P. (2007). Educational expectations, parental social class, gender, and postsecondary attainment: A 10 year perspective. Youth and Society, 39 (2), 135 163. Arbona, C., & Nora, A. (2007). The influence of academic and environmental factors on Hispa nic college degree attainment. The Review of Higher Education, 30 (3), 247 269.

PAGE 140

131 Armor, K. (2006, August 22). Can NCLB close achievement gaps? Retrieved March 30, 2008, from Teachers College Record: http://www.tcrecord.org Attewell, P., & Domina, T. (2008). Raising the bar: Curricular intensity and academic performance. Educational Evaluation and Policy Analysis, 30 (1), 51 71. Babbie, E. (1998). The proactive of social research. Belmont, CA.: Wadsworth. Baer, J. C., & Schmitz, M. F. (2007). Ethnic differenc es in trajectories of family cohesion for Mexican American and non Hispanic white adolescents. J Youth Adolescence 583 592. Barlow, D. R. (1990). Consequences of growing up poor (book review). The Education Digest, 64 (7), 77 81. Barth, R. S. (1990). Impr oving schools from within: Teachers, parents, and principals can make the difference. San Francisco, CA, US: Jossey Bass. Battle, J., & Pastrana, A. (2007). The relative importance of race and socioeconomic status among Hispanic and white students. Hispani c Journal of Behavioral Sciences, 29 (1), 35 49. Becker, G. S. (1964). Human capital: A theoretical and empirical analysis, with special reference to education. New York, NY., U.S.: National Bureau of Economic Research. Belsley, D. A. (1988). Conditioning in models with logs. Journal of Econometrics 38, 127 143.

PAGE 141

132 Brennan, S., & Bliss, T. (1998). Increasing minority representation in the teaching profession through alternative certification: A case study. The Teacher Educator 34 (1), 1 6. Burris, C. C., Heubert, J. P., & Levin, H. M. (2006). Accelerating mathematics achievement using heterogeneous grouping. American Educational Research Journal, 43 (1), 105 136. Carpenter, D. M., Ramires, A., & Severn, L. (2006). Gap or gaps: Challenging the singular defi nition of the achievement gap. Educational and Urban Society, 39 (1), 113 127. Chen, G. (2008). Communities, students, schools, and school crime in U.S. high schools. Urban Education, 43 (3), 301 318. Chen, X., Ender, P., Mitchell, M., Eyben, R., & W ells, C. ( n.d. ). SAS Web Books. Retrieved June 17, 2009, from Regression with SAS: http://faculty.mdc.edu/mshakil/SAS%20Web%20Books%20Regression% 20with%20SAS.htm Coatsworth, J. D., Duncan, L. G., Pantin, H., & Szapocznik, J. (2006). Differential predictors of Af rican American and Hispanic parent retention in a family focused preventive intervention. Family Relations, 55 240 251. Cohen, J. (1988). Statistical power analysis for the behavioral sciences Hillsdale, N.J.: L. Erlbaum Associates. Corporate Author. (20 07 2008). Florida School Grades Retrieved March 19, 2008, from Florida Department of Education: http:schoolgrades.fldoe.org

PAGE 142

133 Corporate Author. (2001). Florida Department of Education Retrieved April 15, 2009, from School Accountability Reports: http://www .tcmstornadoes.com/0809AYPReportCard.htm Corporate Author. (1994). Minority youth in the juvenile justice system: A judicial response. Juvenile & Family Court Journal, 41 (3A), 1 16. Drummond, K., & Stipek, D. (2004). Low income parents' beliefs about thei r role in children's academic learning. The Elementary School Journal, 104 (3), 197 202. Duncan, G. (1994). Families and neighbors as sources of disadvantage in schooling decisions of white and black adolescents. The Elementary School Journal, 104 (3), 197 202. Fletcher, M. A. (2002, March 18). Diversity's future? Socioeconomic criteria, not race, to desegregate San Francisco schools. Washington Post p. 02. Flowers, T A. (2008). Factors affecting urban A frican A merican high school students. Urban education 43(2), 154 71. Freisthler, B., Bruce, E., & Needell, B. (2007). Understanding the geospatial relationship of neighborhood characteristics and rates of maltreatment for Black, Hispanic, and White children. Social Work, 52 (1), 7 16. Fry, R. (2005). The hi gh schools Hispanics attend: Size and other key characteristics. Pew Hispanic Center i ii. Graziano, J., Slavkovich, V., Liu, X., Factor Litvak, P., & Todd, A. (2004). A prospective study of prenatal and childhood lead exposure and

PAGE 143

134 erythropoietin producti on. Journal of Occupational and Environmental Medicine, 46 924 929. Guglielmi, R. S. (2008). Native language Proficiency, English literacy, academic achievement, and occupational attainment in limited English proficient students: A latent growth modeling perspective. Journal of Educational Psychology, 100 (2), 322 342. Hanushek, E., Kain, J., & Rivkin, S. (2004). New evidence about Brown v. Board of Education: The complex effects of racial composition on achievement. Retrieved April 01, 2008, from Institut e for Research on Poverty: http://www.ssc.wisc.edu/irp/ Hoxby, C. M. (2002). The power of peers. Education Next, 2 (2), 57 63. Hoy, W. K., Tarter, S. J., & Hoy, A. W. (2006). Academic optimism of schools: A force for students achievement. American Educational Research Journal, 43 (3), 425 446. Ingersoll, R. M. (2004, November). Why do high poverty schools have difficulty staffing their classrooms with qualified teachers? Retrieved February 09, 2008, from Renewing Our Schools, Securing Our Future: ht tp://www.americanprogress.org Jacob, B. A. (2007). The challenges of staffing urban schools with effective teachers. The Future of Children, 17 (1), 129 153. Johannessen, L. (2003). Achieving success for the "resistant" student. The Clearing House, 77 (1), 6 13.

PAGE 144

135 Johnson, D. P. (2007). Challenges to No Child Left Behind: Title 1 and Hispanic students locked away for later. Education and Urban Society, 39 (3), 382 398. Kahlenberg, R. D. (2004, April 23). Can separate be equal? The overlooked flaw at the cente r of No Child Left Behind. Reality Check 1 14. Kahlenberg, R. D. (2002). Economic school intergration: New ideas for a new century. 2 Kahlenberg, R. (2000). Economic school integration. Retrieved September 10, 2008, from The New Century Foundation New Id ea for a New Century: http://www.policyideas.org Kennedy, G., Green, G. K., Huerta, L. A., & Richards, A. (2007). Getting real: A different perspective on the relationship between school resources and student outcomes. Journal of Education Finance 33 (1), 49 68. Kromrey, J., Onwuegbuzie, A., & Hogarty, K. (2006). The continua of discplined inquiry: Quantitative, qualitative, and mexed methods. In S. Permuth, & R. Mawdsley, Research methods for studying legal issues in education (pp. 91 129). Dayton, OH: Ed ucation Law Association. Lewis, T., & Cheng, S. (2006). Tracking, expectations, and the transformation of vocational education. American Journal of Education, 113 67 98. Livingston, A. (2008). The Condition of Education 2008 in Brief (NCES 2008 032).Natio nal Center for Education Statistics, Institute of Education Sciences, U.S. Department of Education. Washington, DC. Love, D. (2007). From the cradle to jail in America. Retrieved April 31, 2008, from Common Dreams News Center: http://www.commondreams.org

PAGE 145

136 L ULAC et al. v. Florida Dept. of Education et al., 90 1913 Civ. Scott (S.D. Fla., Miami Div. Aug. 14, 1990). Martin, I., Karabel, J., & Jaquez, S. (2005). High school segregation and access to the University of California. Educational Policy, 19 (2), 308 330. Meier, D., Kohn, A., Darling Hammond, L., Sizer, T. R., & Wood, G. (2004). Many children left behind: How the No Child Left Behind Act is damaging our children and our schools. Boston, MA, US: Beacon Press. McLaughlin, B. (2009, April 01). T oo many free lunches for those who can pay? Retrieved April 11, 09, from ABC Action News: http://www.abcactionnews.com/news/local/story/Too many free lunches for those who can pay/Ps6NQV6JK0ujRCknqrr0Rg.cspx Naglieri, J. A., Rojahn, J., & Matto, H. C. (200 6). Hispanic and non Hispanic children's performance on PASS cognitive processes and achievement. Center for Cognitive Development 568 579. Naglieri, J. A., Winsler, A., & Booth, A. L. (2004). Comparison of Hispanic children with and without limited Engli sh proficiency on the Naglieri Nonverbal Ability Test. Psychological Assessment, 16 (1), 81 84. Nichols, J. D. (2003). Prediction indicator for students failing the state of Indiana high school graduation exam. Preventing School Failure 47 (3), 112 120. No Child Left Behind Act of 2001 (2002), P.L. 107 110.

PAGE 146

137 Osborne, J. W., & Waters, E. (2002). Four assumptions of multiple regressions that researchers should always test. Retrieved April 12, 2009, from Practical Assessment, Research & Evaluation: http://par eonline.net Permuth, S., & Mawdsley, R. D. (2006). Research methods for studying legal issues in education. Dayton, OH, U.S.: Education Law Associates. Perna, L. W. (2005). The benefits of higher education: Sex, racial/ethnic, and socioeconomic group differences. The Review of Higher Education, 29 (1), 23 52. Peskin, M. F., Tortolero, S. R., & Markham, C. M. (2006). Bullying and victimization among Black and Hispanic adolescents. Adolescence, 41 (163), 467 484. Raver, C. C., Aber, J. L., & Gershoff, E. T. (2007). Testing equivalence of mediating model of income, parenting, and school readiness for White, Black, and Hispanic children in a national sample. C hild Development, 78 (1), 96 115. Reyes, P., & Capper, C. A. (1991). Urban principal: A critical perspective on the context of minority student dropout. Educational Administration Quarterly 530 557. Rumberger, R. W., & Palardy, G. J. (2005). Does segregat ion still matter? The impact of student composition on academic achievement in high school. Teachers College Record 107 (9), 1990 2045. Russo, P. (2004, February). OSWEGO State University of New York. Retrieved September 10, 2008, from OSWEGO: http://www .oswego.edu

PAGE 147

138 Schiller, K. S. (2003). Raising the bar and equity? Effects of state high school graduation requirements and accountability policies on student' mathematics course taking. Educational Evaluation and Policy Analysis, 25 (3), 299 318. Shaw, R. (2 007) motivating them to learn Routledge. Solochek, S. S. (2008, August 8). The gradebook: Your daily report on educational news. Make the FCAT harder p. Blog. Stein, S. J. (2004). The culture of education policy. New York, NY., U.S.: Teachers College Press. Stevens, J. P. (2007). Intermidiate statistics: A modern approach (3rd ed.). New York, NY., US.: Lawrence Erlbaum Associates. Stevens, T. J., Olivarez, A., & Hammon, D. (2006). The role of cognition, motivation, and emotion in explaining the mathematics achievement gap between Hispanic and White students. Hispanic Journal of Behavioral Sciences, 28 (2), 161 186. Sus, S., Houston, J., & Suh, J. (2007). Predictors of categorical at ris k high school dropouts. Journal of Counseling & Development, 85 196 203. Thornton, B., Collins, M., & Daugherty, R. (2006). A study of resiliency of American Indian school students. Journal of American Indian Education, 45 (1), 4 16. Tillman, L. C. (2005) Mentoring new teachers: Implications for leadership practice in an urban school. Educational Administration Quarterly, 41 (4), 609 629.

PAGE 148

139 U.S. Census (n.d.). Retrieved March 25, 2008, from http://quickfacts.census.gov U.S. Census Bureau. (2007). Retrieved April 1, 2008, from American Fact Finder: http://factfinder.census.gov Unger, J. B., Ritt Olson, A., Wagner, K., Soto, D., & Baezconde Garbanati, L. (2007). A comparison of acculturation measures among Hispanic/Latino adolescents. J Youth Adolescence 555 565. Viadero, D. (2000). Lags in minority achievement defy traditional explanations. Education Week, 19 (28). Warren, J. R., & Edwards, M. R. (2005). High school exit examination and high school completion: Evidence from the Early 1990s. Educational Evalu ation and Policy Analysis, 27 (1), 53 74. Warren, J. R., & Jenkins K. N. (2005). High school exit examination and high school dropout in Texas and Florida, 1971 2000. Sociology of Education, 78 122 143. Weaver, C. N. (2005). The changing image of Hispani c Americans. Hispanic Journal of Behavioral Science, 27 (3), 337 354. Weaver, C. N. (2007). The effects of contact on the prejudice between Hispanics and non Hispanic Whites in the United States. Hispanic Journal of Behavioral Science, 29 (2), 254 274. Weisberg, S. (1985). Applied linear regression Wiley series in probability and mathematical statistics. New York: Wiley.

PAGE 149

140 Wilkins, A., & Staff, E. T. (2006). Yes we can: Telling the truth School and dispelling myth about race and education in America. Retrieved March 28, 2008, from The Education Trust: http://www.edtrust.org Willson, V. L., & Hughes, J. N. (2006). Retention of Hispanic/Latino students in first grade: Child, parent, teacher, school, and peer predictors. Journal of School Psychology 31 49. Zeichner, K. (2006). Reflections of university based teacher educator on the future of college and university based teacher education. Journal of Teacher Education, 57 (3), 326 340. Zuniga, K., Olson, J. K., & Winter, M. (2005). Science education for rural Latino/a Journal of Research in Science Teaching, 42 (4), 376 402.

PAGE 150

About the Author from the University of South Florida. He taught elementary, middle, and high school before returning to USF to complete a Masters Degree in Educational Leadership. Mr. Piedra had been an assistant principal in both the middle and high school levels servin g students, parents, and teachers in grades six through twelve.


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 2200397Ka 4500
controlfield tag 001 002068373
005 20101007133020.0
007 cr bnu|||uuuuu
008 100413s2009 flu s 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0003138
035
(OCoLC)606875661
040
FHM
c FHM
049
FHMM
090
LB2805 (Online)
1 100
Piedra, Osvaldo.
3 245
An explanatory examination of relationships between measures of school and student socioeconomic status and reading and math achievement of Hispanic limited English proficient (LEP) high school students
h [electronic resource] /
by Osvaldo Piedra.
260
[Tampa, Fla] :
b University of South Florida,
2009.
500
Title from PDF of title page.
Document formatted into pages; contains 140 pages.
Includes vita.
502
Dissertation (Ed.D.)--University of South Florida, 2009.
504
Includes bibliographical references.
516
Text (Electronic thesis) in PDF format.
520
ABSTRACT: Hispanics are the largest and fastest growing minority population in the U.S. requiring limited English proficiency services (Guglielmi, 2008). Many schools and school districts place heavy emphasis on high stakes testing. The No Child Left Behind (NCLB) Act of 2001 requires all children to have access to a fair, equal, and high-quality education. Many schools and school districts focus on individual students to show grade level performance for small populations such as limited English proficient (LEP) students. The purpose of this study was to examine the question, "What was the relationship between measures of school and student socioeconomic status, and reading and math achievement of Hispanic LEP high school students?" Guiding this study was the conceptual framework, the middleclass education concept.This concept suggested that underprivileged students attending an impoverished school had lower academic success than did underprivileged students attending a middleclass school with middleclass values. This study was developed from secondary data analysis of archived data from the Pinellas County School Board database. Pearson correlation coefficients were computed between the variables in this study. The correlation analysis was followed by a multiple regression analysis to estimate the capacity of the subsidized meal program, student absences, gender, grade point average (GPA), and the Norm Referenced Test (NRT) reading and math to explain reading and math achievement, as defined by the Florida Comprehensive Assessment Test in reading and math.The results of this study provided insight that the role played by indices of school and student SES may be different for Hispanic LEP students than for other minority students, such as African American and Hispanic or Latino high school students. The results showed that measures of SES were not as strong as expected based on the current research literature. GPA and NRT provided the strongest explanation of variability for FCAT reading and math scores. Other indicators such as student absences and gender were not statistically significant for explaining the FCAT reading and math scores. Thereby, indicating that the conceptual framework, the Middleclass Education Concept, may not apply to Hispanic LEP high school students and thus, further research for this population is needed.
538
Mode of access: World Wide Web.
System requirements: World Wide Web browser and PDF reader.
590
Advisor: Darlene Bruner, Ed.D.
653
SES
Minority education
High stakes testing
Middleclass education concept
Academic achievement
0 690
Dissertations, Academic
z USF
x Educational Leadership
Doctoral.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.3138