|USFDC Home | USF Electronic Theses and Dissertations||| RSS|
This item is only available as the following downloads:
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001681052
007 cr mnu|||uuuuu
008 051221s2005 flu sbm s000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0001070
Harris, Karen Monk.
An examination of the relationship between urbanicity and children with emotional disturbances served in restructuring public schools
h [electronic resource] /
by Karen Monk Harris.
[Tampa, Fla.] :
b University of South Florida,
Thesis (Ph.D.)--University of South Florida, 2005.
Includes bibliographical references.
Text (Electronic thesis) in PDF format.
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
Title from PDF of title page.
Document formatted into pages; contains 172 pages.
ABSTRACT: Efforts to affect changes in student achievement through altering the manners in which schools operate have been countless. However, there are few empirical studies on the relationship between these reform activities and student outcomes, especially outcomes for students with emotional disturbances from geographically diverse locations. The current study was a secondary analysis of data collected as part of the School and Community Study and the Urban School and Community Study conducted by the Research and Training Center for the Childrens Mental Health at the Louis de la Parte Florida Mental Health Institute at the University of South Florida. Both studies examined the relationship between student exposure to school restructuring efforts and change in academic and behavioral functioning.The primary purpose of this study was to investigate the relationship between student outcomes and school reform activities and to compare students attending suburban/rural schools and students attending urban schools on academic achievement, psychopathology, and mental health service utilization. Using baseline data from the School and Community Study to match students from the Urban School and Community Study on the variables gender, income, and age; 66 matches (i.e., 132 students) comprised the study sample. Differences between the suburban/rural students and the matched sample of urban students were statistically significant in reading achievement, math achievement, functional impairment, and mental health service utilization. There were no significant differences between students on the variable of level of behavior problems, all of the students scoring in the clinical range.
Adviser: Dr. Albert Duchnowski.
x Special Education
t USF Electronic Theses and Dissertations.
An Examination Of The Relationship Between Urbanicity and Children With Emotional Disturbances Served In Restructuring Public Schools by Karen Monk Harris A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Special Education College of Education University of South Florida Major Professor: Albert Duchnowski, Ph.D. Krista Kutash, Ph.D. Daphne Thomas, Ph.D. John Ferron, Ph.D. Date of Approval: April 4, 2005 Keywords: special education, school reform, achievement, psychopathology, mental health Copyright 2005, Karen Monk Harris
ii DEDICATION Now glory be to God who by His mighty power at work within us is able to do far more than we would ever dare to ask or even dream of infinitely beyond our highest prayers, desires, thoughts, or hopes. Ephesians 3:20 The accomplishment of this goal would not have been possible without the love and support of my family and it is to them this is dedicated: To my husband, Dyfierd, who embraced my desire to pursue academic excellence and held it tightly even when my own resolve weakened, To my children, Adrienne, Angelique, and Alex, who modeled determination and commitment, and To my mother, Blonnie Monk, who taught me the importance of being an advocate for children and has always been my inspiration.
iii ACKNOWLEDGEMENTS I would like to demonstrate my appreciation by recognizing the following people who were instrumental in the completion of my graduate studies and research. Albert Duchnowski, for his professional guidance as my major professor, directed my work to become a champion of childrens mental health and an advocate of the system of care. Krista Kutash, for her mentorship, encouragement, and expertise, the research experience I gained working with the team is invaluable. John Ferron, who helped me befriend statistics and Daphne Thomas, who shares my interest in culture and family, I am grateful for the insight and knowledge you shared and your participation on my committee. I especially want to express my gratitude to Deborah and Donnie for their wise counsel, moral support, and making their home, my home.
i TABLE OF CONTENTS LIST OF TABLES iv LIST OF FIGURES vii ABSTRACT viii CHAPTER ONE INTRODUCTION 1 Introduction 1 Rationale 3 Purpose 5 Research Questions 5 Study Limitations 6 Definitions 8 Urbanicity 8 Academic and Psychological Functioning Outcomes 9 School Reform and Restructuring Activities 9 Significance of the Study 9 Organization of the Study 10 CHAPTER TWO REVIEW OF LITERATURE 11 Defining Educational Setting 12 Urbanicity 12 Proportion Characteristics 13 Poverty Status 15 Urban Settings and Academic Outcomes 17 Suburban/Rural Settings and Academic Outcomes 21 Students with Emotional Disturbances 22 Definition 22 Characteristics 26 Disproportionality 28 Legal Assessment of Disproportionality 28 Academic Outcomes of Students with Emotional Disturbances 30 Mental Health Service Usage of Children with Emotional Disturbances 37 Educational Services 40 School Reform 41 General Education Reform 42 Special Education Reform 43
ii Systemic Reform 45 Summary 47 CHAPTER THREE METHOD 48 Purpose 48 Research Questions 48 Data Sources 50 The School and Community Study 51 The Urban School and Community Study 53 Participants 56 Schools 59 Study Variables and Instrumentation 60 Demographic Information 61 Academic Achievement 61 Wide Range Achievement Test 61 Psychological Functioning 63 Child Behavior Checklist 64 Child and Adolescent Functional Assessment Scale 66 Columbia Impairment Scale 68 Mental Health Service Use 69 Teacher Report 69 Child and Adolescent Services Assessment 70 Service Assessment for Children and Adolescents 72 School Reform and Restructuring Index 74 Research Design 78 Research Question 1 78 Research Question 2 78 Research Question 3 79 Research Question 4 80 Research Question 5 81 Summary 84 CHAPTER FOUR RESULTS 85 Sample 86 Unmatched subjects 98 Study Sample Size Summary 101 Research Question 1 101 Time Spent in School Setting 103 Research Question 2 103 Psychological Functioning 104 Functional Impairment 106 Research Question 3 109 School Staff Report 109
iii Parent Report 111 Research Question 4 115 School Reform and Restructuring Measures 116 Suburban/Rural Schools 116 Urban Schools 117 Conversion to Percentage Scores 118 Schools and Study Sample Subjects 121 School Reform and Restructuring Interviews 121 Governance 121 Accountability 122 Curriculum and Instruction 122 Includeness 122 Parent Involvement 123 Pro-Social Discipline 123 Research Question 5 124 Reading Achievement 125 Math Achievement 125 CBCL Total Problem Score 126 Mental Health Service Utilization 126 Summary 129 CHAPTER FIVE DISCUSSION Overview of the Study 131 Review of the Method 131 Discussion of Findings 132 Academic Functioning 133 Emotional Functioning 133 Mental Health Service Utilization 135 School Reform and Restructuring Levels 136 Study Limitations 138 Recommendations for Future Research 139 Summary 143 REFERENCES 145 ABOUT THE AUTHOR End Page
iv LIST OF TABLES Table 1 Characteristics of the Core Research Studies 7 Table 2 Characteristics and Outcomes for Children 16 Table 3 Characteristics of Urban Public School Students as Compared to Suburban/Rural Public School Students 20 Table 4 Areas of Outcomes for Children and Youth with Emotional Disturbances 27 Table 5 Summary of Characteristics and Functioning of Children and Youth with Emotional Disturbances 34 Table 6 Research Design for the Two Studies 51 Table 7 Characteristics of Students Participating in the Research Studies 57 Table 8 Domains, Sources, and Instruments Used in Both Studies 61 Table 9 Services Assessed by the Child and Adolescent Services Assessment 72 Table 10 Services Assessed by the Service Assessment for Children and Adolescents 74 Table 11 Reform and Restructuring Propositions 76 Table 12 An Example Layout of the Analysis for Research Question #4 82 Table 13 An Example of the Four Regression Equations to Answer Research Question #5 83 Table 14 An Example of School within the Factor of Urban/Suburban/Rural Settings 84 Table 15 Income Conversion from Categorical to Continuous Variables For the Suburban/Rural Study 88
v Table 16 Age and Income Data on the Matched Sample 94 Table 17 Summary Statistics of Matched Sample 97 Table 18 Age and Income Data for Matched Subjects using New Parameters 100 Table 19 Age and Income Data on Remaining Subjects 100 Table 20 Descriptive Statistics for WRAT3 Reading and Math Achievement Standard Scores 103 Table 21 Percent of Day Spent in Educational Settings 104 Table 22 Child Behavior Checklist (CBCL) T-scores 106 Table 23 Child and Adolescent Functional Assessment Scale (CAFAS) 107 Table 24 Columbia Impairment Scale (CIS) 109 Table 25 Investigation of the Relationship between the Impairment of Suburban/Rural Subjects and Urban Subjects 110 Table 26 Therapeutic Services Received in the School, during the School Day, from School Personnel or from Agency Professionals 111 Table 27 Frequencies for type of Services Ever Used and Used in the Past Six Months for Suburban/Rural Subjects (N = 69) 113 Table 28 Frequencies for Type of Service Ever Used and Used in Last Six Months for Urban Students (N = 69) 115 Table 29 Investigation of the Relationship between Mental Health Service Utilization (Ever Used) and Setting 116 Table 30 Results of Ratings of Propositions on School Reform and Restructuring Suburban/Rural Schools 118 Table 31 Scores for Urban Schools by School Reform and Restructuring Proposition 119 Table 32 Reform and Restructuring Data on Study Schools 120
vi Table 33 Rank Order of School Reform and Restructuring Index (SRRI) Scores from Lowest to Highest 121 Table 34 Multiple Regression Analysis for Reading Scores of Matched Subjects (N = 138) 128 Table 35 Multiple Regression Analysis for Math Scores of Matched Subjects (N = 138) 128 Table 36 Multiple Regression Analysis for CBCL Total Scores of Matched Subjects (N = 138) 129 Table 37 Multiple Regression Analysis for Mental Health Service Utilization of Matched Subjects (N = 138) 129
vii LIST OF FIGURES Figure 1. A graphic representation of the matching method for female students in the two studies 59 Figure 2. A graphic representation of the matching method for male students in the two studies. 59 Figure 3. A scatterplot of income and age for males over the age of 14 by income and age. 89 Figure 4. A scatterplot of income and age for males between the ages of 10 and 14 by income and age. 90 Figure 5. A scatterplot of income and age for males under the age of 14 by income and age. 90 Figure 6. A scatterplot of income and age for females over the age of 10 by income and age. 91 Figure 7. A scatterplot of income and age for females over the age of 10 by income and age. 91 Figure 8. .A scatterplot of income and age for matched male subjects by income and age 92 Figure 9. A scatterplot of age and income for matched female subjects by income and age 93
viii An Examination of the Relationship between Urbanicity and Children with Emotional Disturbances Served in Restructuring Public Schools Karen Monk Harris ABSTACT Efforts to affect changes in student achievement through altering the manner in which schools operate have been countless. However, there are few empirical studies on the relationship between these reform activities and student outcomes, especially outcomes for students with emotional disturbances from geographically diverse locations. The current study was a secondary analysis of data collected as part of the School and Community Study and the Urban School and Community Study conducted by the Research and Training Center for the Childrens Mental Health at the Louis de la Parte Florida Mental Health Institute at the University of South Florida. Both studies examined the relationship between student exposure to school restructuring efforts and change in academic and behavioral functioning. The primary purpose of this study was to investigate the relationship between student outcomes and school reform activities and to compare students attending suburban/rural schools and students attending urban schools on academic achievement, psychopathology, and mental health service utilization. Using baseline data from the School and Community Study to match students from the Urban School and Community Study on the variables gender, income, and age; 66 matches (i.e., 132 students) comprised the study sample. Differences between the
ix suburban/rural students and the matched sample of urban students were statistically significant in reading achievement, math achievement, functional impairment, and mental health service utilization. There were no significant differences between students on the variable of level of behavior problems, all of the students scoring in the clinical range. Schools in the suburban/rural settings were more highly engaged in reform and restructuring activities than the schools in the urban settings. Multiple regression equations were used to compare differences in school reform mechanisms and student outcomes. Family income and degree of engagement in school reform had a positive impact on reading achievement. Future research on the relationship between student outcomes and school reform and restructuring activities is needed to guide school efforts to improve student academic performance.
1 CHAPTER ONE INTRODUCTION Critics of urban schooling presume that urban public schools are in a state of crisis and in need of drastic changes and solutions (Hess, 1998). While urban public school students lag behind students attending suburban/rural schools academically, the achievement gap may be a factor of the students environment. Students who attend urban public schools, generally, come from neighborhoods with poverty and crime levels above the national average, are more likely to belong to a racial or ethnic minority, and are more prone to engage in risk-taking behaviors (Lippman, Burns, McArthur, & NCES, 1996; Crosby, 1999). The schools that service these students are often characterized as bureaucratic and inflexible and educational outcomes for urban students are generally dismal. Conversely, students attending suburban/rural schools are perceived as having economic stability, secure and safe neighborhoods, and have active parent involvement, that lessens the likelihood of risk-taking behaviors. Their schools tend to be less bureaucratic, adaptable, and education outcomes are better (Crosby, 1999; Lippman et al., 1996). It should be noted, however, that there is a general dissatisfaction with educational outcomes for all students in this country. The publication of A Nation at Risk (National Commission on Excellence in Education, 1983) lambasted the nations schools, igniting an educational reform movement that continues today. Educators responded to
2 the grim reports of educational outcomes and searing criticism by making changes in standards, assessments, accountability, and governance in public school systems. These systemic reforms have been beneficial in describing how schools should function and what students should know, but have not totally been effective for many reasons (Kutash, Duchnowski, Kip, Greeson, Sheffield, & Oliveira, 2001). The policies are established in state capitals and district offices and are prescribed to large numbers of schools in local communities. Although these reform policies have created a formula that is to produce improvement in both urban and suburban/rural schools, this generic approach to school improvement lacks the research and documentation that supports a one-size-fits-all strategy (Cuban, 2001; Slavin, 2001). The complex nature of school reform has been cited as a contributing factor for scant empirical research. Methodological and practical challenges create barriers hindering research and evaluation efforts (Frechtling, 2000). The exclusion of students with disabilities from standardized academic assessments heightens those barriers by preventing any assessment at all of the progress that students who have disabilities are making under school reform initiatives (Vanderwood, McGrew, & Ysseldyke, 1998). Additionally, there is sparse educational literature that examines the discrepancies between urban and suburban/rural school populations as it relates to students educated in special education classrooms, specifically those students identified as having emotional disturbances. While there has been little literature relating school reform activities and students with emotional disturbances, there has been a vast amount of literature characterizing the students. It has been documented that students with emotional disturbances have lower graduation rates, lower grade point averages, higher dropout rates and higher absenteeism
3 in comparison to students in other disability categories (Cullinan, Epstein, & Sabornie, 1992; U.S. Department of Education, 1996). Over the years, the prevalence and severity of problems experienced by these students has shifted dramatically from mental and behavioral problems (i.e., depression and social isolation) to critical behavioral events (i.e., severe aggression, antisocial behavior, and interpersonal violence) (Walker, Sprague, Close, & Starlin, 2000). These characteristics are not symbolic of all students identified as having emotional disturbances but the intensity of the behaviors is greater and occurs at a higher rate than with their same age peers. Reasonable prevalence estimates based on the best available research are that 3 to 6 percent of the school-age population are in need of special education and related services because of their emotional or behavioral disabilities (Kauffman, 1997, p. 58). Post-school outcomes, such as unemployment, independent living, and involvement with the criminal justice, mental health, and welfare systems, contrast sharply with those of their peers (Silver, Unger, & Friedman, 1994; Walker & Brunson, 1995). These data indicate that current systemic efforts to positively impact the lives of students with emotional disturbances are failing (Koyangi & Gaines, 1993). Rationale If the outcomes for children with emotional disturbances are going to improve, in both the academic and the social/emotional domains, members of the research community need to continually investigate and evaluate educational reform approaches. Current school reform literature lacks the theoretical framework, specificity of reform mechanisms, and whole school reform evaluation data to enhance student outcomes (Kutash, Duchnowski, Rivera, Oliveira, & Kelly, 1997). While school reform efforts
4 offer the opportunity to improve outcomes for students with emotional disturbances, there is a need to determine whether outcomes do improve, to identify the key elements of change related to the improvement, and to investigate the variability in school reform strategies based on school location. The School and Community Study, conducted by a research team at the Research and Training Center for Childrens Mental Health (Kutash, Duchnowski, Calvanese, Rivera, & Oliveira, 1997), examined the effects of service system reform in both the education and social service systems. The focus of this investigation was on the effects of school restructuring and special education reform on children with serious emotional disturbances. Employing a multi-modal design, the study sought to identify exemplary schools, identify the reform activities at these schools, and relate these reform activities to a variety of academic and behavioral outcomes for students with serious emotional disturbances. This study also obtained descriptions of the restructuring and reform activities in 6 areas: governance, curriculum and instructional reform, accountability, parent involvement, includeness, and prosocial discipline. Finally, a systematic description was obtained of social and mental health services offered to students. Expanding upon the School and Community Study whose schools were located in suburban or rural areas, the Urban School and Community Study (Kutash, Duchnowski, Kip, Oliveira, Greeson, & Sheffield, 2001) was begun. The goal of this study is threefold. First, the study identifies and describes the common features of the reform models operating in the selected urban schools serving a diverse population of students. Second, the study examines the results of these reform models in comparison to similar efforts in other sites. Lastly, the study investigates the links between mental health
5 services programs and special education programs. This study used a mixed method methodology to gather information that was similar to the design used in the School and Community Study. Purpose Evaluations of school reform initiatives are in their infancy (Frechtling, 2000). The complex nature of the reform process poses methodological barriers to conducting comprehensive, empirical studies to evaluate the various school reform mechanisms operating in Americas schools. Additionally, there have been no empirical investigations of the effects these reform mechanisms have on the academic and emotional functioning of students in special education due to emotional disturbances. The purpose of the current study is to add to the knowledge base by describing and contrasting school reform activities in urban and suburban/rural communities and the effects associated with these activities for students in special education due to emotional disturbances. The results of the current study will supply the field with much needed empirical information on both, the activities of reform operating in urban and suburban/rural schools, and the characteristics of students in special education due to emotional disturbances. Potential differences between urban and suburban/rural reform activities and the differential effects on students in special education will also be documented. Research Questions 1. Are there differences in academic functioning, i.e., reading and math achievement, between urban students in special education classrooms due to
6 emotional disturbances and a matched sample of suburban/rural students in special education classrooms due to emotional disturbances? 2. Are there differences in the psychological functioning (i.e., symptomatology and functional impairment) between urban students in special education classrooms due to emotional disturbances and a matched sample of suburban/rural students in special education classrooms due to emotional disturbances? 3. Are there differences in the mental health service usage between urban students in special education classrooms due to emotional disturbances and a matched sample of suburban/rural students in special education classrooms due to emotional disturbances? 4. Are there differences in the school reform activities (e.g., governance, accountability, prosocial discipline, accountability, inclusion, family involvement, curriculum and instruction) between the urban public schools attended by students with emotional disturbances and the suburban/rural schools attended by students with emotional disturbances? 5. What is the contribution of school reform activities in explaining those differences (i.e., academic functioning, psychological functioning, and service usage) between urban students in special education classrooms due to disturbances and suburban/rural students in special education classrooms due to emotional disturbances? Study Limitations As is true with most studies, limitations are inherent and pose potential threats to the reliability and validity of the results. This study is a secondary analysis of data from
7 two federally funded research projects. Many of the same advantages and disadvantages inherent in secondary analyses apply to this study as well. As noted by Yegidis and Weinbach (1991), the use of an existing data set decreases costs, increases time efficiency, and is a non-intrusive means of analysis. However, this form of research limits the researchers in variable selection, research design, instrumentation, participants, and the data collection process. The participating schools, in both of the core studies, were purposively selected based on their urbanicity and level of reform activities. Schools in the urban communities had majority black student populations whereas the schools in the suburban/rural communities were attended by a majority of white students. While race is not the focus of this study, a complex relationship exists between race and urbanicity. A clear distinction cannot be drawn to distinguish effects directly attributable to race or to urbanicity. Thus, the schools were studied as they exist, in the context within which they operate, and with the students that they serve (see Table 1). Table 1 Characteristics of Core Research Studies Research Study Urbanicity Level of Reform Activities Urban School and Community Study Urban Actively Engaged/Less Actively Engaged School and Community Study Suburban/Rural Actively Engaged
8 This is a descriptive study that seeks to build upon the descriptions (e.g., characteristics and outcomes) of students with emotional disturbances in the core studies. Both studies selected schools based on their ongoing reform and restructuring strategies. Schools who participated in the School and Community Study were selected due to their exemplary approach to restructuring and reform. The selection process varied slightly in the Urban School and Community Study. In this study, schools were invited to participate that exhibited active approaches to reform as measured by the School Reform and Restructuring Index and then a control group of schools was selected based on their less than active approaches. This sampling strategy may reduce the widespread use of the results. The sample for the current study was achieved by matching participants, from both studies, on selected variables. Matching was used to reduce errors or extraneous variables. However, the relationship between these matching variables may net different results in other studies. Additionally, all of the students in the study were identified as having emotional disturbances. This further limits the generalizability of the findings. Finally, the data were collected within a specific social and historical context that also limits its widespread use. The process linking urbanicity to poverty and the connection to racial/ethnic differences may change over time and vary across locations. Definitions Urbanicity This is a term used to describe a school by its geographical location. Categories of urbanicity are: urban, suburban, and rural (Lippman et al., 1996). For the purposes of this study, only public schools participated in the data collection effort.
9 Academic and Psychological Functioning Outcomes Scores on standardized instruments that measure areas of academic achievement and psychological functioning was used as outcomes. A description of the instruments utilized in this study appears in Chapter 3. School Reform and Restructuring Activities These are activities that promote changes in the organizational structure and the performance of school personnel. Typically, the changes alter the policies, procedures, practices, and fundamental assumptions of the school that should result in improvements in student outcomes. Significance of the Study While explaining and understanding the differences in school performance has been a longstanding topic of educational research, there has been a lack of studies investigating student outcomes from a school reform and restructuring perspective. This study focused on the differences in approaches to reform and restructuring of urban and suburban/rural schools and compared those differences to educational outcomes for students with emotional disturbances. The differences discussed may have implications for educational policy and practice. However, there is a more pressing significance of this study. Empirical literature concerning urbanicity, children with emotional disturbances, and school reform outcomes is limited. According to the U.S. Census (2000) the majority of children live in impoverished urban environments and a large percentage of them have been identified as having some type of disability. This means that urban schools are taxed with meeting the multiple needs of their student population. Studies are needed that investigate the
10 relationships between these variables and add to the knowledge base. This study sought to accomplish this task. Organization of the Study Chapter One provides a brief introduction to the current status of children with serious emotional disturbances enrolled in public urban and suburban/rural schools. This overview describes the importance of this study by reviewing current practices and research that indicates a prevailing need to improve outcomes for these children through school reform and restructuring activities. Chapter One concludes with a section on limitations, definitions and the significance of the study. Chapter Two provides a review of the relevant literature and critically reviews associated information relating to the purpose of this study. In Chapter Three, the participants, methodological procedures, and statistical analyses to be utilized in this research study are described. The results of the analyses are presented in Chapter Four. Chapter Five provides a discussion of the findings and the relevance they have on the current climate of schools.
11 CHAPTER TWO REVIEW OF THE LITERATURE This chapter presents a review of the literature that guides this study. The purpose of the study was to examine the relationship between urbanicity, school reform and restructuring activities, and the functionality of children and youth with emotional disturbances. The chapter begins with an examination of the relevant literature on public school location, in terms of urbanicity. Descriptions of the public school student populations in urban and suburban/rural locations (e.g., student characteristics, and academic outcomes) will be presented. Next, the target population of this study, children with emotional disturbances, is described. This description includes a definition of emotional disturbances, characteristics (including the disproportionate representation of minority students in this disability category), academic outcomes, and mental health service utilization for this population. The study explored the extent to which school reform activities impact the academic achievement, psychological functioning, and mental health service utilization of children with emotional disturbances. A historical review of the dual reform efforts in general and special education and the movement towards whole school reform will be presented. The limited empirical knowledge base on the relationship between school reform and restructuring mechanisms and student outcomes and the extant literature
12 documenting the impact of school reform activities on children with emotional disturbances will be included in this discussion. Defining Educational Settings Because educational context or setting is generally associated with academic and behavioral outcomes, it is a major component of any discussion that focuses on education. In order to better understand the variations in student populations and their educational communities, researchers designed a framework that describes school location in terms of population density and poverty concentration. This framework provides guidance in understanding the following characteristics of educational settings. Urbanicity Recent studies have described school location in terms of urbanicity. The term, urbanicity, is used to convey a schools location in relation to the closest large city. Urbanicity can be divided into three categories: urban, suburban, or rural. Urban refers to the area within the boundaries of a large city and is densely populated. Suburban settings are extensions of the large city with fewer inhabitants while rural areas are the most sparsely populated and located furthest away from the large city (Economic Research Service, 1993). Researchers and statisticians have used these categories to group and describe schools, students, and various other educational factors (e.g., resources and achievement). For example, the U.S. Census (1998) reported that 29% of U.S. students are enrolled in urban schools, 51% in suburban schools, and 20% in rural schools. In developing an urban-suburban-rural method of describing or measuring locations, city populations were a natural starting place (Goodall, Kafadar, & Tukey, 1998). The most commonly used method uses categorical variables given by ranges of
13 total population, population density, percent urban population, or a combination of those variables. According to this method, urban refers to cities with populations exceeding 500,000 inhabitants. Cities with populations less than 500,000 fall into the suburban/rural category (National Center for Education Statistics, 1988; U.S. Census Bureau, 2000). Cities, in particular urban areas, can more accurately be described by their population nucleus, commuting ties, and metropolitan character using a statistical method. This method is based on the following criteria: 1) the area must include a city with a population of at least 50,000 or 2) the area must meet the Census Bureau definition of urban and have a total Metropolitan Statistical Area population of at least 100,000 (National Center of Education Statistics, 1998). Recent data show that blacks were more likely than non-Hispanic whites to live in metropolitan/ urban areas (86% compared with 77%) (McKinnon & Humes, 2000, p.2). Population Characteristics Additional comparative data show that in 1999 there were 53.1 million white families and 8.4 million black families living in the United States. Of those families, less than one-half (47%) of all black families were married couple families, 45 % were maintained by women with no spouse present, and 8% were maintained by black men with no spouse present. The corresponding figures for white families were 82%, 13% and 5%, respectively (McKinnon & Humes, 2000, p.2). Although the majority of economically depressed families in the United States are white (15.8 million; 9.1 million blacks), the poverty rate was 26% for blacks and 8% for whites. Poverty is highest among families headed by women with no spouse present
14 (41%) compared to 21% for white families. The concentration of these single parent, black families can be found in the urban cities (McKinnon & Humes, 2000). The family structure in terms of the number of parents in a household has been linked to a childs academic success (Kretovics & Nussel, 1994; Mulkey, Crain, & Harrington, 1992; Weist, Paskewitz, Warner, & Flaherty, 1996). In a study conducted by Lippman, Burns, McArthur, and NCES (1996), approximately 30 percent of urban students lived in a one-parent household as compared with 20 percent of suburban/rural students. With only one parent in the home, that parent is likely to have less time to spend with the child and the household income is generally lower than in a two-parent home. There are numerous possible family configurations (i.e., mother and father, mother only, father only, mother and grandmother, grandmother only, or older siblings as parents) that can define family structure and all of them have implications for children and their educational development (Entwisle & Alexander, 1995). There is an estimated 3.3 million children under the age of 18 who live with their grandparents or other relatives. Comparative data across ethnic groups indicate that grand parenting is particularly prevalent in black communities (12%), 5.8% in Hispanic communities and 3.6% in white communities (Rodgers & Jones, 1999). It has also been found that children raised by their grandparents are more likely to be diagnosed with developmental delays and attention deficit disorders (Goldberg-Glen, Sands, Cole, & Cristofalo, 1998). Clearly, children in a one-parent household, headed by a single mother and in households headed by a grandparent, are reported to experience the greatest risk for economic and environmental disadvantage. These disadvantages have been linked to
15 school failure and special education placement (Gottlieb, Gottlieb, & Wishner, 1994; Shapiro, 1996). Poverty Status Another means of discriminating between school settings is done through examining the schools poverty concentration or the socioeconomic (SES) level of the schools students and families. Poverty concentrations are derived through information provided by school administrators who report the number of students receiving free or reduced priced meals (Lippman et al., 1996). These data are often misleading because the federal definition of poverty is different from the eligibility criteria for reduced priced meal service and students in upper grades are often reluctant or embarrassed to apply for the meal service. Therefore, family income level is a more accurate indication of socioeconomic status. The failure to report or the misreporting of family income can also lead to erroneous estimations and both instances can lead to inaccurate calculations of the schools poverty concentration. Regardless of these limitations, researchers have concluded that children living in urban areas attend public schools with high poverty concentrations. For children under the age of 18, the poverty rate was 19 % but three times as high for black children (37%) as for non-Hispanic white children (11%) (McKinnon & Humes, 2000). Lippman et al. (1996) reported that in 1991, 30 percent of children in urban locations were living in poverty, more than twice the rate for children living in the surrounding suburbs (13 percent) and only slightly higher than the rate for children living in rural areas (22 percent). These data clearly show that urban students face tougher life circumstances than their suburban/rural counterparts (seeTable 2).
16 Table 2 Characteristics and Outcomes for Children Urban Suburban Rural Percent living in poverty 30 13 22 Percent receiving free/reduced school meals 38 16 28 Percent attending public schools with a high poverty concentration1 40 10 25 Percent of students that graduated on time 66 74 80 National Math and Reading Mean Standard Scores: 8th grade level 47 51 49 10th grade level 49 51 49 Percent of young adults living in poverty2 26 14 10 Percent of young adults unemployed2 11 5 7 1 High poverty concentration is defined as 40% of the students receive free or reduced school meals. 2 Calculated 7 to 15 years after high school. There are distinct differences between urban school environments and suburban/rural school environments that affect students respective school experiences. Urban school students have a myriad of things to confront which make their environment and worldview different from their suburban/rural counterparts. Compared with suburban/rural public school students, students attending urban public schools are more likely to be confronted by crime, violence, and unemployment. As a consequence, schools assigned to educate students living in urban areas must be sensitive to environmental characteristics.
17 Urban Settings and Academic Outcomes In the 1930s, black migration to northern settings caused dramatic shifts in the national population map. The loss of political power after Union troops withdrew from the South, the destruction of the cotton crops by boll weevils, and Jim Crow laws motivated blacks to leave the southern states. Aside from economic considerations, blacks were drawn to Americas major urban areas because they were seeking space within a highly oppressive society space to assume control of their own institutions, and thus reclaim those institutions from the control of a repressive white power structure (Carlson, 1998, p.282). Major urban industrialization meant employment opportunities for blacks. To some extent racial minorities were drawn to urban areas by the new service industry and jobs which many working class white males refused to take (Carlson, 1998, p.280). The end of World War II also caused significant changes to the urban landscape. Before and during World War II, metropolitan cities were home to middle and upper income white families. As the war was ending those families began migrating to the outlying areas. Returning soldiers were awarded federal loans, enabling them to easily secure new affordable homes in the suburbs. Left behind in the city was the poor, the elderly, the new migrants (Asian and Hispanic) and a large proportion of blacks (Kretovics & Nussel, 1995, p.43). By the 1940s, urban ghettos were firmly established and weak learning institutions were beginning to emerge (Balfranz, 2000). These schools, attended primarily by black students, were decaying buildings with large student populations often
18 overcapacity. This led to large class sizes, housing students in makeshift facilities, and the adoption of double sessions. The accommodations made for the large student populations had negative consequences. Researchers have documented that urban students were attending one hour less per school day than their suburban/rural peers. Urban students were also being guided into a general track of studies that had a weak academic focus, designed for noncollege bound students (Balfranz, 2000). A steady educational decline in the urban schools continued through the 1960s. Conan (1961) found that the academically weaker schools were linked to the segregated, minority populated neighborhoods and the housing and employment discrimination felt by the urban inhabitants. According to the U.S. Census (2000), minority populations represented a majority in 51 American cities. The minority school population in Washington, DC was 73%; 66% of which were black; Detroit was 79% minorities; 76% of which were black. Children living in these cities have a stronger likelihood to be living in poverty than those children living in suburban/rural cities. Likewise, these urban children are more than twice as likely as suburban/rural children to receive a free or reduced price lunch. Previous research suggests that students from both schools with high concentrations of low income students and students from urban schools would be expected to have less successful educational outcomes (Lippman et al., 1996). Schools located in urban areas are plagued with larger student enrollments and fewer resources as compared to their suburban/rural counterparts (Lippman et al., 1996). These factors can be related to the characteristics of their student populations. The urban students, characteristically, have more behavior problems, higher levels of absenteeism,
19 and feel less safe at school than students that attend suburban/rural schools. Notably, urban teaching staffs have higher rates of absenteeism than their suburban/rural peers, although they have more years of teaching experience and garner higher salaries. The combination of poverty and urbanicity has the tendency to produce less than desirable educational environments for children. Yet, despite these factors, more than two-thirds of the children (66%) from urban schools graduated on time and several years later, were employed or in school, and living above the poverty level (Lippman et al, 1996) (see Table 2). In discussing urban public schools, President George W. Bush talked about the soft bigotry of low expectations. His comment was in response to the dismal achievement scores that have emerged from urban public schools. President Bush stated that the urban students most affected are poor or minority or have limited English proficiency (U.S. Department of Education, 2001). As a result, a comprehensive plan to address achievement scores and educational outcomes was initiated. Bushs plan could also have been in response to a report released in 1996. The Urban Schools: The Challenge of Location and Poverty (Lippman et al., 1996) report is one of the largest longitudinal research projects investigating the educational outcomes, school experiences, and family characteristics of children attending urban public schools as compared to children attending public schools in other locations. One of the studys goals was to determine if and to what extent urban students were being undereducated. Using data from several national sources and data collected independently, the study amassed one of the largest study samples of public school students. The study controlled for poverty concentration allowing the researchers to closely examine other variables that
20 could affect achievement (e.g., family background, educational attainment, and school resources). Conclusions drawn from the study were that urban public school students varied from their suburban/rural counterparts on several measures (e.g., educational attainment, family characteristics and school experiences). Urban public school students are more likely to live in a one-parent home, have a disability, and attend schools with limited resources and large enrollment (see Table 3). Table 3 Characteristics of urban public school students as compared to suburban/rural public school students Urban public school students are more likely: to live in a one-parent home to have changed schools more than once to have disciplinary problems to watch a lot of television to have difficulty speaking English to engage in risk-taking behaviors to have attended a preschool program to have a disability They are less likely: to have attended schools with gifted and talented programs to have parents who completed college to participate in school-sponsored extracurricular sports activities to have access to medical care Studies have found that urban public schools have: fewer resources larger enrollments teachers have less control over their curriculum teachers had comparable levels of experience and salaries higher concentrations of less advantaged student
21 Suburban/Rural Settings and Academic Outcomes About one-quarter of Americans live in big cities, half live in suburbs and a quarter live in small towns or rural areas. Unlike the diverse population of urban areas, suburban/rural areas tend to be homogenous with a low poverty concentration. School enrollments in these areas reflect these characteristics. According to Lippman et al. (1996), 72% of the nations students are enrolled in schools located in suburban or rural settings. These schools make up the largest portion of public elementary and secondary schools in the nation. In another report by the National Center of Education Statistics (1997), schools, located in small rural school districts, averaged 10 students per high school grade and 25 students per elementary grade. These schools enroll a smaller percentage of students whose primary language is not English and nonwhite students than urban high schools. Poverty rates for rural areas average 22% well below the rates reported for urban areas (30%) (Lippman et al., 1996). The academic outcomes of students enrolled in suburban/rural schools remain consistently higher than those of students enrolled in urban areas. Besides the lower school populations, smaller class sizes and lack of diversity, suburban/rural schools have larger fiscal budgets and higher levels of parental support that may account for the differences (Lippman et al., 1996). Larger school budgets allow administrators to furnish their schools with equipment and supplies urban schools lack. The added component of parental support increases the level of community involvement and investment in the student body. Interestingly, the Urban Schools report (Lippman et al., 1996) reported differences between urban students and suburban/rural students on 8th grade achievement
22 test scores and high school completion rates but the differences disappeared when the students were retested in the 10th grade. The variance in high school graduation rates also disappeared after two years. The reduction in differences could have resulted from several methological factors: 1) high schools are larger and more heterogeneous, 2) 8th grade students that dropped out of school were not included in the follow-up sample, and 3) the 10th grade sample included those students who graduated later than scheduled. The conclusion is made that when poverty is accounted for, the differences between urban schools and suburban/rural schools are not as dramatic as have been perceived. Students with Emotional Disturbances Definition The inception of Public Law 94-142 in 1975, and the recent reauthorization of the Individuals with Disabilities Education Act (IDEA) in 1997, mandated that children identified as having emotional disturbance are required by law to be provided a free appropriate public education in the least restrictive environment (Yell & Shriner, 1997). It was hoped that the identification of this disability group would lead to much needed services to this population of children however, definitional issues continued to plague parents and professionals. While parents and professionals define the disability in different contexts (e.g., parents focus on home/familial interactions, psychiatrists focus on pharmacological treatments, and psychologists focus on behaviorally based therapy), there is little agreement about when the behaviors become deviant (Forness, 1997). Several definitions exist that have sparked debates about the most appropriate method to identify and provide services to children characterized as: noncompliant, depressed, aggressive, and anti-social (Cullinan, Epstein, Sabornie, 1992, Dunlap & Childs, 1996).
23 However, special education services designated to children who experience serious emotional problems as defined under the category of Emotional Disturbance. The definition is as follows: 1. The term means a condition exhibiting one or more of the following characteristics over a long period of time to a marked degree, which adversely affects educational performance: a. An inability to learn which cannot be explained by intellectual, sensory, or health factors; b. An inability to build or maintain satisfactory interpersonal relationships with peers and teachers; c. Inappropriate types of behavior or feelings under normal circumstances; d. A general, pervasive mood of unhappiness or depression; or e. A tendency to develop physical symptoms or fears associated with personal or school problems. 2. The term includes children who are schizophrenic (or autistic). The term does not include children, who are socially maladjusted, unless it is determined that they are seriously emotionally disturbed. (Education of the Handicapped Act of 1977) This definition grew out of the findings of research conducted by Bowers (1981). After completing an extensive investigation of students in the California school system, he derived a definition for emotionally handicapped students. The definition lists five characteristics of types of behaviors exhibited by students with emotional handicaps, ranging from school-related problems with learning to social interactions. Significant to his definition was the inclusion of levels of severity. Bowers definition was highly criticized for its high degree of inaccuracy and subjectivity (Kaufman, 1997). One of the greatest barriers to the provision of services to children with emotional disturbances is the difference among professionals in the interpretation and implementation of the IDEA definition of emotional disturbances. The definition is
24 controversial because of the ambiguities in its content and wording. Phrases such as, over a period of time, to a marked degree, and adversely affects educational performance confound professionals as they attempt to identify children with the disability. Believing that professionals and children would be better served with a definition that eliminated the need to be pigeonholed into one of the five criterion areas for serious emotional disturbance, a working group of educators developed an alternative definition that relied more on functional assessments. The definition calls for: a. establishing the level of difference of the childs behavioral or emotional responses through standard diagnostic procedures, interviews, checklists, case histories, observations, or the like; b. establishing that significant impairment indeed exists in at least one area of educational performance; c. considering other differential diagnoses or alternative reasons for the childs difficulty; and d. ensuring that pre-referral interventions and multiple sources of case data have been adequately assessed (Coalition Work Group on Definition, 1992). The proposed definition aligns with the diagnostic concepts of other disability categories and sought to address the persistent evidence of the underidentification of children with emotional disturbances. The wording in this definition would increase earlier identification and service provision. Unfortunately, as with the other definitions, the quality of the interventions and other factors could not be guaranteed (Forness & Knitzer, 1992). The Center for Mental Health Services (CMHS) developed a definition for children with serious emotional disturbances in response to the Alcohol, Drug Abuse, and Mental Health Administration Reorganization Act in 1992. While similar to the U.S.
25 Department of Education definition, the CMHS definition was designed for state utilization in mental health services planning and block grant funding (Friedman, Kutash, & Duchnowski, 1996). Children with a serious emotional disturbance, according to the CMHS definition, are described as persons who are: From birth to age 18, who currently or at any time during the past year, have had a diagnostic mental, behavioral or emotional disorder of sufficient duration to meet diagnostic criteria specified within the DSM-III-R, that resulted in functional impairment that substantially interferes with or limits the childs role or functioning in family, school, or community activities. These disorders include any mental disorder (including those of biological etiology) listed in the DSM-III-R or its ICD-9CM equivalent (and subsequent revisions), with the exception of DSM-III-R V codes, substance use, and development disorders, which are excluded, unless they co-occur with another diagnosable serious emotional disturbance. All of these disorders have episodic, recurrent, or persistent features; however, they vary in terms of severity and disabling effect. Functional impairment is defined as difficulties that substantially interfere with or limit a child or adolescent from achieving or maintaining one or more developmentally appropriate social behavioral, cognitive, communicative, or adaptive skills. Functional impairments of episodic, recurrent, and continuous duration are included unless they are temporary and expected responses to stressful events in the environment. Children who would have met functional criteria during the referenced year without the benefit of the treatment or other support services are included in this definition. (Substance Abuse and Mental Health Services Administration, 1993, p.29425). The inability of professionals in the field to agree upon a definition of the disability impairs the quality of services and outcomes for children with emotional disturbances. Kauffman (1997) emphasized the conceptual framework that a succinct definition provides practitioners and therefore reflects the intervention strategies selected. A clear, concise, agreed upon definition would enhance the identification and service provision of the population to be served. Forness (1997) reflected that children with emotional disturbances are very diverse, they need many different types of services, and
26 that the children being treated are still just children. He chides professionals to remember, we treat a child, not a disorder (Forness, 1997, p.36). Characteristics There are approximately 6 million to 9 million children and adolescents in the United States with emotional disturbances, as defined by the federal definition, accounting for 9 to 13 percent of all children (Friedman, Katz-Leavy, Manderscheid, & Sondheimer (1998). Some of the common behaviors exhibited by children with this disability are: hyperactivity, aggression, withdrawal, immaturity, and learning difficulties. Additionally, children with the most emotional disturbances may demonstrate distorted thinking, bizarre motor acts, and abnormal mood swings. Many children who do not have emotional or behavioral disabilities may display some of these behaviors at various times during their development. However, when children have emotional disabilities these behaviors continue over long periods of time and it is the frequency of the behaviors that separate them from their peers. Their behavior thus signals that they are not coping with their environment or peers (General Info Fact Sheet, 1997, p. 1). Several consistent findings have emerged about children with emotional disabilities (Friedman et al., 1996; Greenbaum et al., 1998). They are characteristically male, culturally or ethnically diverse, being reared in a low income, single parent home, primarily diagnosed with conduct disorders, and are at risk of delinquency and substance abuse. Children with emotional disturbances have significant outcomes in the areas of academic and psychological functioning, and mental health service use when compared to their general education counterparts (see Table 4).
27 Despite children with emotional disturbances being described as culturally or ethnically diverse, Achenbach and Edelbrock (1981) did not find racial differences in their study investigating behavioral problems in children. They did discover discrepancies in the behavioral ratings reported from various social classes, with children from lower classes exhibiting higher problem scores and lower competence scores. It was determined that when social class is controlled, ethnicity has little or no relationship to emotional or behavioral disabilities. However, the number of black children labeled as emotionally or behaviorally disabled and served in special education classrooms greatly outnumbers those of children of other ethnicities, particularly white children served in special education classes (National Research Council, 2002). Table 4 Areas of Outcomes for Children and Youth with Emotional Disturbances Area of Functioning Outcomes Academic Functioning Higher rates of absenteeism Below grade level in academic achievement Higher retention rates Less likely to graduate Psychological Functioning Rejection by others High rates of incarcerations Difficulty establishing and maintaining relationships with peers More restrictive placements Problems in the community Table 4 (continued)
28 Table 4 (continued) Area of Functioning Outcomes More restrictive placements High rate of comorbidity Mental Health Service Utilization Several year lag between identification and service provision Disproportionality Disproportionality can be calculated two different ways with significantly different meanings. One method examines whether the number of identified children with emotional disturbances is in proportion to those whose achievement or behavior indicates a need for special supports. The other, more commonly used method compares the number of students identified with emotional disturbances and receiving special services to the total student population (National Research Center, 2002; Harry & Anderson, 1994). The first method provides percentages that reflect appropriate numbers of students in their disability placements. Whereas, the second method renders substantial evidence that minority students are identified and placed in classrooms for students with emotional disturbances at a higher proportion than white students. Legal Assessment of Disproportionality In California, the case of Larry P. et al. versus Wilson Riles et al. (1979) accused the San Francisco school district of discriminating against five black children who had been placed in educable mental retardation (EMR) classes. At the time of the case, 29% of the school population was black while 69% of the students in the EMR classrooms were black. The judge ruled that IQ tests could not be used for the purpose of special
29 education placement for black students and the state was ordered to monitor and eliminate the disproportionate number of black students in the EMR classes (Chinn & Hughes, 1987; Harry & Anderson, 1994). The 1979 California judicial ruling had little effect nationally. The Office of Civil Rights (OCR) (1992) report found an overrepresentation of blacks in special education. Black males accounted for 8.23% of the total school enrollment nationally but accounted for more than twice that percentage in the categories of trainable mentally retarded (TMR), emotional disturbances (ED), and EMR. Similar results were reported by the National Longitudinal Transition Study (NLTS) in 1992. This study tracked a nationally representative sample of over 8,000 secondary school-aged special education students and found that there were twice as many black students represented in special education classrooms (24%) than black students in general education classrooms (12%). These percentages were true for all the disability categories. Explanations for the overrepresentation have been varied. Harry and Anderson (1994) surmise that behavioral and verbal styles of black students, in particular black males, results in misinterpretation and inappropriate emotional disturbances classification. Other researchers blame poverty levels as the culprits in overand misidentification (Oswald, Coutinho, & Best, 1998). Persistent patterns in the overrepresentation of black children can be traced over the past twenty years. Numerous causal factors have been cited in the literature ranging from failure of the general education system (Patton, 1998; Artiles & Trent, 1994) to inequalities associated with special education referrals and assessments (Harry & Anderson, 1994). Adding to these factors are the known definitional problems of the
30 emotional disturbances category that have serious implications for black students. These concerns and the attendant cultural variability of student behavior and teacher judgment place African American youth at great risk of being falsely labeled as SED (Patton, 1998, p. 27). Unfortunately, disproportionality persists even after causes have been defined, researched, and documented. Academic Outcomes of Students with Emotional Disturbances While the definition and the assessment procedures for the disability continue to be examined and debated, the outcomes are well documented. Findings from a number of national studies have provided clarity in our understanding of the complexities associated with this disability category (see Table 5). The National Adolescent and Child Treatment Study (NACTS; Greenbaum et al., 1996; Silver et al., 1992), was a seven-year longitudinal study of children and youth with emotional disturbances who ranged in ages from 8 to 18 years and were either served in a residential mental health facility or in the public school system. At its initiation in 1985, 812 children and youth from six states participated in the study. The majority of the sample was white and male with an average age of about 14 years old and lived in a twoparent home. The study revealed that the children were below grade level in math and reading achievement and were in the low-normal range in intelligence. According to the measures of adaptive and psychological functioning, the children displayed substantial levels of impairments and almost all of the children received mental health services. A major objective of NACTS was to obtain information on how emotional, behavioral, and academic functioning changed for children and youth over time (Greenbaum et al., 1998). At the end of the study with a remaining sample of 628
31 children, 77% of the original sample, educational levels for children were not encouraging. Of children who were under 18 years of age, 85% were below grade level in reading and 94% were below grade level in math. Children and youth over age 18 were also below grade level in reading (75%) and in math (97%). Consistent with national graduation data for children with emotional disturbances about 25% had obtained a regular high school degree, about 17% had received their General Education Development (GED) degrees, and 16% were currently enrolled in an educational program. Adaptive functioning as measured by the Vineland Adaptive Behavior Scales showed a significant decline. The Child Behavior Checklist (Achenbach, 1991), used to measure psychological functioning, netted mixed results. Students in the middle and youngest age groups improved but remained in the borderline and clinical ranges, respectively, while students in the oldest age group showed significant improvements moving out of the clinical range. Another objective of NACTS was to investigate service utilization of children and youth with emotional disturbances. Over the 7 years of the study, 93% of the children accessed mental health services most frequently followed by educational services, vocational services, child welfare, and nonroutine health care. In a survey of a nationally representative sample, Cullinan, Epstein, and Sabornie (1992) collected information on 269 students identified as having emotional disturbances, aged 12 to 17 years and served in public school settings. The majority of the students were white (76%; 22% African American, and 2% Hispanic) and male (79%) residing in a single-parent home (56%). The average intelligence score for the students was in the low-normal range, with females and minority students scoring lower than the students who were male and white. In an examination of educational placement, about one-third
32 of the students spent more than half of the school day with their non-disabled peers, while 39% of the students were served in self-contained classrooms. The average time spent in special education was 4.6 years, approximately 70% received at least one related service, and 16% were taking some form of medication. The Alternatives to Residential Treatment Study (ARTS; Duchnowski, Hall, Kutash, & Friedman, 1998) investigated the effectiveness of five innovative communitybased programs for children with emotional disturbances. The average age of the group (N = 163) was 14 years. The majority were white (65%) males (66%) and lived in an out-of-home placement at the initiation of the study. Academically, the children were functioning below grade levels in reading (80%) and math (90%). The ARTS data revealed that, on average, the reported age of onset for behavior problems was 6.8 years, while there was a two-year lag before the receipt of professional services (8.7 years). The children exhibited severe emotional and behavioral problems as measured by the Child Behavior Checklist (CBCL) and moderate to severe levels of impairment as measured by the Child and Adolescent Functional Assessment Scale (CAFAS). One hundred forty four children remained in the study after one year and showed improvement in academic functioning and a decline in psychological functioning. Wagner (1995) reported the results of the National Longitudinal Transition Study (NLTS), a nationally representative sample of approximately 8,000 youth with disabilities, aged from 13 to 21 years. Emotional disturbance was an identified subgroup (N = 777), characterized as 67% white, 76% male, and 44% lived in a single-parent home. On average, the students were functioning below grade levels, 2.2 grade levels behind in reading and 1.8 grade levels behind in math.
33 In a study by Quinn and Epstein (1998) the characteristics of children and families being served by local interagency systems of care in a large suburban county outside of Chicago were reported. Of the 238 children and youths that participated in the study, 42% were identified as having an emotional or behavioral disorder as either their primary or secondary educational disability. The majority of the participants were white (77%; 10% African American, 8% Hispanic) males (75%) and about half of the children lived with one or both parents. Despite the fact that most scored in the average intelligence and IQ range, 27% of the children had experienced course failure and 14% had been retained. More than half of the children and youth had accessed special education services (79%), juvenile justice services (63%), or mental health services (57%). From these studies several consistent characterizations have emerged. With regard to gender and ethnicity, most children with emotional disturbances are males and most are functioning below grade level despite scoring in the low-normal range of intelligence. Their grade point averages are lower than children in other disability categories and significantly lower than students not served in a special education setting (Cullinan et al., Koyangi & Gaines, 1993; Peacock Hill Working Group, 1991). Children with emotional disturbances have been found to receive special education services for longer periods of time, spend the majority of their schooling in restrictive educational settings, and have a significantly higher dropout rates and lower graduation rates than other disability groups. They also have a higher likelihood to interact with the juvenile justice system. Longitudinal studies that examined the characteristics of children with
34 Table 5 Summary of Characteristics and Functioning of Children and Youth with Emotional Disturbances Study Sample GenderAge Ethnicity Functioning Outcomes over Time National Adolescent And Child Treatment Study (Greenbaum et al., 1998; Greenbaum et al., 1996; Silver et al., 1992) 812 at Entry 628 at Year 7 Youth with emotional disturbances served in schools or residential settings 75% Male Avg.: 14 yrs. White: 70% Black: 22% Hispanic: 5% Average IQ: 85.8 Math: 93% below grade level Reading: 59% below grade level Average CBCL Total Score: 69.6 Average Vineland Composite: 78.0 At Year 7: Math: 94% below grade level Reading: 85% below grade level Improved 1.84 T-score points per year Declined 1.00 point each year Nationally representative sample of children and youth served in public schools (Cullinan et al., 1992) 269 children and youth identified with emotional disturbances 79% Male Ages 12-17 White: 76% Black: 22% Hispanic: 2% Average IQ: 93 Special Education Placement: Self-contained: 39% Resource: 23% Consultation: 19% Alternative school: 14% Res/Homebound: 2% Avg. Hrs. Mainstreamed: 13/week Related services: 70% --(Continued on next page)
35 Table 5 (continued) Study Sample Gender Age Ethnicity Functioning Outcomes over Time Alternatives to Residential Treatment Study (Duchnowski et al., 1998) 163 at Entry 144 at Year 1 Youth with emotional disturbances served in five innovative community-based programs across the country 66% Male Avg: 14 yrs. White: 65% Black: 14% Hispanic: 9% Native American/ Alaskan: 11% Asian Pacific Islander: 1% Average IQ: 84 WRAT-R Math (Raw): 23.77 SORT Reading (Raw): 125.05 CBCL Total: 71.8 Avg. Total Score on CAFAS: 14.34 WRAT-R Math (Raw): 25.35 SORT Reading (Raw): 136.01 CBCL Total: 68.31 Avg. Total Score on CAFAS: 10.85 National Longitudinal Transition Study (Wagner, 1995) Nationally Representative sample of 8,000 children and youth with a subsample of 777 children with emotional disturbances 76% Male Ages 13-21 White: 67% Black: 25% Hispanic: 6% Math: average 1.8 grade levels below Reading: average 2.2 grade levels below Higher dropout, higher arrest, and lower employment rates than any other disability category and nondisabled peers (Continued on next page)
36 Table 5 (continued) Study Sample Gender Age Ethnicity Functioning Outcomes over Time Sample of children and youth served by local interagency systems of care in a large suburban county outside Chicago (Quinn & Epstein, 1998) 238 children and youths. 42% (N=99) identified with emotional disturbances as the primary or secondary educational disability 75% Male Avg.: 15 yrs. White: 77% Black: 10% Hispanic: 8% (For total sample) Average IQ: 94 Failed a course: 27% Retained: 14% Depression: 31% Out-of-Home Placement: 90% Psychiatric Hospital Stay: 62% Ever Adjudicated: 68% --
37 emotional disturbances with regard to their academic, behavioral, and adaptive functioning have yielded mixed results. The NACTS study found improvements in emotional and behavioral functioning while findings in academic and adaptive functioning were not encouraging. The children in the ARTS study showed improvement in all functioning areas (i.e., academic, behavioral, emotional, and adaptive). These results demonstrate that while children with emotional disturbances display challenges in many areas, they are capable of improving over time. With evidence that children with emotional disturbances can achieve, why are the outcomes for these children not better? Mental Health Service Utilization of Children with Emotional Disturbances A frequent finding in the literature is that children in need of mental health services do not get them (Friedman et al., 1998; Knitzer, 1996). About 75 to 80 percent fail to receive specialty mental health services, and the majority of these children fail to receive any services at all (U.S. Department of Health and Human Services, 1999). The present childrens mental health service system is inadequate to provide the multifaceted assistance children with emotional disturbances need. Their needs are dynamic and require services from a number of agencies including mental health, child welfare, juvenile justice, and special education (Burns, 1991; Friedman, 1995; Knitzer, 1996). Currently, there is considerable evidence that many children with emotional disturbances are not receiving appropriate or sufficient mental health services. According to Mental Health : A Report of the Surgeon General (1999), 70% of children and adolescents in of need of mental health treatment do not receive services. Other findings in the report: only one in five children with emotional disturbances used specialty services; 70% of children with a diagnosis and impaired functioning received mental
38 health services from their school; and for nearly half the children with emotional disturbances who received services, the public school system was the sole provider. Schools as the primary or sole mental health provider was corroborated by the Great Smoky Mountain Study in North Carolina, investigating a community with an enriched service system (Farmer, Stangl, Burns, Costello, & Angold, 1999). The results from the study indicated that only 40% of youth with emotional disturbances received any specialty mental health services in the course of one year and the majority of services that were received were provided in the schools. Burns (1995) showed that children, in need of services, were receiving services in schools (70%), specialty mental health facilities (40%), health sector (11%), child welfare (16%), and juvenile justice (4%). For nearly half the children with emotional disturbances who received services, the public school system was the sole provider. Hoagwood and Erwin (1997) supported the findings through a review of other studies; concluding that schools were the primary providers of mental health services for children. Expanding upon the research base, Marcenko, Keller, and Delaney (2000) investigated the service needs, expectations, and use of children with emotional disturbances and their families in an urban area. Families reported, on average, they needed 17 different services. Services receiving the highest responses were focused on children (recreational opportunities, counseling and support services for the children). Parent or caregiver services reflected needs for parent training, counseling, and employment-related programs. Although parents wanted assistance in changing their childs behavior, educational services were noted least frequently.
39 Referrals to service providers play a large part in determining the number of children and families that utilize the agency services. In a study of 696 children and their families, living in urban areas, significant facts emerged regarding the characteristics of those children referred for services. Black children and their families were referred for services at the highest rates by juvenile justice, social services, and the school system as compared to all of the other ethnic/racial groups. The urban public school system also had the highest percentage of referrals of younger children. These children had higher levels of impairments, had moderate histories of previous service use, and had higher risk factors than other children. The study suggests that schools may be ill equipped to handle the additional services that children with emotional disturbances and their families may need. Results from the School and Community Study (Kutash, Duchnowski, Rivera, Oliveira, & Kelly, 1997) indicated that the majority of students used school-based services (81%) for their emotional and behavioral problems. Schools in this study were forced to employ or make available those services necessary to meet the needs of students with emotional disturbances (e.g., individual counseling, group counseling, case management, and medication monitoring). In another study, Quinn and Epstein (1998) found that more than half of the students had accessed special education services (79%), juvenile justice services (63%), mental health services (57%), and child welfare (45%). The high percentage of students relying on schools to provide special educational and mental health services is unmistakable.
40 Educational Services Children with emotional disturbances receive special services in school when they are formally identified under the category of seriously emotionally disturbed. Without formal identification and eligibility, these children are very likely to remain in a regular education class with little or no assistance provided to the teacher in dealing with his or her behavioral or emotional problems, no matter how severe (Forness, Kavale, & Lopez (1993). Overall, less than 1% of all school-aged children are identified by school systems as having emotional disturbances (U.S. Department of Education, 1996) compared to the generally accepted prevalence rate of approximately 3% (Federal Register, 1998). The School and Community Study (Kutash et al., 1997) investigated the effects of school restructuring and reform activities on outcomes for students who were identified as having emotional disturbances. In the 10 schools selected as actively using models of restructuring and reform, 16% of all students received special education services. Three percent of the students were identified as having emotional disturbances as compared to the national average of less than 1% (U.S. Department of Education, 1996). These actively engaged schools appear to be identifying children with emotional disturbances at a rate more consistent with the most recent prevalence estimate (Federal Register, 1998). In the actively engaged schools, the majority of the students received an array of services from multiple agencies with 81% receiving school-based services and 78% receiving outpatient services some time in their life for their emotional and behavioral challenges. School Reform Schools have engaged in educational reform for most of the 20th century (Cuban, 2000). During this time, educators have grappled with designing a model for the best
41 school environment, which would result in improved student outcomes. However, a model has not emerged that has significantly impacted educational outcomes (Cuban, 2000; Frechtling, 2000). Some of the failure can be attributed to the gap between research and practice (Malouf & Schiller,1995). Attempts to implement models of reform have been met with many obstacles, including a failure to understand the model, unwillingness to implement, and inconsistent procedures. Any one of these barriers could yield failure (Vanderwood et al., 1998). In retrospect, failures could also be attributed to the manner in which the reform strategies were introduced to the educational community. The reform movement has been launched into the various education camps. The general education sector has developed models and strategies with the regular education student as its focal point, ignoring the needs of children with disabilities (Vanderwood et al., 1998). In response, special educators created their own reform strategies. These parallel processes fragmented schools, as teachers were following different agendas trying to reach the same ultimate goal, improved student outcomes. At last, an integrated or whole school approach to reform, combining the efforts of both regular and special educators, is emerging. The whole school reform movement embraces the total school population and improving student outcomes remains the goal. Not only were the reform strategies distinguishable by their educational sector (i.e., regular education and special education), there were variations in implementation based on geographical settings (i.e. urban versus suburban/rural). Urban schools focused their reform initiatives on basic skills with little success while suburban/rural schools put their energies into advanced level courses and programs (Carlson, 1998).
42 General Education Reform Characterized as the decade of reform, the 1980s saw a series or wave of efforts to improve a mediocre educational system. A Nation at Risk (National Commission on Excellence in Education, 1983) was the most notable study citing dismal academic outcomes. A series of reform efforts followed with a scope and vigor unparalleled in past efforts. Although seen by many as a single effort, reform has been described in terms of three waves. Each wave refers to a phase of the history: repair, restructuring, and inclusion. The early 1980s efforts were spent repairing the existing educational system. Problems were viewed as existing in the quality of the staff and in the educational tools utilized by the schools. The solution was to implement top-down initiatives, based on a theory of centralized controls and standards. This wave of reform drew criticism for failing to address the real problem, which was the educational system itself. Wave 2 efforts (1986-1988) focused on restructuring the entire system. There was movement towards empowering teachers and parents, decentralizing the school management system, and attending to topics neglected during the previous wave. The bottom-up approach of Wave 2 was criticized for the continued low academic outcomes and decline in parent involvement. Childrens policy became the hallmark of Wave 3 reform efforts. School focused policies were replaced with comprehensive service delivery systems. The uncoordinated, fragmented service systems were obsolete and children deemed at-risk or disadvantaged received attention for the first time. The Goals 2000 Act, signed into law by President Clinton on March 31, 1994, marked the culmination of the general education reform movement. Eight educational
43 goals are embodied in the act: school readiness, school completion, student achievement and citizenship, science and mathematics, adult literacy and lifelong learning, school environment, teacher education and professional development, and parental participation. The language of inclusion exists in the Goals 2000 legislation with terms such as, all children, all students, and students or children with disabilities. Thus in Wave 3, students with special needs gained federal government consideration in school reform efforts (Danielson & Malouf, 1994). The interests of the general and special education communities began to merge; however, an integrated system is far from being realized (Paul & Roselli, 1995). Special Education Reform Likened to the Tower of Babel (Mitchell, 1988), special education began to develop its own reform initiatives in response to poor student outcomes. They were according to Mitchell, constructing a similar vision of what schools should become, but they are not building it together (p. 49). The special education community was concurrently undertaking its own series of reform programs in response to several reports documenting the poor outcomes of children with disabilities, including the annual report to Congress on the Implementation of the Individuals with Disabilities Education Act (National Council on Disabilities, 1989) and the results of a longitudinal study of students with disabilities (Wagner, Newman, DAmico, Jay, Butler-Walum, Marder, & Cox, 1991). The special education reform initiatives like the general education reform movement occurred in waves, beginning with the Regular Education Initiative (REI) of the 1980s and culminating with the Inclusive Schools Movement (ISM) of the 1990s.
44 Passage of PL 94-142 in 1975 mandated a free, appropriate public education be provided for all students and that students be served in the least restrictive environment. In the context of increased special education centers and self-contained classrooms, the REI marked the first major reform effort impacting special education. The initiative proposed that pull-out programs, which educated students in settings other than a regular classroom, be abandoned and a more inclusive model of instruction be adopted (Kauffman, 1997). The Goals of REI included: (a) a merger of special and general equation into an inclusive model of schooling, (b) a substantial increase in the number of students with special needs being educated in mainstream classrooms, and (c) an increase in the academic achievement of students with mild to moderate disabilities and those without disabilities who experienced underachievement (Fuchs & Fuchs, 1994). The inclusion movement quickly followed. The Inclusive Schools Movement (ISM; Muscott, 1995) dominated the 1990s policy debates. As with the REI, the ISM attempted to partner the general education and special education sectors in a collaborative effort to support all students. The goals of the ISM are to eliminate the continuum of services, abolish special education, and focus on social competency (Fuchs & Fuchs, 1994). Critics of this movement argue that reforms in special education can only be achieved by separating the students with special needs, restoring and rebuilding the concepts of the field, and increasing the empirical base (Kauffman, 1993). Thus, the debate continues as to whether all students with special needs should and can be fully included in general education settings (Muscott, 1995).
45 Systemic Reform Integrated, whole school or systemic reform is the latest reform movement. Systemic reform policies emerge from federal, state, and district offices as directives to large numbers of schools and classrooms (Slavin, 2000). Frechtlings (2000) conceptualization of systemic reform includes three concepts: 1) a set of standards that includes high expectations for all students, 2) aligning all of the components of the educational system (i.e., curriculum, instructional materials, student assessment, educational policies, educational policies, professional development, and evaluation) with these standards, and 3) collaborative relationships between people and institutions based on shared decision-making, rather than hierarchical arrangements. Suggested in this definition are levels of impact. Level one is the student, level two is the school and the classroom, and level three is the educational system itself. A comprehensive evaluation of systemic reform must address each level and is about changing the system itself in ways that are sustainable and scalable (Frechtling, 2000). Developing a method to evaluate the complexities of systemic reform activities has proved challenging. Some of the challenges focus on the diversity of the disciplines and prospectives integrated in the system. Other issues confronting researchers are the multiple, simultaneous efforts with multiple outcomes (Knapp, 1995). As a result few empirical evaluations have been conducted. Evaluations are further complicated when researchers seek to investigate the outcome of children with special needs. Vanderwood, McGrew, & Ysseldyke (1998) found in their investigation of national education data sources that significant numbers of children with disabilities were excluded from data collection programs. Assessing the
46 outcomes for this population of students was impossible given the paucity of available information. What does this mean for students with disabilities? It means that efforts to integrate the general education philosophy of broad systematic reform with the special education philosophy of improving social and academic outcomes for students with disabilities has not yet been realized. The principles of an integrated, systemic school reform model have been embraced but remain regular education strategies. In a study of the relationship between students exposure to school restructuring efforts and changes in academic functioning and symptomatology (Rivera, 1999), exposure to restructuring efforts failed to significantly predict change in academic achievement, symptomatology, or functioning for students with emotional disturbances (after controlling for age, level of cognitive functioning, family income level, and school attendance). In a study by Shouse and Mussoline (2000), the disadvantaged schools in their study did not accrue any long-term achievement improvements and those recently adopting restructuring reforms appeared to lag substantially behind the other schools. Summary The geographical location or urbanicity of a school can have a significant effect on the educational outcomes of its students. Varying conditions between urban and suburban/rural settings contribute to the differences in academic achievement scores. These discrepancies have been well documented. Also, the overrepresentation of minorities in special education, especially in emotional disturbances classrooms has been equally well documented. Results from
47 research studies consistently show a larger percentage of minorities, particularly black males, being identified with emotional disturbances and served in special education settings. There is overwhelming evidence that these students tend to reside in high poverty, urban areas. Classrooms serving children with emotional disturbances, whether in urban or suburban/rural locations, are affected by school reform and restructuring initiatives. Despite the extensive literature base on school reform, there has been little empirical research on the effects of school reform on student outcomes of students with special needs. A research effort is needed that investigates the relationship between school reform and student outcomes and also looks at how urbanicity impacts that relationship.
48 CHAPTER THREE METHOD This chapter describes the methodology used to address the research questions posed in Chapter One. The chapter is organized into four sections. The first section describes the two studies from which the participants and schools for the current study drawn. The procedures used to generate matched pairs of students are described in the next section. This is followed by a description of the study variables and instruments used to measure the constructs of interest while the final section outlines the analytic strategies used to answer each research question. A summary of these sections closes the chapter. Purpose The differences between urban school environments and suburban/rural school environments have been well documented. These differences have dominated professional educators discussions and have produced a number of school reform and restructuring programs. However, schools have had to adopt and implement these reform strategies without the assistance of empirical support for their use. There is a paucity of research on the extent to which interventions with documented positive outcomes are used and the difference in utilization among schools in districts with widely financial and demographic characteristics (National Research Council, 2002, p. 4-1). The knowledge
49 base could greatly benefit from empirical results regarding educational reform and restructuring efforts and the effects on student outcomes. It has been well documented that the academic and social/emotional outcomes for children with emotional disturbances served in public schools are poor. This is despite the fact that there are numerous empirically supported interventions to meet their academic, social, and emotional needs. The knowledge base is also lacking empirical investigations of the relationship between school reform activities and the academic, social, and emotional functioning of students in special education settings due to emotional disturbances. The primary purpose of this study is to examine the academic, social, and emotional functioning of children with emotional disturbances and the relationship between this functioning and the level of school reform operating in urban and suburban/rural communities. The results of the current study will supply the field with much needed empirical information on both the activities of reform operating in urban and suburban/rural schools and the characteristics of students in special education due to emotional disturbances. Differences in levels of reform activities and the effects on the students in special education due to emotional disturbances are also explored. In order to investigate the relationships, the following questions are examined. Research Questions 1. Are there differences in academic functioning (i.e., reading and math achievement) between urban students in special education classrooms due to
50 emotional disturbances and a matched sample of suburban/rural students in special education classrooms due to emotional disturbances? 2. Are there differences in the psychological functioning (i.e., symptomatology and functional impairment) between urban students in special education classrooms due to emotional disturbances and a matched sample of suburban/rural students in special education classrooms due to emotional disturbances? 3. Are there differences in the mental health service utilization between urban students in special education classrooms due to emotional disturbances and a matched sample of suburban/rural students in special education classrooms due to emotional disturbances? 4. Are there differences in the school reform activities (e.g., governance, accountability, prosocial discipline, inclusion, family involvement, curriculum and instruction) between the urban public schools attended by students with emotional disturbances and the suburban/rural schools attended by students with emotional disturbances? 5. What is the contribution of school reform activities in explaining those differences (i.e., academic functioning, psychological functioning, and service usage) between urban students in special education classrooms due to emotional disturbances and suburban/rural students in special education classrooms due to emotional disturbances?
51 Data Sources The current research study used data from two studies conducted by staff at the Research and Training Center for Childrens Mental Health, a federally funded Center at the University of South Florida to increase the knowledge base on children with emotional disturbances. The research design for the two studies, the School and Community Study (Kutash et al., 1997) and the Urban School and Community Study (Kutash et al., 2001) are presented in Table 6 and each study is described in greater detail in the following sections. Table 6 Research Design of the Two Studies Methodology School and Community Study Urban School and Community Study Number of Participants1 115 2002 Number of Schools 10 202 Design Longitudinal Point-in-time Length of follow-up (in months) 24 Domains Measured: School Reform X X Demographics X X School Functioning X X Emotionality X X Mental Health service use X X 1Participants in these studies are restricted to students in special education due to emotional and behavioral disorders. 2Projected number of participants/schools at the completion of the study.
52 The School and Community Study The School and Community Study employed a multi-method, multi-source approach to examine the various aspects of school restructuring and special education reform. This study examined the outcomes of school reform on 115 students formally identified as having emotional disturbances and served in a special education classroom in one of ten suburban or rural schools participating in the study (Kutash, et al, 1997). A four-step nomination process was used to solicit schools actively engaged in reform and restructuring. The first step involved sending out a national call for nominations which invited individuals to nominate schools that were actively engaged in reform efforts. During the second step, nominated schools were asked to complete a screening questionnaire and based on their responses, some schools were eliminated. The remaining schools met the following criteria: (a) a regular public school as opposed to an alternative site; such as a special education center or specialized day treatment center serving students formally identified as having emotional and behavioral disabilities; (b) had identifiable reform and restructuring policies and procedures in place; (c) been engaged in restructuring activities for a minimum of two years; and (d) demonstrated genuine parent involvement, including parents of students with emotional and behavioral disabilities. The third step involved sending the remaining schools a second screening questionnaire requesting additional information about their reform and restructuring activities. Schools were then ranked according to their responses and the highest ranked schools were selected for site visits. The final step involved intensive onsite interviews
53 with key stakeholders (district level school administrators, state level officials, parents, teachers, and service providers). Ten of the 216 nominated schools were invited to participate in the study These ten schools were located in six states: Georgia (2 schools), Kentucky (2 schools), Iowa (1 school), Maryland (2 schools), Vermont (2 schools), and Wisconsin (1 school). Of these ten schools, there were two high schools, one middle school, five elementary schools, and one school with preschool or kindergarten through the 8th grade. Five of the ten study schools (50%) were located in rural areas and five (50%) were located in suburban settings, with one of the rural schools located on an American Indian reservation. These schools varied widely in their enrollments, ranging from 192 to 2,149 students, with an average of 763 students. In these ten schools, 16% of all students received special education services, and 3% were identified as having an emotional disturbance as compared to the national average of 13% in special education and less than 1% who were identified as having emotional disturbances (U.S. Department of Education, 1996). Only those students who were formally identified as having emotional disturbances by their school and served in a special education program were eligible to participate in the study. Of the 145 eligible children, 115 (79%) parents/caregivers returned signed consent forms (Kutash et al., 1997). Participants in the study were mostly male (81%), predominately white (79%), and on average were 11 and a half years old. The mean grade level was 5.4 with the greatest number of students being in the 5th and 6th grades. The majority lived in two-parent homes (57.8%), with 20% living with two biological parents, and 16% living with a biological mother with stepfather.
54 Approximately 13% of the students had been retained in grade at least once and about one-third had medical concerns and were on medications. On average, students were absent from school about 12 times a year and had an average of 14 total discipline incidents per year. The average IQ was in the low average range (70-89), 67% were below grade level in reading, and 72% were below grade level in math. The Urban School and Community Study This study is currently being conducted and will examine the impact of school reform and restructuring activities on approximately 200 children with emotional disturbances in special education classrooms that attend one of sixteen schools in urban areas across the nation. To date, eight schools in two urban cities and 99 students and their parents have participated in the study. The overall design of the study calls for the comparison of the academic outcomes and service utilization patterns of children with emotional disturbances attending urban schools actively engaged in reform activities with the academic outcomes and service utilization patterns of children attending urban schools less actively engaged in reform activities. All schools were selected based on the following overarching criteria: (a) serve students identified as having emotional disturbances, (b) be located in an urban area, (c) a regular public school as opposed to an alternative site; such as a special education center or day treatment center; and (d) serve at least 40% of its student body from ethnically and diverse backgrounds. The school selection and data collection procedures were similar to those followed in the School and Community Study except that this study is using a singlepoint-in-time design while the School and Community Study was longitudinal.
55 Nominations for schools actively engaged in reform and restructuring activities were solicited nationally with 37 schools from 13 states nominated to participate in the study. At the conclusion of the selection process, schools located in 5 urban areas remained. Once the active schools were located within a city, schools that were less actively engaged in reform and restructuring activities were selected with the assistance of district personnel. To date, complete data have been collected on eight schools in two urban areas with extensive data collected on 99 students in special education due to emotional and behavioral disabilities. Research staff visited each school twice in order to complete data collection. The first visit by research staff was to collect information regarding school reform and restructuring activities. This information was collected through a structured interview process and is described in greater detail in the instrument section of this chapter. On the second visit, information on the students in special education due to emotional and behavioral disabilities was collected. In order to ensure that data were only collected on students actively engaged in the school, each student had to meet the following criteria to be eligible for the study: (a) be over four years at the start of the school year; (b) actively attending the school for the 30-day period prior to data collection; and (c) enrolled in the school since the start of the school year or prior to January 15th of the current school year. Through interviewing parents and teachers, as well as reviewing student records, information regarding the students levels of academic achievement, rates of attendance and disciplines referrals, and basic demographic information were collected.
56 Students participating in the study were mostly male (83%), predominately black (83%), on average 13 years old. The mean grade level was 7.0 with the greatest number of students being between the 7th and 9th grades. The majority lived in single parent homes (57%), with 65% living with a biological mother. One third of the households (33%) fell at or below the poverty level as defined by the poverty thresholds provided by the U. S. Census Bureau. Seventy-one percent of the students received their school meals free. On average, students were absent from school 21 days for the year. The average IQ was in the low average range (70-89), 78% were below grade level in reading, and 92% were below grade level in math. Summary statistics on the participants in both studies are presented in Table 7. Table 7 Characteristics of Students Participating in the Research Studies Characteristic School and Community Study (N=115) Percentage Urban School and Community Study (N = 99) Percentage Gender Male 80.9 82.8 Female 19.1 17.2 Race White 79.1 16.2 Black 9.6 82.8 Hispanic 0.9 --Native American 9.6 --Other 0.9 1.0 (Continued on next page)
57 Table 7 (continued) Age 5-7 14.8 6.1 8-9 17.4 10.1 10-11 20.9 14.1 12-13 20.9 23.2 14-15 13.9 19.2 16-18 12.2 25.3 19-20 --2.0 Mean Age (SD) 11.6 (SD = 3.2) 13.1 (SD = 3.0) Family Structure Two parent home 58.3 28.3 One parent home 33.9 56.6 Cost of School Meal(s) Free 64.3 70.7 Reduced 2.6 8.1 Full Price 33.0 21.2 Participants The participant pool for the current study is made up of students who participated in either of the studies just described. Four research questions for the current study require a comparison between the characteristics of students in urban schools and students in suburban/rural schools. This required students from the urban schools to be paired with students in the suburban/rural schools on several demographic variables so the areas of interest are not clouded by extraneous factors. A hierarchy of demographic variables was used to match students from the urban schools with students from the suburban/rural schools. The demographic variables that students were matched include :
58 gender family income level age of student Using gender as categorical variable, students can be easily paired. The first step in the matching process, therefore, is to generate a data set of all male students from the Urban School and Community Study and a list of all male students from the School and Community Study. The same two data sets were generated for female students. With this variable (gender) isolated, the youth can be further matched on the variables of family income and age. Both family income and age are continuous variables. Having two students with exactly the same yearly family income and age would be difficult, if not impossible. Therefore, youth with similar family incomes and ages were matched using two methods. The first method generated a visual representation of these two variables for the potential males and females in the study. As can be seen in Figure 1 and Figure 2, graphs for both males and females were generated with yearly family income and age plotted on the Y and X axis, respectively. With these graphs, students from each study can be placed and the distance between students can be examined. As can be seen in Figure 1, it appears that a female student in the urban study (U01) is close in yearly family income and age to the female in the suburban/rural study (R09). The same graph can be created for the male students in the two studies. As seen in Figure 2, it appears that student U23 is close in yearly family income and age to R04.
59 Yearly U01 Family R09 FEMALES Income R04 U13 Student Age Figure 1 A graphic representation of the matching method for female students in the two students (R = Rural, U = urban). Yearly Family *U23 Income R04 MALES *U45 R10 Student Age Figure 2 A graphic representation of the matching method for male students in the two students (R = Rural, U = urban). All participants were placed in the graphs just described and a student from each study was paired with a student in the other study that is similar in yearly family income and age. In addition to this visual matching of students, an analytic technique will also be employed to ensure the best possible matches are made.
60 To ensure that the space was minimized between students when matching on family income and age, Mahalanobis Distances was also calculated on the distance between students on these graphs shown in Figure 1 and Figure 2. Mahalanobis distance is a technique used to measure the distance between two points in the space defined by two or more correlated variables (Stevens, 1996). Calculation of these distances did not add more precision to the matching process than the visual matching process; therefore participants included in this study were visually paired. Once all students are matched or paired, independent t -tests were conducted between the group of students from the urban study and the group of students in the suburban/rural study on the variables of age and yearly family income. These t -tests are necessary to ensure the two groups of students, on average, do not significantly differ from one another. Likewise, a Chi-square analysis was calculated on gender. Schools The schools these youth attend can also be considered participants. Each school was described in detail, for example, the number of students attending, grades covered, number of teachers, and the number of special education students. Additionally, the implementation of school reform techniques in the areas of (1) governance, (2) accountability, (3) curriculum and instruction, (4) parent involvement, (5) special education practices, and (6) pro-social discipline methods were described through the School Reform and Restructuring Index (SRRI) conducted at each school. The SRRI is described more fully in the study variables and instrumentation section of this chapter.
61 Study Variables and Instrumentation This study has several study variables that cluster into five areas. These areas include: demographic information, academic achievement, psychological functioning, mental health service use, and school reform and restructuring. Each of these areas is discussed below as to how the information about this area was collected, the instruments used, and the range and interpretation of the scores generated. The domains, sources and instruments used in the two studies are summarized in Table 8. Table 8 Domains, Sources, and Instruments Used in Both Studies Domains Source Demographic Information Academic Functioning Psychological Functioning Mental Health Service use School Reform and Restructuring School record review X X Staff interview X SRRI Parent interview X CBCL CAFAS/CIS CASA SACA Student interview WRAT X = standardized protocol designed for the current study. CAFAS used to measure impairment in the School and Community Study CIS used to measure impairment in the Urban School and Community Study CASA used to measure Mental Health service use in the School and Community Study SACA used to measure Mental Health service use in the Urban School and Community Study
62 Demographic Information Through both review of student records by either research staff or school staff and interviews from parents, basic demographic information was gathered about each student. This information included students date of birth, gender, race, level of family income, and family composition. From the date of birth information, the age at which the student was interviewed was calculated. Therefore, the age variable in the current study refers to the age of the student at the time of participation in the study. For the School and Community study, which was longitudinal in design, age refers to the first time the student participated in the study. Level of family income was measured differently in the two studies. For the School and Community Study, income information was collected as a categorical variable (i.e., Which of the following categories best represents how much income your family brings in a month?). In the Urban School and Community Study, this information was collected as a continuous variable (i.e., What is your familys annual income?). To match students on this variable, information on income were converted to the same metric (e.g., annual income) and the mid-point for each of the categories were used as a proxy for the actual income for those students participating in the School and Community Study. Academic Achievement Wide Range Achievement Test In both studies, either research staff or school staff administered the reading and math portions of the Wide Range Achiev ement Test (WRAT-III; Wilkinson, 1993) to obtain a standardized measure of achievement on each student. The WRAT-III is an
63 individually administered standardized instrument designed to assess academic achievement in arithmetic, reading, and spelling in individuals aged 5-75 years. There are two equated forms (Blue Form and Tan Form) that can be used individually or together to provide a more comprehensive evaluation of skills (Combined Form). The arithmetic subtest assesses the ability to count, read number symbols, solve oral problems, and perform written computations. The reading subtest assesses the ability to recognize and name letters and pronounce words out of context while the spelling subtest assesses spelling ability from dictation. The WRAT-III subtests yield four types of scores for each of the participants, including standard scores, percentiles, grade equivalents, and absolute scores. Standard scores (M =100, SD =15) can be used to calculate achievement levels and compare the individuals score to the normative sample of nearly 5,000 people. Psychometric properties of the WRAT-III have been well documented. Four sets of reliability indices were calculated: coefficient alpha, alternate form, person separation, and test-retest. Internal consistency as measured by median test coefficient alphas ranged from .85 to .95 over the nine WRAT-III tests (3 subtests x 3 forms). Alternate form correlations were .92 for reading, .93 for spelling, and .89 for math. Rasch person separation indices ranged from .98 to .99 for the nine WRAT-III tests. Finally, the test-retest method was used to measure the stability of the WRAT-III and yielded corrected stability coefficients that ranged from .91 to .98 on the nine tests given to a subsample of the norm group. The WRAT-IIIs content validity was measured by item and person separations using the Rasch analysis of tests (Wright & Stone, 1979). Item separation indicates how
64 well items define the variable being measured, while person separation indicates the tests capacity to distinguish among a sample of persons on the basis of the total number of items answered correctly. For each test of the WRAT -III, the highest item separation score possible of 1.00 was found. This provided strong evidence that there is content validity on each of the WRAT-III subtest. Several indices of the WRAT-IIIs construct validity include: the skills measured by the WRAT-III are developmental in nature; the various skills assessed by the WRAT-III are related to one another b ecause they are measures of cognitive ability; the WRAT-III has a close relationship with the earlier version, the WRAT-R; the WRAT-III is similar to other standardized instruments of academic achievement; and the WRAT -III is sensitive to differences of academic skill within the school population. The WRAT-III is a widely used assessment instrument in educational settings and has been used extensively with at-risk students and those with emotional disturbances. A discriminant analysis was also conducted on the WRAT-III using special education students and a matched control sample of students from the norm data. Results showed significant differences at the .001 level indicating that the WRAT-III can successfully group this sample at a 68% confidence level. Results from the School and Community study indicated an average reading score of 86.6 (SD = 17.4) and an average arithmetic score of 86.8 (SD = 15.3) for students. Achievement scores were also reported in the Urban School and Community study, with an average reading score of 78.4 (SD = 16.6) and an average arithmetic score of 74.5 (SD = 12.0) obtained.
65 Psychological Functioning Several instruments were used to measure the psychological functioning of the participants. The Child Behavior Checklist (CBCL) was used in both studies to measure levels of psychopathology. To measure the amount of impairment the youth experienced due to having emotional disturbances, the Child and Adolescent Functioning Assessment Scale (CAFAS) was used in the School and Community Study and the Columbia Impairment Scale (CIS) was used in the Urban School and Community Study. The CAFAS and CIS vary in format but measure the same construct, functional impairment. The CAFAS is a longer, more detailed survey while the CIS is a shorter, more streamlined survey. While there are no studies in the literature that have measured the correlation between scores on these two measures, several experts in the field of childrens mental health research agree that both instruments essentially measure the same construct. A trained data collector administered these instruments to a parent or caregiver of the student either in person or on the phone. Child Behavior Checklist The Child Behavior Checklist (CBCL; Achenbach, 1991), developed in 1983 and revised in 1991, is one of the most frequently used measures of problem behavior in child psychopathology and provides a standard against which the validity of other instruments are often measured. The CBCL is an individually administered instrument designed to measure childrens competencies and behavior problems compared to a national representative sample. The normed sample reflected the United States population in 1991 in terms of ethnicity, socioeconomic status, and region and urban-suburban-rural
66 residence. The parent/caregiver provides information on 20 competence items and on 118 problem behaviors using a rating scale for how true the item is of the child now or within the past six months. The CBCL yields normalized T scores (M = 50, SD = 10) and percentiles on numerous scales: three competence scales, a total competence scale, and eight syndrome scales (Withdrawn, Somatic Complaints, Anxious-Depress, Social, Thought, Attention, Delinquent Behavior and Aggression). The Externalizing Behavior scale includes antisocial and aggressive behaviors such as stealing, truancy, fighting, and running away; the Internalizing Behavior scale includes withdrawn and anxious behaviors such as fearfulness, worrying, crying, and feelings of worthlessness. The Total Problem Behavior scale includes social problems and attention problems in addition to items from the Externalizing and Internalizing Behavior scales. A T-score above 63 is considered to be in the clinical range and a score between 60 and 63 is considered borderline. The psychometric properties concerning reliability and validity of the CBCL have been well established and reported in several studies (Achenbach, 1991; Dedrick, Greenbaum, Friedman, & Wetherington, 1997). The CBCL is a widely used instrument for assessing the psychological functioning of four to eighteen year old males and females (Achenbach, 1991). The checklist has been used extensively in a number of settings with children experiencing emotional disturbances (see for example, Duchnowski et al., 1998; Greenbaum et al., 1998). The CBCL was administered and completed as part of the data collection process in both the School and Community study and the Urban School and Community study.
67 Child and Adolescent Functional Assessment Scale In addition to the CBCL, the Child and Adolescent Functional Assessment Scale (CAFAS; Hodges, 1989) was administered to measure the amount of impairment the youth experienced due to having emotional disturbances. In contrast to measures of symptomatology such as behavior checklists that measure the number or frequency of symptoms, the CAFAS indicates the level of impairment or how the youths emotional, behavioral, or substance use problems interfere in various life roles, such as student, family member, friend, or member of the community. Thus, behavioral checklists and symptomatology inventories may be only moderately correlated with the CAFAS because the number of symptoms is not necessarily equivalent to the level of functional impairment (Hodges, Doucette-Gates, & Liao, 1999). The CAFAS is a multidimensional measure comprised of items that describe behaviors organized into eight domains of functioning (i.e., Role Performance in School, Home and Community, Behavior Toward Others, Moods/Emotions, Self-harmful Behavior, Substance Use, and Thinking). A parent/caregiver gives a score that best describes the severity of the youths behavior on a particular domain: 30 for severe (severe disruption or incapacitation), 20 for moderate (persistent disruption or major occasional disruption of functioning), 10 for mild (significant problems or distress), and 0 for minimal or no impairment (no disruption of functioning). The total score refers to the sum of the five subscales with a range from 0 to 150 with a higher score reflecting greater impairment. An individual whose total score equals 40 or above is considered to be in the clinical range of functioning (Walrath, Nickerson, Crowel, & Leaf, 1998)
68 There is considerable psychometric data available on the CAFAS. Reliability for the scales has been assessed by examining the level of agreement between trained raters and a criterion value and agreement among raters (Hodges, Bickman, Kurtz, & Reiter, 1991). Internal consistency, test-retest reliability, and inter-rater reliability were at satisfactory levels (Hodges & Wong, 1996). Hodges and Wong (1996) reported evidence for construct, concurrent, and discriminant validity on the CAFAS using data from the Fort Bragg Demonstration Evaluation Project. Construct validity was supported by significant zero-order correlations found between the CAFAS Total Score and four other related measures. Concurrent Validity was evidenced by significant positive relationships between CAFAS and independent ratings reported by parents, teachers, and youth of specific problem behaviors (i.e., interpersonal problems, risk behaviors, involvement with juvenile justice, and problem behaviors at school). Discriminant validity was supported by the report of higher CAFAS scores related with individuals being served in the Fort Bragg Demonstration Evaluation Project as compared to those being served in outpatient settings. For the purposes of the School and Community Study, the parent report form of the CAFAS was modified for use in telephone interviews. The adapted version, the CAFAS-Research instrument, is not a new version of the CAFAS but gathers only the information necessary to determine an impairment score for each domain, thus reducing the respondents burden. Reliability of the CAFAS-Research instrument was confirmed in an analysis by Kutash, Oliveira, and Rivera (1996).
69 Columbia Impairment Scale The Columbia Impairment Scale (CIS; Bird, Shaffer, Fisher, Gould, Staghezza, Chen, & Hoven, 1993) replaced the CAFAS in the Urban School and Community Study to measure levels of impairment the youth experienced due to emotional or behavioral disabilities. The CIS is a global measure of functional impairment in four domains: interpersonal relations, psychopathology, job or schoolwork, and use of leisure time. The CIS is a fully structured questionnaire consisting of 13 items that requires a parent/caregiver to rate how problematic each behavior has been for the child in the past year. A single score is yielded that can range from 0 to 52. A score of 16 or above is considered to be in the clinical range of impairment. Studies have shown adequate reliability through examinations of internal consistency. Further support for reliability was done through test-retest reliability. Bird et al. (1993) report validation of the CIS through concurrent and discriminant validity using small samples of a general population and a clinical group. Concurrent validity was determined based on its relationship to the Child Global Assessment Scale (CGAS). Additional concurrent validity was evidenced through moderate agreement with the CBCL. Differences between CIS mean scores for a community sample and a clinical sample were used to determine moderate discriminant validity. The largest descriptive sample used for the validity studies was deemed diverse ethnically, socioeconomically, by gender, and urban-suburban-rural residence.
70 Mental Health Service Use Three instruments (Teacher report, CASA, SACA) and two sources (parent and teachers) were used to collect information on the types of mental health services used and the intensity of these services (i.e., daily, weekly, or monthly). The lead special education teacher at the school or the school social worker was asked to provide information on services provided during the school day. Parent/caregivers were asked to complete either the Child and Adolescent Service Assessment (CASA) or Service Assessment for Children and Adolescents (SACA) through structured interviews conducted by study staff. These instruments asked parent/caregivers to report on mental health service use over the youths lifetime and during the last six months. Teacher Report For both studies, the lead special education teacher or school social worker reported on any related services the student may have received at the school during the school day. The categories of mental health services that special education teachers were asked to report on included: individual counseling, group counseling, case management, medication monitoring, and other services. Further, the services were categorized as either provided by a school employee (e.g., school psychologist, guidance counselor) or provided by a person employed by an agency (e.g., mental health therapist, social services counselor). The number of times a week the services were received was also recorded.
71 Child and Adolescent Services Assessment The Child and Adolescent Services Assessment (CASA; Burns, Angold, Magruder-Habib, Costello, & Patrick, 1996) is a parent-report instrument designed to assess the utilization of mental health services by children ages 8 to 18 years. It works well in diverse ethnic and cultural groups and it is the only self-report instrument of childrens mental health service use with documented psychometric properties. Services are broadly defined to include 33 settings organized under five broad categories of service: overnight/inpatient treatment, outpatient mental health services, other professional help, nonprofessional help, and other services. Several different services are listed within each broad category. For example, the overnight/inpatient treatment category has nine different services (e.g., medical inpatient unit, group home, therapeutic foster care, or boarding school). If the child has used a service, information is obtained on (a) the date the child first used the service; (b) his or her perceived benefit of the service; and (c) use of the service in the recent past, which is defined as the last 3 months (modified to last 6 months for both the School and Community Study and the Urban School and Community Study) (see Table 9). The reliability and validity of the scores from the CASA have been examined using clinical samples. Test-retest reliability was assessed using a sample of 77 children who were new admissions to either an outpatient clinic or an inpatient facility (Farmer, Angold, Burns, & Costello, 1994). Concurrent validity was assessed by comparing CASA data with data from a mental health centers management information system. The validity sample included 56 children and 50 parents who were participating in a
72 research project associated with the Robert Wood Johnson Foundations Mental Health Services Program. Table 9 Services Assessed by the Child and Adolescent Services Assessment (Burns et al., 1996) Overnight/Inpatient Treatment Outpatient Mental Health Treatment Other Professional Help Non professional Help Other Services Psychiatric hospital In-home counseling School guidance counselor/ psychologist/ social worker Crisis hotline Case manager General hospital psychiatric unit Outpatient drug or alcohol clinic Special class Self-help group Respite services Inpatient alcohol or drug treatment unit Mental health center Educational tutoring Adult relatives/ friends Medical inpatient unit Community health center Social services Residential treatment center Crisis center Probation officer Detention center/training school/jail Day treatment program Family doctor Group home or emergency shelter Private professional help Hospital emergency room Therapeutic foster care Religious counselor/ alternative healer Boarding school
73 for Youth in Western North Carolina (Ascher, Farmer, & Burns, 1996). The percentage of children receiving a service according to mental health center records was compared to the report of this service on the CASA. Results indicated the validity was quite good, although nonintensive services tended to be unreported. In the School and Community Study, the majority of the students had used at least one type of outpatient mental health service during their lifetime (78%) and 48% of the students had used at least one service within the last six months. Less than half of the students reported using overnight/inpatient treatment (35%) and of those students, 17% had used the service in the past six months. All of the study participants had utilized some type of professional service (e.g., school-based related services, tutor, probation officer) during their lifetime and 95% had used at least one of the services within the last six months. Service Assessment for Children and Adolescents. In response to the need for an assessment instrument to assess underinvestigated mental health service issues, the CASA was modified, creating the Service Assessment for Children and Adolescents (SACA ; Stiffman, Horwitz, Hoagwood, Compton, Cotler, Bean, Narrow, & Weiz, 2000). The SACA is designed to assess the types of mental health services children use, the treatments they receive within service settings, the reasons for service use, and the quality of the services. Service utilization is examined by the SACA across three service categories: inpatient (e.g., hospital, residential treatment center, and group home), outpatient (e.g., community mental health center and day treatment), and school (e.g., special education and school counseling). The services
74 can be provided in a variety of settings: public sector (e.g., juvenile justice, mental health center), private providers (e.g., physicians, psychiatrists), and informal, personal, and community resources (e.g., minister). A modified version of the SACA, focusing on inpatient and outpatient services, was used in the Urban School and Community Study (see Table 10). Table 10 Services Assessed by the Service Assessment for Children and Adolescents (Stiffman et al., 2000) Inpatient Outpatient Psychiatric hospital Community mental health center Psychiatric unit in a general hospital Psychologist/ psychiatrist/ social worker/ family counselor Drug or alcohol treatment unit Day or partial hospital Residential treatment center Drug or alcohol treatment unit Group home In-home therapist/counselor Foster home Emergency room Detention center/jail Family doctor Emergency shelter Probation officer/ court counselor Religious counselor Alternative healer Acupuncturist/chiropractor Self-help group Respite care provider
75 Service utilization history was measured by asking the parent/caregiver if the child had ever received inpatient and/or outpatient services for behavioral, emotional, or substance use problems. Additional information such as the age the child first used the service, the number of total life uses, and recent use (within last year) was collected to establish a service history. The SACA had good test-retest reliability for both lifetime service use and previous 12month service usage. Validity studies showed moderate evidence of predictive validity. In the Urban School and Community Study, more than half of the students had used an inpatient service during their lifetime (52%) and almost all of those students had used an inpatient service in the past year (91%). Of the students using inpatient services (N = 42), 19 students had used the services in the past year with an average of 1.6 different service types used during that time. Outpatient service use found 88% of students had used outpatient services with 62% of these students having used the services within the past year. For the 61 students (N= 87) who had used outpatient services in the past year, an average of 2.1 different service types had been used during that time. School Reform and Restructuring Index (SRRI) A special technique has been developed to measure the amount and types of school reform operating within a school by Kutash and Duchnowski (see Kutash, 1999). The approach is based on the exploratory case study methodology proposed by Robert Yin (1993) and is generalizable to a theory, not to a population. Research staff began by reviewing the literature on school reform and by documenting several recurring themes.
76 These themes were re-written as six propositions that characterize the elements of school reform and restructuring. (see Table 11) Table 11 Reform and Restructuring Propositions Area Proposition 1. Governance There will be evidence of de-centralization of authority from the district to the school building level. At the building level, there will be evidence of shared decisionmaking between the principal, the faculty and parents. Schools will have mechanisms such as an Advisory Council that will engage in shared decision making. There will be evidence that issues specific to children who have emotional and behavioral difficulties are discussed by those councils. 2. Accountability A system of measuring outcomes were developed and implemented. Such a system may be part of state or district level mandates, however, there will be evidence at the school building level of a commitment to demonstrate student progress. For example, an annual school report card will be presented to the community. Authentic assessment techniques such as portfolios, criteriareferenced tests, and other methods of documenting the educational progress of children were evident. Additionally, test scores of children with emotional and behavioral disabilities were included in the overall school averages. 3. Curriculum and Instructional Reform There will be evidence of systematic reform in instruction, both in regular and special education. There will be sustained activity to improve the instruction if children through the adoption of innovative techniques and instructional models. Examples of such evidence are multi-age grouping, instructional teams, and consultativecollaborative models of special education. These models were used in both regular and special education. (Continued on next page)
77 Table 11 (continued) 4. Includedness There will be evidence that all school staff, including regular and special education teachers, share the value that children with emotional and behavioral disabilities should be educated in a community school setting. School staff will exhibit a high degree of shared responsibility for the progress of all students, including children with emotional and behavioral disabilities. There will be a support mechanism in the schools to help achieve this includedness. 5. Parent Involvement A high level of parent involvement will be evident particularly for parents of children who have emotional and behavioral disabilities. Parent involvement should consist of more than attendance at school functions. There will be evidence of collaboration between parents and teachers in the education of their children. 6. Pro-Social Discipline The school will discipline their students using strategies such as conflict resolution, peer counseling, and other ways that enable students to learn from the experiences. There will be evidence that discipline will be handled in a positive manner that is individualized for all students, including students with emotional behavioral disabilities. At each participating school, semi-structured interviews were conducted with five members of the school: 1) the principal, 2) School Advisory Council (SAC) member, 3) a general education teacher, 4) a special education teacher, and 5) a person knowledgeable about the schools activities (e.g., assistant principal). The interviews were transcribed and sported into indicators that reflected a significant part of the proposition. Five raters, independent of the study staff, reviewed the interviews and determined if the interviews provided evidence as to whether there was strong, moderate, or mild evidence that supported the proposition or strong, moderate, mild, or no evidence
78 against the proposition. School restructuring and reform index (SRRI) scores were generated for each school by averaging the ratings for each indicator across raters and these averages were added together to obtain a total of six proposition scores. Further, these six proposition scores were added together to obtain a SRRI score for each school. SRRI scores for schools in the School and Community Study ranged from 54 to 54 and reflected the level of reform and restructuring for a school. Studies conducted by the research staff documented the reliability and validity of the SRRI (Kutash, Oliveira, & Robbins, 1997). Intraclass correlation coefficients (ICCs) were computed to determine the number of raters and the reliability of the ratings. The ICC results provided evidence of reliability of the SRRI. Further support of the reliability of the SRRI was conducted through test-retest analyses. Discriminate validity was established by determining that the SRRI could differentiate between those schools nominated to be actively restructuring and those schools, which were less active. At the conclusion of the School and Community study and before the Urban School and Community Study, some items were added to the SRRI while other items were re-written to increase clarity. These improvements changed the range of scores for the SRRI. The scores for the SRRI used in the Urban School and Community Study can range from 72 to 72. Therefore, the scores on the SRRI used in the two studies are not on the same matrix. The two different SRRI scores were converted to Z scores before inferential statistics are performed.
79 Research Design and Analysis This study was a secondary analysis of data collected as part of the School and Community Study and the Urban School and Community Study conducted by staff at the Research and Training Center for Childrens Mental Health, Louis de la Parte Florida Mental Health Institute at the University of South Florida. Measures of academic and behavioral outcome data utilized in this study were taken from the first year of data collection for the School and Community Study and from the first eight schools participating in the Urban School and Community Study. The current study seeks to explore the relationship between levels of urbanicity and school reform and restructuring activities and their impact on academic and emotional functioning. Research Question 1 The first research question examines the academic functioning of youth in special education settings in urban schools as compared to youth in special education settings in suburban/rural schools. Standardized scores from the Wide Range Achievement Test (WRAT) in reading and math were used in this analysis. Because children from the two research studies have been matched on several demographic variables (i.e., gender, family income, and age), dependent ttests were conducted on these scores to determine if the two groups of youth are statistically different from one another. Research Question 2 This research question centers on the differences in psychological functioning between the two groups of students. The Child Behavior Checklist (CBCL) is the
80 measure of psychopathology used in the two core studies from which the current sample was drawn. The CBCL yields a total score and scores for both the internalizing and externalizing aspects of emotional functioning. All three scores of this instrument were analyzed to address this question through the use of dependent t-tests. Additionally, the Child and Adolescent Functioning Assessment Scale (CAFAS) was used in the suburban/rural area study while the urban area study used the Columbia Impairment Scale (CIS) to measure the amount of impairment in a youth due to emotional or behavioral disabilities. While each scale measures the same construct (i.e., impairment due to emotional disturbance), the relationship of these scales with each other has yet to be fully investigated. As such, the results from the two instruments were discussed and compared descriptively before any further analyses were performed. Research Question 3 This question investigates whether there is any differential use of mental health services between the two groups of students. Within both studies, parents (caregivers) were asked what mental health services were used in the last 6 months while teachers were asked what mental health services were provided in the school during the school day to the youth. The services reported by both parents and teachers have been arranged into categories. From these categories, average use per student can be calculated for each group of students. These averages were analyzed using dependent t -tests.
81 Research Question 4 This question examines if differences exist between the selected urban schools and suburban/rural schools in their use of various school reform mechanisms. Using scores from the School Reform and Restructuring Index (SRRI), schools from the Urban School and Community Study can be compared to the schools from the School and Community Study in their use of reform mechanisms in the six areas of (1) governance, (2) accountability, (3) curriculum and instruction, (4) parent involvement, (5) special education climate/practices, and (6) pro-social discipline methods. These 6 scores for each school were averaged for all the urban schools and likewise for the suburban/rural schools. These means were analyzed using a mixed model Analysis of Variance (ANOVA) where there is within participant analysis using the 6 average scores of reform and between participant analysis by the two types of schools (urban versus suburban/rural) (see Table 12). Due to the increased probability of Type I errors occurring when numerous comparative analyses are conducted, a modified Bonferroni procedure will be used (Williams, Jones, & Tukey, 1999). Any differences found in the analysis were followed by an examination of the transcribed narrative interviews conducted with school staff at each school. This was conducted so the specific mechanism used at each school can be isolated and discussed.
82 Table 12 An Example Layout of the Analysis for Research Question 4. Urban Schools Urban Mean Scores Rural/Suburban Schools Rural/Suburban Mean Scores Difference in Means between school types? Areas of Reform School A School B School F School G Governance Score Score Mean Score Score Mean Yes / No Accountability Score Score Mean Score Score Mean Yes /No Curriculum and Instruction Score Score Mean Score Score Mean Yes /No Parent Involvement Score Score Mean Score Score Mean Yes /No Includedness Score Score Mean Score Score Mean Yes /No Pro-Social Score Score Mean Score Score Mean Yes /No Research Question 5 This research question addresses if differences in reform mechanisms used in the urban schools compared to the suburban/rural schools account for any differences in academic achievement, psychological functioning or mental health service utilization of the youth served in these schools. This area was explored because of differences found in reform activities between urban schools and suburban/rural schools found in Research
83 Question 4 and differences between schools types documented in academic achievement, psychological functioning, or mental health service use. To answer this question, a series of multiple regression equations were employed. The criteria variables included: reading achievement (WRAT Reading), math achievement (WRAT Math), psychopathology (CBCL Total), and service utilization (a composite variable). The predictor variables included both status variables (gender, age, annual family income level) and the SRRI total score. Therefore, there were four regression equations calculated as illustrated in Table 13. Table 13 An Example of the Four Regression Equations to Answer Research Question 5 Equation Number: Criterion Variable Predictor Variable 1. Reading achievement IQ, Gender, age, income, and SRRI score 2. Math achievement IQ, Gender, age, income, and SRRI score 3. Psychopathology IQ, Gender, age, income, and SRRI score 4. Mental health service use IQ, Gender, age, income, and SRRI score These regression equations will be used to identify if the variation in reading achievement, for example, can be accounted for by the youths gender, age, family income level, and the amount of reform and restructuring used with in the school the youth is attending. The status variables will be entered first into the regression equation
84 and the SRRI entered last. Entering the SRRI score last will allow the unique contribution this variable has on reading achievement to be examined. Within Question 5 is the situation in which the school from which the participants were recruited is not a variable or factor in the four multiple regression equations resulting in school being a nested factor (Maxwell & Delaney, 2000). As previously discussed, students were recruited from a defined set of schools and thus, some students share the same school. This is illustrated in Table 15. As can be see in Table 15, Participants 1, 2, and 3 all share being in urban environments, having CBCL scores, reading scores, and attend School A. Participants 6 and 7 also are in an urban environment, have CBCL scores, and reading scores. However, Participants 6 and 7 attend School B. Should not the factor of School be included in the analysis when examining the differences between students in urban and suburban/rural schools? When this factor of school is not included in the analyses, two problems arise (1) the independence of the participant may be compromised and (2) the variance accounted for by this factor of school will not be accounted for in the model (See Table 14). Table 14. An example of school nested within the factor of Urban/Suburban/Rural. Participant # 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 School A A A B B C C C CBCL Urban Reading (Continued on next page)
85 Table 14 (continued) Participant # 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 School G G H H J K CBCL Rural/ Suburban Reading The nested factor of school was acknowledged in the current design. However, the number of participants from each school was limited and there was insufficient N for each school to include school as a factor in the current analyses. Summary Furthering the knowledge base on school reform and restructuring activities is the overall goal of this study. More specifically, this study will examine the differences between urban schools and suburban/rural schools reform strategies and explore the differential effects on students with emotional disturbances. Following the described statistical analyses, empirical information will add to the knowledge base.
86 CHAPTER FOUR RESULTS The present study is a secondary analysis of data collected as part of two larger studies, the School and Community Study (suburban/rural) and the Urban School and Community Study (urban). The purpose of this study is to investigate the academic, social, and behavioral functioning of children with emotional disturbances and the relationship between this functioning and the level of school reform operating in suburban/rural versus urban communities. The results from this study answered five research questions and are reported in four major sections. The first section provides a description of how the sample for the current study was determined and the demographic characteristics of this sample of students. The following sections address the study research questions. Descriptive statistics are presented for the academic achievement outcomes (Research Question1); psychological functioning levels, including total problem behaviors, internalizing problem behaviors, and externalizing problem behaviors (Research Question 2); and mental health service utilization (Research Question 3) of the students attending both suburban/rural and urban schools. Inferential statistics are also presented for the school reform and restructuring index (SRRI) for schools located in suburban/rural areas and those located in urban areas (Research Question 4). Results of the multiple regression analyses examining the relationship between the predictor variables, IQ, gender, age,
87 income, and SRRI, and the four outcome variables, reading achievement, math achievement, psychological functioning, and mental health service use are presented (Research Question 5). The final section consists of a summary of the results. Sample The first four research questions required a comparative analysis of the characteristics of students in the suburban/rural study versus the students in the urban study. Students from the two studies were matched or paired on three demographic variables: gender, income, and age to control for extraneous factors. The matching process allowed an investigation of the characteristics of the students with emotional disturbances who were similar in ages and on similar socioeconomic levels from two different studies to occur. Only those students who had both student data and parent data files were included in the matching process (N = 213). The first step in generating a participant pool for this research study was to divide the studies into groups based on the gender of each participant. As a categorical variable, it was a straightforward process to separate the students into a data set that contained all female participants and another data set that contained all male participants. The isolation of this variable (gender) allowed the students to be matched on the two remaining variables: income and age. Income data were collected differently in the two studies. In the suburban/rural study, income data were collected as a categorical variable (e.g., $2,000 $2,999) where parents were asked to select the category that best described their monthly household income. In the urban study, income data were collected as a continuous variable where
88 parents were asked to report the income for the entire family. Parents reported their income in varying forms (e.g. hourly, weekly, monthly) and these amounts were converted to annual figures. The income categories from the suburban/rural study were converted to midpoints and these midpoint figures were used to calculate a new annual income figure for each participant (see Table 15). With this calculation, both studies had a continuous variable for income and this aided in the matching process. Table 15 Income Conversion from Categorical to Continuous Variables for the Suburban/Rural Study. Monthly Income Categories Monthly Income Midpoints Annual Income < $100 $50 $600 $100 499 $250 $3,000 $500 999 $750 $9,000 $1,000 1,999 $1,500 $18,000 $2,000 2,999 $2,500 $30,000 $3,000 3,999 $3,500 $42,000 $4,000 4,999 $4,500 $54,000 $5,000 5,999 $5,500 $66,000 Graphs were created in order to visually match participants from the two studies. Income was plotted on the X axis and age plotted on the Y axis for the male participants and the female participants. Markers (e.g. triangles, squares) representing each participant were placed on the graphs and the distance between the participant-markers was examined. Participant-markers that were closest to each other on both income and
89 age were considered a match. A match was a close similarity in income and age between participants in the two studies. Matches were deemed valid if the participants incomes were within $10,000 of each other and their ages were within a 1 year range. Both income and age had to be within these parameters to be considered a successful match (see Figures 1, 2, and 3 for the male participants and Figures 4 and 5 for the female participants). Income90000 80000 70000 60000 50000 40000 30000 20000 10000 0Age19 18 17 16 15 14 Urbanicity of schoolUrban Suburban/rural Figure 3 A scatterplot of income and age for males over the age of 14 by income and age.
90 Income90000 80000 70000 60000 50000 40000 30000 20000 10000 0Age14 13 12 11 10 Urbanicity of schoolUrban Suburban/rural Figure 4 A scatterplot of income and age for males between the ages of 10 and fourteen by age and income. Income90000 80000 70000 60000 50000 40000 30000 20000 10000 0Age10 9 8 7 6 5 4 Urbanicity of schoolUrban Suburban/rural Figure 5 A scatterplot of income and age for male participants under the age of 10 by income and age.
91 Income70000 60000 50000 40000 30000 20000 10000 0Age20 19 18 17 16 15 14 13 12 11 10 Urbanicity of schoolUrban Suburban/rural Figure 6 A scatterplot of income and age for female participants over the age of 10 by income and age. Income70000 60000 50000 40000 30000 20000 10000 0Age10 9 8 7 6 5 Urbanicity of schoolUrban Suburban/rural Figure 7 A scatterplot of income and age for female participants under the age of 10 by income and age.
92 For the total participants in the study, N = 114 in the suburban/rural study and N = 99 in the urban study, 66 participants in both studies were successfully matched for a total of 132 participants for the current sample. In the male participant group, 60 pairs were made (N = 120). The successful matches are displayed in Figure 6. Additionally, successful matches were made for six pairs (N = 12) of female participants. Figure 7 represents the six successful female matches (N = 12) that were made. Income 70000 60000 50000 40000 30000 20000 10000 0Age20 18 16 14 12 10 8 6 4 Urbanity of school Urban Suburban/rural Figure 8. A scatterplot of income and age for matched male participants by income and age.
93 Income70000 60000 50000 40000 30000 20000 10000Age18 16 14 12 10 8 6 Urbanicity of schoolUrban Suburban/rural Figure 9. A scatterplot of age and income for matched female participants by income and age. Table 16 displays the 66 participant pairs (120 male participants and 12 female participants) that were matched on age and income with the demographic data on the matched pairs reported in Table 17. In summary, using the variables of age and income, successful matches were made for 60 pairs (N = 120) of male participants and 6 pairs (N = 12) of female participants. This resulted in a total sample size of 66 pairs or 132 students. As a group, there were no differences found between the suburban/rural students (N = 66) and the urban students (N = 66) on age (t (65) = -1.18, p >.05), income (t (65) = 1.80, p > 0.5),
94 Table 16 Age, Income, and IQ Data for the Matched Sample Student ID Age Income IQ Suburban/ Rural Urban Suburban/ Rural Urban Suburban/ Rural Urban Suburban/ Rural Urban GAM01 MDM11 8 9 $30,0000 $30,000 79 79 GAM02 MDM02 7 7 $30,000 $32,700 96 53 GAM05 OHC04 10 10 $30,000 $26,087 109 69 GAM07 MDT09 11 11 $30,000 $24,000 86 82 GAM09 MDM01 8 8 $60,000 $63,000 95 97 GAM17 MDT05 10 10 $9,000 $9,244 53 80 GAM20 MDT06 11 11 $9,000 $10,260 91 57 GAM21 OHC08 10 11 $18,000 $18,000 89 73 GAR01 OHA05 16 16 $18,000 $15,576 76 62 GAR02 OHA26 16 16 $9,000 $9,600 73 74 GAR03 OHW01 17 17 $42,000 $45,000 84 99 GAR05 OHW09 15 16 $18,000 $15,000 102 56 GAR06 OHA20 15 16 $18,000 $19,048 82 68 GAR07 MDK02 17 16 $30,000 $36,400 110 83 GAR08 MDK11 15 14 $30,000 $24,000 93 67 IAA01 MDF07 11 11 $18,000 $21,000 70 --M IAA03 MDK12 13 13 $42,000 $36,400 95 93 IAA05 MDM07 10 9 $30,000 $26,400 114 75 IAA06 MDK10 12 13 $18,000 $12,288 81 61 IAA09 MDT10 9 10 $42,000 $42,000 119 98 IAA10 MDK26 13 13 $18,000 $15,420 76 75 IAA11 MDM04 9 9 $30,000 $30,000 105 81 IAA12 OHG04 10 11 $60,000 $66,000 98 --M IAA13 MDT02 10 9 $18,000 $19,884 117 88 IAA15 MDM03 7 6 $18,000 $24,000 50 71 IAA19 MDT03 8 8 $18,000 $22,810 91 85 IAA20 MDF08 12 12 $30,000 $33,800 102 83 IAA21 MDT01 9 9 $18,000 $18,000 133 78 IAA22 MDT08 10 10 $18,000 $15,100 99 63 IAA23 MDK25 12 12 $18,000 $14,160 98 83 IAA25 MDF05 11 11 $30,000 $22,200 83 76 IAA27 MDK18 12 12 $30,000 $28,000 97 68 IAA28 MDK17 11 11 $30,000 $24,024 115 81 KYE01 MDK22 12 12 $18,000 $14,264 107 83 KYE02 OHG10 10 9 $9,000 $14,400 84 86 KYE05 OHG02 8 8 $54,000 $50,000 93 95 KYV02 MDK03 13 13 $30,000 $30,000 83 62 (Continued on next page)
95 Table 16 (continued) Student ID Age Income IQ Suburban/ Rural Urban Suburban/ Rural Urban Suburban/ Rural Urban Suburban/ Rural Urban KYV04 MDF02 13 13 $9,000 $9,360 57 72 KYV06 MDK21 12 13 $30,000 $29,340 71 94 MDE03 MDF04 13 13 $30,000 $28,000 68 81 MDE04 MDK20 14 14 $18,000 $12,288 83 67 MDE05 OHW08 15 15 $18,000 $18,000 87 78 MDE07 MDK01 13 14 $9,000 $12,288 83 92 MDE12 MDK15 13 13 $18,000 $20,000 106 69 MDE13 MDT07 12 12 $9,000 $14,400 90 83 MDW01 OHW14 17 17 $54,000 $45,838 79 83 MDW03 OHA10 16 16 $18,000 $14,400 112 89 MDW05 OHA18 18 18 $54,000 $43,992 90 82 MDW06 OHA32 15 16 $30,000 $32,360 109 82 MDW08 OHA13 16 15 $42,000 $50,400 112 97 MDW09 OHW02 15 16 $9,000 $5,400 105 --M MDW10 OHW13 16 16 $30,000 $24,000 109 70 MDW11 OHA03 17 17 $30,000 $24,000 98 70 MDW12 OHA24 18 17 $18,000 $12,672 87 87 MDW13 OHA04 18 17 $30,000 $20,464 99 69 VTE03 MDF09 11 12 $18,000 $20,400 82 ---M VTE04 MDK04 12 13 $18,000 $12,000 82 104 VTE05 MDK09 13 13 $30,000 $20,800 97 70 WIL01 MDK06 14 14 $30,000 $27,600 97 100 WIL02 MDK16 14 14 $3,000 $6,000 102 85 WIL03 OHC07 12 12 $18,000 $21,600 90 74 WIL04 MDK07 10 11 $9,000 $9,480 96 75 WIL05 OHC13 14 14 $9,000 $12,000 100 71 WIL10 OHA15 14 15 $42,000 $50,000 85 74 WIL11 MDF03 12 13 $18,000 $15,060 76 97 WIL13 MDM05 6 5 $18,000 $24,000 86 70 Means (SD) 12.45 (2.94) 12.53 (3.00) $24,700 ($12,800) $24,155 ($13,200) 91.91 (15.71) 78.53 (11.79) M=missing IQ data and gender (Z = .0, N = 66, p>.05). There were also no significant differences found on cost of school lunch (Z = -1.80, N = 36, p > .05), race (Z = -2.74, N = 11, p < .05), or
96 household size (t (65) = -1.88, p>.05). Differences were found on the variables IQ (t (55) = 6.86, p < .05) and grade level (t (65) = -3.25, p < .05). An analysis of the IQ levels for the two groups was also performed. The IQ scores for the suburban/rural participants ranged from 50 to 133 while the urban participants scores ranged from 53 to 104. One participant in the suburban/rural group had an IQ score of 133. This score was more than two standard deviations above the group average (M = 91.91, SD = 15.71). Therefore, this participants IQ score was recoded as missing and not included in the group mean. Likewise, three participants had IQ scores two standard deviations below the group average (50, 53, and 57). These IQ scores were also recoded as missing and not included in the group mean. An examination of the urban group revealed that one participant had an IQ score of 104, which was two standard deviations above the group average (M = 78.53, SD = 11.79) and another urban participant had an IQ score of 53, which was two standard deviations below the group average. As with the suburban/rural group, these IQ scores were recoded as missing and not included in the group mean. Group means were recalculated for both the suburban/rural group (M = 93.11, SD = 12.66) and the urban group (M = 78.53, SD = 11.03). Transforming the outlying scores to missing variables did not greatly influence the group means but did reshape the distribution closer to normality. Statistical analysis, however, revealed statistically significant differences in IQ levels with the suburban/rural participants scoring higher on the intelligence tests than the urban participants (t (55) = 6.85, p < .05).
97 While the suburban/rural participants were matched on three key variables (gender, income, and age) with the urban participants, the intelligence and grade levels of the two groups varied. However, the two groups were found to be similar on several other demographic variables (meal plan and household size). Therefore, the group was deemed similarly matched. Table 17 provides the descriptive statistics of the resulting sample of matched students. In summary, the resulting sample of 132 students (66 participants from each study) was predominately male, around 12 years of age, and in the sixth grade. There were on average four people living in a household with an annual income of $25,000. Table 17 Summary Statistics of Matched Sample Suburban/Rural (N = 66) Urban (N = 66) Total Sample (N = 132) n % M (SD) n % M (SD) n % M (SD) Gender 66 66132 Male 60 90.96090.9120 90.9 Female 6 9.169.112 9.1 Income (SD) 66 $25,137 ($13,629) 66$24,155 ($13,200) 132 $24,672 ($13,897) <$10,000 11 16.769.117 12.9 $10,000-$19,999 24 36.32233.346 34.8 $20,000-$29,999 0 ---2131.821 15.9 $30,000-$39,999 21 31.8812.229 22.0 $40,000-$49,999 5 7.646.09 6.8 $50,000-$59,999 3 4.634.66 4.5 >$60,000 2 3.023.0 4 3.0 Age (SD) 66 12.45 (2.94) 6612.53 (3.00) 132 12.49 (2.96) 5-7 yrs old 3 4.534.56 4.5 8-10 yrs old 16 24.31319.729 22.0 Continued on next page
98 Table 17 (continued) Suburban/Rural (N = 66) Urban (N = 66) Total Sample (N = 132) n % M (SD) n % M (SD) N 11-13 yrs old 24 36.42639.450 37.9 14-16 yrs old 16 24.21827.334 25.8 17-19 yrs old 7 10.669.113 9.8 Race 66 66132 White 54 81.81116.765 49.2 Black 2 3.05380.355 41.7 Hispanic 1 1.50---1 .8 Native American 8 12.10---8 6.1 Other 1 1.523.03 2.3 Meal Plan 66 66132 Free/Reduced 44 66.75481.898 74.2 Full 22 33.31218.234 25.8 Grade level (SD) 66 6.12 (2.97) 666.53 (2.86) 132 6.33 (2.91) Kindergarten 2 3.023.04 3.0 1-3 12 18.21015.222 16.7 4-6 26 39.41827.344 33.3 7-9 16 24.22740.943 32.6 10-12 10 15.2913.619 14.4 IQ Score 66 91.91 (15.71) 62 78.53 (11.79) 128 85.43 (15.43) 50-69 4 6.1 1421.2 18 13.4 70-89 24 36.3 3756.1 66 49.3 90-109 30 45.5 1116.7 42 31.3 110-129 7 10.6 0--7 5.3 >=130 1 1.5 0--1 .7 Missing 0 --4--4 --Household Size (SD) 66 4.08 (1.47) 664.62 (2.03) 132 4.35 (1.79) 2-3 30 45.5 2334.8 53 40.2 4-5 25 37.8 2334.8 48 36.4 6-7 10 15.2 1218.2 22 16.7 8-9 1 1.5 710.7 8 6.1 >10 0 --1 1.5 1 0.8
99 To address the representativeness of the current sample to the original study samples, the matched study participants were compared with the non-matched participants from the two core studies. Those students matched (N = 132) did not differ significantly from those students that were not matched (N = 81) on age, t (211) = .966, p > .05), grade level, t (211) = 1.05, p > .05), and cost of school lunch (Z = -.807, N = 213, p > .05). There were also no significant differences found between the two groups on income, t (105) = .019, p > .05), IQ (suburban/rural: t (113) = .488, p >.05 and urban: t (91) = .142, p > .05), or the number of people living in the household, t (204) = .037, p > .05). Differences were found on the variable: gender (Z = -4.25, N = 213, p < .05). There were more participants in the suburban/rural study (N = 114, n = 93 males, n = 21 females) than in the urban study (N = 99, n = 82 males, n = 17 females) resulting in a significant difference. Unmatched Participants The matching process described previously resulted in 66 pairs or 132 participants, leaving 49 participants in the suburban/rural study unmatched and 33 participants in the urban study unmatched. Of these remaining participants (N = 82), nine pairs or 18 participants were matched using new parameters for age and income. The new parameters were increased to ages within a two year range and incomes within a $20,000 range. The nine pairs are shown in Table 18.
100 Table 18 Age and Income Data for Matched Participants using New Parameters. Suburban/Rural Participants Urban Participants Differences Pair Student ID Age Income Student ID Age Income Age Income 1. GAM04 12 $30,000OHA2214 $16,800 -2 $13,200 2. IAA02 7 $9,000MDM065 $10,000 2 $-1,000 3. IAA04 9 $66,000OHG019 $50,000 0 $16,000 4. KYE04 11 $18,000MDF1012 $6,336 -1 $11,664 5. KYV01 8 $18,000MDT0410 $8,568 -2 $9,432 6. KYV03 14 $9,000OHW1116 $14,400 -2 $-5,400 7. MDE11 13 $42,000MDK1914 $60,000 -5 $-18,000 8. MDW02 17 $60,000OHA0115 $67,600 2 $-7,600 9. VTE02 6 $18,000MDM086 $5,388 0 $12,612 Mean differences -0.889$3,434 After the second match procedure, 63 participants remained in the two studies. Of this group, 39 participants were from the suburban/rural study and 24 participants were from the urban study. The age and income data on these participants are listed in Table 19. Table 19 Age and Income Data on Remaining Participants Suburban/Rural Urban Differences Student ID Age Income Student ID Age Income Age Income GAM08 7 $18,000 OHC01 10 $8,400 -3 $9,600 GAM18 10 $30,000 OHW06 15 $29,000 -5 $1,000 GAM24 11 $42,000 OHW17 14 $60,000 -3 -$18,000 GAM27 7 $18,000 OHA17 15 $18,000 -8 $0 (Continued on next page)
101 Table 19 (continued) Suburban/Rural Urban Differences Student ID Age Income Student ID Age Income Age Income GAR04 18 $54,000 OHW15 19 $80,568 -1 -$26,568 IAA07 10 $9,000 OHA02 18 $13,000 -8 -$4,000 IAA16 10 $18,000 OHA23 16 $14,460 -6 $3,540 MDE02 13 $18,000 MDF01 13 $30,000 0 -$12,000 MDE06 12 $60,000 OHA11 15 $79,000 -3 -$19,000 MDE08 12 $18,000 OHW16 19 $17,400 -7 $600 MDE10 14 $600 OHA19 15 $40,000 -1 -$39,400 MDE14 12 $30,000 OHA07 17 $21,000 -5 $9,000 MDW07 17 $60,000 OHA29 17 $32,000 0 $28,000 MDW14 18 $60,000 OHW07 18 $37,000 0 $23,000 MDW15 16 $60,000 OHA09 16 $28,000 0 $32,000 WIL06 8 $30,000 OHA16 17 $20,760 -9 $9,240 WIL12 10 $18,000 OHA14 15 $16,760 -5 $1,240 GAM10 9 $9,000 MDF06 13 ---M GAM25 10 $9,000 MDF11 13 --M IAA14 8 $18,000 MDM09 7 --M IAA17 7 $18,000 OHA21 15 --M KYE03 7 $18,000 OHA25 17 --M KYV05 12 $42,000 OHG07 13 --M KYV07 7 $9,000 OHW03 17 --M GAM06 7 $30,000 GAM11 7 $18,000 GAM13 10 $60,000 GAM14 7 $30,000 GAM15 10 $60,000 GAM16 8 $42,000 GAM19 12 $54,000 GAM23 8 $42,000 MDE09 13 $3,000 VTE01 7 $60,000 VTM01 12 $30,000 VTM03 8 $66,000 VTM04 7 $42,000 IAA26 13 --M MDE01 14 ---M MDW04 17 ---M M = missing income data.
102 Study Sample Size Summary As a precursor to the investigation of the relationship between school location and academic achievement, psychological functioning, and mental health service utilization, a matching procedure was conducted. Participants from the suburban/rural study were paired with participants from the urban study on three variables: gender, income, and age. The first sequence produced 66 pairs or 132 participants that were within one year of age and had an annual family income within $10,000. A second sequence was conducted with larger parameters that yielded nine additional pairs of participants. It was concluded that expanding the matching criteria to within two years and $20,000 to include the additional nine pairs of participants would result in the two samples being very different. It should be noted that no urban participants were matched with the suburban/rural participants attending VTM, due to the low number of participants (N = 3) from that school. The exclusion of these participants did not effect the study sample or student outcomes. Therefore, the sample for the current investigation will remain at 66 pairs or 132 participants. Research Question 1 The following section examines the research question: Are there differences in academic functioning between suburban/rural participants and a matched sample of urban participants? Reading achievement scores (standardized score with 100 being average) for suburban/rural participants ranged from 45 to 123 and their math achievement scores ranged from 45 to 114. The matched sample of urban participants reading achievement scores ranged from 47 to 112 and their math achievement scores ranged from 45 to 99.
103 The frequencies and percentages are reported in Table 20. Dependent t-tests revealed that the suburban/rural participants achievement scores were consistently higher than the achievement scores of the urban participants in both reading, t (65) = 2.57, p < .05) and math, t (65) = 4.27, p < .05). Table 20 Descriptive Statistics for WRAT3 Reading and Math Achievement Standard Scores Suburban/Rural (N = 65) Urban (N = 66) T or Z Values Reading Standard Scores 85.43 (18.46) 77.88 (16.22) 2.57* <50 1 1.5 1 1.5 0 59 3 4.6 10 15.2 60 69 9 13.8 9 13.6 70 79 12 18.5 15 22.7 80 89 13 20.0 17 25.8 90 99 10 15.4 7 10.6 >100 17 26.2 7 10.6 Grade level Above grade 10 15.446.1 At grade 8 12.3710.6 Below grade 47 72.35583.3 Math Standard Scores 84.69 (16.19) 75.15 (11.95) 4.27* <50 3 4.6 1 1.5 50 59 3 4.6 5 7.6 60 69 5 7.7 18 27.3 70 79 9 13.8 17 25.8 80 89 17 26.2 16 24.2 90 99 18 27.7 9 13.6 >100 10 15.4 0 0.0 Grade level Above grade 34.600 At grade 913.857.6 Below grade 5381.56192.4 *p < .05
104 Time Spent in School Setting The differences found in academic achievement prompted an investigation of how the participants spent their academic day. Reflected in the participants school schedules, an examination was conducted of the type of classroom setting (academic or nonacademic and either for special education students only or for all students) and the amount of time the participants spent in the settings each school day (see Table 21). For the suburban/rural participants, 57% (SD = 35) of their day was spent in special education settings and 43% (SD = 35) of their day was spent in regular education settings. The urban participants spent on average 73% (SD = 25) of their day in special education settings and 27% (SD = 25) of their day in regular education settings. Dependent t-test results for special education settings were significant, t (65) = -3.05, p < .05), indicating that urban participants spend more time in special education settings than their suburban/rural counterparts. Table 21 Percent of Day Spent in Educational Settings Suburban/Rural (N = 66) Urban (N = 66) t Values n % M (SD) n % M (SD) Special Education 57.50 (35.12) 72.19 (24.99) -3.05* Academic 39.11 (28.01) 51.70 (21.21) 0.0% 16 24.2 4 6.1 0.1 19.9 1 1.5 1 1.5 20.0 39.9 12 18.2 5 7.6 40.0 59.9 23 34.8 3451.5 (Continued on next page)
105 Table 21 (continued) Suburban/Rural (N = 66) Urban (N = 69) t Values n % M(SD) n % M (SD) 60.0 79.9 7 10.6 1725.8 80.0 100 7 10.6 5 7.6 Non-Academic 18.39 (20.19) 20.24 (19.67) 0.0% 21 31.8 1015.2 0.1 19.9 18 27.3 3248.5 20.0 39.9 17 25.8 4 6.1 40.0 59.9 8 12.1 2030.3 60.0 79.9 1 1.5 0 0.0 80.0 100 1 1.5 0 0.0 Regular Education 43.07 (34.77) 27.68 (24.99) 3.23* Academic 18.27 (23.66) 5.25 (16.03) 0.0% 31 47.0 5380.3 0.1 19.9 12 18.2 8 12.1 20.0 39.9 5 7.6 1 1.5 40.0 59.9 13 19.7 2 3.0 60.0 79.9 4 6.1 2 3.0 80.0 100 1 1.5 0 0.0 Non-Academic 24.80 (19.06) 22.68 (16.75) 0.0% 8 12.1 1 1.5 0.1 19.9 27 40.9 3654.5 20.0 39.9 16 24.2 2030.3 40.0 59.9 14 21.2 8 12.1 60.0 79.9 1 1.5 0 0.0 80.0 100 0 0.0 1 1.5 p < .05 Research Question 2 The following section addresses the research question: Are there differences in the psychological functioning (i.e., symptomatology and functional impairment) between
106 suburban/rural participants and a matched sample of urban participants? First, the descriptive and inferential statistics are presented for psychological functioning, followed by the statistical data for functional impairment. Psychological Functioning The behavioral problems of participants attending suburban/rural schools and participants attending urban schools were measured by the Child Behavior Checklist (CBCL). Scores for the suburban/rural participants were lower than the scores for the urban participants in the areas of CBCL Internalizing and CBCL Externalizing. Lower scores on this assessment scale mean that the participants exhibited fewer behavioral problems. Paired samples t-tests revealed no significant differences in any of the three CBCL areas: Total CBCL T-scores, t (65) = -.393, p>.05; CBCL Internalizing T-scores, t (65) = .754, p > .05; CBCL Externalizing T-scores, t (65) = -1.48, p > .05). Skewness and kurtosis values did not indicate substantial departure from normality for these three measures. Table 22 Child Behavior Checklist (CBCL) T-scores Suburban/Rural (N = 66) Urban (N = 66) n % M (SD) n % M (SD) t Value Total Problem Score 67.20 (10.00) 67.86 (8.77) -.393 Clinical (> 63) 49 74.2 41 62.1 Borderline (60 63) 7 10.6 12 18.2 Normal (< 60) 14 21.2 13 19.7 (Continued on next page)
107 Table 22 (continued) Internalizing Score 63.68 (10.79) 62.09 (12.77) .754 Clinical (> 63) 35 53.0 30 45.5 Borderline (60 63) 8 12.1 12 18.2 Normal (< 60) 23 34.8 24 36.4 Externalizing Score 65.95 (11.17) 68.53 (8.76) -1.48 Clinical (> 63) 40 60.6 44 66.7 Borderline (60 63) 9 13.6 11 16.7 Normal (< 60) 17 25.8 11 16.7 Functional Impairment Two different assessment tools were used to determine the participants functional impairment: Child and Adolescent Functional Assessment Scale (CAFAS) and Columbia Impairment Scale (CIS). For the suburban/rural participants, average CAFAS scores across the five subscales of schoolwork, home, behavior towards others, and moods and emotions placed 96% in the clinical range of impairment (see Table 23). Table 23 Child and Adolescent Functional Assessment Scale (CAFAS) Suburban/Rural (N = 66) Domains n % M SD Schoolwork1 27.58 6.58 Home1 21.06 10.54 Community1 8.33 11.17 (Continued on next page)
108 Table 23 (continued) Suburban/Rural (N = 66) Domains n % M SD Behavior toward Others1 20.15 9.36 Moods and Emotions1 19.39 10.06 5 Domain Total Score2 79.09 29.07 0-10 (None or minimal dysfunction) 1 1.5 20 30 (Mild impairment) 2 3.0 40 60 (Moderate impairment) 19 28.8 70 80 (Marked impairment) 14 21.2 90 or higher (Severe impairment) 30 45.5 Clinical Range3 (> 40) Non-Clinical Range3 (< 40) 63 3 95.5 4.5 1Item response options: 0=Minimal/no impairment; 10=Mild impairment; 20=Moderate impairment; 30=Severe impairment. 2The 5 Domain Total Score has a range of 0 to 150. 3Clinical Range: 5 Domain Total Scores > 40; Non-Clinical Range: 5 Domain Total Scores< 40. For the urban participants, average CIS scores placed 41 participants or 62% in the clinical range of impairment, having the most problems with behavior at schools, problems with schoolwork, getting into trouble, and problems at home. Based on a total item score of 16, 60% or 41 participants were determined to be clinically impaired (see Table 24).
109 Table 24 Columbia Impairment Scale (CIS) Suburban/Rural (N = 66) n % M SD CIS Item Scores1 Getting into trouble 1.88 1.35 Gets along female 1.26 1.04 Gets along male 2.63 2.78 Feeling unhappy/sad 1.36 1.34 Behavior at school 2.41 1.53 Having fun .85 1.60 Gets along w/adults 1.27 1.59 Feeling nervous/afraid .97 1.16 Getting along w/siblings 1.93 1.78 Gets along w/other kids 1.52 1.38 Involved sport/hobby 1.27 1.59 Problems w/schoolwork 1.98 1.51 Problems at home 1.88 1.23 CIS Total score 19.73 (9.60) Clinical range (> 16) 41 62.1 Non-Clinical range (< 16) 25 37.9 1Item response options: 0=No problem; 1=Very small problem; 2=Some problem; 3=Moderate problem; 4=Very bad problem. A Wilcoxon signed-rank test was conducted between the number of participants who scored in the clinical range and the non-clinical range of the CAFAS and CIS for the suburban/rural participants and urban participants. Differences were found: Z = 26.0, p < .05. As the data show in Table 25, more suburban/rural participants were functionally impaired than urban participants.
110 Table 25 Investigation of the relationship between the impairment of suburban participants and urban participants Clinical Range Non-Clinical Range Z Value CAFAS Suburban/Rural Participants 63 (95.7%) 3 (4.3%) CIS Urban Participants 43 (65.2%) 23 (34.8%) 26.0* p < .05 Research Question 3 The following section addresses the research question: Are there differences in the mental health service utilization between suburban/rural participants and a matched sample of urban participants? Mental health service data were collected in two categories: those services received at school, during the school day as reported by the school staff and those services received as reported by the parent/guardian. School Staff Report School personnel reported the number and type of therapeutic services received by students in school, delivered by school personnel during the school day and those services received by students in school delivered by personnel from a community agency during the school day. Forty-three or 65% of the suburban/rural participants received mental health services from school personnel while 73% or 48 of the urban participants
111 received this type of service. A Wilcoxon sign-rank test found no significant differences in the number of participants who received mental health services from school personnel from those who did not by suburban/rural or urban setting. However, differences were found in the number of participants who received mental health services from agency personnel (Z = -3.40, N = 3, p < .05) with more suburban/rural youth (38 or 58%) receiving services from mental health agency personnel than urban youth (20 or 30%). Further analysis revealed that the number of participants who received mental health services from either school personnel or agency personnel varied with 80% of the suburban/rural matched participants receiving services and 78% of the urban matched participants. Suburban/rural educational settings provided a wider array of mental health services and used more agency personnel to meet the needs of their students (see Table 26). Table 26 Therapeutic Services Received in the Sc hool, during the School Day, from School Personnel or from Agency Professionals Suburban/Rural Urban n % n % Z Value Services from School Personnel 43 65.2 48 72.7 -.962 Individual Counseling 1015.2 4466.7 Group Counseling 2639.4 4466.7 Case Management 2842.4 1827.3 Medication Monitoring 57.6 46.1 Other Medical Services 46.1 0-Other1 1116.7 0-Services from Agency Professionals 5857.6 2030.3 -3.40* Individual Counseling 913.61319.7 Continued on next page
112 Table 26 (continued) Suburban/Rural Urban n % n % Z Value Group Counseling 1421.20--Case Management 0---69.1 Medication Monitoring 34.51116.7 Other Medical Services 0---0--Other2 2639.457.6 Received Services from Either School Personnel or Agency Professionals 53 80.3 51 77.3 1Other category service was health services. 2Other category services were: family counseling, in-home intervention, one-on-one behavioral aide, police liaison, probation officer, counseling with a licensed clinical social worker, and group activities. p < .05 Parent Report Mental health services utilization was reported by the suburban/rural parents in five broad categories: overnight/inpatient treatment, outpatient treatment, other professional help, other non-professional help, and other services used during the last six months as well as a lifetime use. Within each of these broad categories, several different services are listed. Table 27 presents information on service utilization and the number of different services used within each service category for the matched sample of suburban/rural participants. According to their suburban/rural parents, seven of the nine inpatient mental health services had been utilized by at least two participants. In the outpatient mental health treatment category, all of the services had been utilized by at least one participant with the highest number of participants using School Guidance Counselor, Psychologist, or Social Worker (54 participants or 82%). An examination of services used over the past six months revealed that most of the sample participants used
113 services under the category Other Professional Help through outpatient mental health treatment facilities. Table 27 Frequencies for Type of Services Ever Used and Used in the Past Six Months for Suburban/Rural Participants (N = 66) Ever Used Used in Past 6 Months Service n (%) n (%) Overnight/Inpatient Treatment Psychiatric hospital inpatient unit 1522.7 2 3.0 General hospital psychiatric unit 1015.2 3 4.5 Alcohol or drug treatment unit 23.0 1 1.5 Medical inpatient unit 0--0 --Residential treatment center 1725.8 4 6.1 Detention center/Training school/Jail 46.1 0 --Group home or emergency shelter 46.1 1 1.5 Therapeutic foster care 710.6 3 4.5 Boarding school 0--0 --Outpatient Mental Health Treatment In-home emergency services or in-home counseling 11 16.7 3 4.5 Outpatient drug or alcohol clinic 34.5 3 4.5 Mental health center 2436.4 10 15.2 Community health center 57.6 1 1.5 Crisis center 11.5 0 --Day treatment program or hospital 913.6 6 9.1 Private professional help 39 59.1 28 42.4 Other Professional Help School guidance counselor, psychologist, or social worker 54 81.8 43 65.2 Special help in classroom setting 5075.8 44 66.7 Special classes or special services 4060.6 31 47.0 Educational tutoring 1421.2 6 9.1 Social services 1725.8 9 13.6 Probation officer/Juvenile correction counselor 15 22.7 10 15.2 Family doctor 2639.4 15 22.7 Hospital emergency room 812.1 3 4.5 (Continued on next page)
114 Table 27 (continued) Service Ever Used Used in Past 6 Months n (%) n (%) Minister/Priest/Rabbi 34.5 2 3.0 Other healers 0--0 --Other Non-Professional Help Crisis hotline 46.1 3 4.5 Self-help groups 11.5 1 1.5 Adult relatives 1319.7 11 16.7 Other adults 1218.2 9 13.6 Friends 57.6 4 6.1 Other Services Case manager 812.1 4 6.1 Respite 710.6 4 6.1 Table 28 presents information on service utilization and the number of different services used within each service category for the matched sample of urban participants for lifetime use as well as use during the last 12 months. According to their urban parents, the highest percentage of urban participants used Community Mental Health Center services (67% or 44 participants) with none of the urban participants having used Drug/Alcohol Treatment or Self-help Group services. Of the urban participants in this matched sample who had used the community mental health center over half of them had used the service in the last year. The same trend can also be seen in the use of most of the other outpatient services, that is, of the participants who had ever used a service, over half of them had used the service in the past 12 months.
115 Table 28 Frequencies for Type of Service Ever Used and Used in Last Year for Urban Participants (N = 66) Ever Used Used in Last Year Service n % n % Overnight/Inpatient Mental Health Treatment Psychiatric hospital 1116.7 1 1.5 General hospital psychiatric hospital 1522.7 5 7.6 Drug or alcohol treatment unit 0--0 --Residential treatment center 913.6 4 6.1 Group home 69.1 2 3.0 Foster home 69.1 4 6.1 Detention center/Jail 34.5 1 1.5 Outpatient Treatment Mental Health Treatment Community mental health center 4466.7 28 42.4 Psychologist/Psychiatrist/Social worker/Family counselor 37 56.1 22 33.3 Partial hospitalization/Day treatment program 1319.7 9 13.6 Drug or alcohol treatment clinic 11.5 1 1.5 In-home therapist/Counselor/Family preservation worker 21 31.8 11 16.7 Emergency room 1218.2 6 9.1 Pediatrician/Family doctor 2131.8 11 16.7 Probation or juvenile corrections officer/Court counselor 9 13.6 6 9.1 Priest/Minister/Rabbi 69.1 4 6.1 Other healers 23.0 0 --Acupuncturist/Chiropractor 11.5 0 --Self-help group 0--0 --Respite 34.5 1 1.5 Descriptive statistics showed that a slightly higher percentage of urban participants (42%) utilized inpatient mental health services than suburban/rural participants (39%). The same trend was true for outpatient mental health service utilization with urban parents reporting slightly higher usage than suburban/rural parents.
116 A Chi square analysis showed no statistically significant differences between the two groups in their lifetime use of inpatient and outpatient service utilization. This statistical procedure was the most appropriate due to the nature of the data (i.e., the use of different assessment instruments and These results confirm the trend of mental health service utilization between the two study populations. (see Table 29) Table 29 Investigation of the Relationship between Mental Health Service Utilization (Ever Used) and Setting Suburban/Rural Urban "2 Values Inpatient MH Services Suburban/Rural Participants 27 (20%) 29 (21%) .120* Outpatient MH Services Suburban/Rural Participants 57 (41%) 63 (46%) 2.30* p < .05 Research Question 4 The following section addresses the question: Are there differences in the school reform activities between the suburban/rural schools attended by students with emotional disturbances and the urban schools attended by students with emotional disturbances? Scores from the School Reform and Restructuring Index (SRRI) were used to describe
117 the level of school reform activities operating in the suburban/rural schools and the urban schools. School Reform and Restructuring Measures Suburban/Rural Schools Schools in the suburban/rural study were purposely selected as actively engaged in reform and restructuring activities. For the ten schools in this study, 57% of the reform activities measured by the SRRI were implemented by these schools. The area with the highest level of implementation was Curriculum and Instruction (73%); second highest was Governance (67%). Parent Involvement was the area with the lowest level of implementation (37%). The overall implementation level for all six propositions was almost 60% Although the SRRI measured the propositions adequately in the first study, the scale was revised to enhance its sensitivity in the second study. The six propositions contain several parts or items that are discrete. The initial design had a total of 18 parts or items resulting in each proposition having 2, 3, or 4 parts. This resulted in propositions with more items having more weight or influence in the total score than those propositions with fewer items. In the revised SRRI, each proposition has four items for a total of 24 items.
118 Table 30 Results of Ratings of Propositions on School Reform and Restructuring Suburban/Rural Schools School Governance AccountAbility Curriculum and Instruction Includedness Parent Involvement ProSocial Discipline Total Restructuring Score Range -12 to +12 -9 to +9 -6 to +6 -9 to +9 -12 to +12 -6 to +6 -54 to +54 GAM 8.2 7.2 5.2 5.4 1.8 3.8 31.6 GAR 7.6 6.2 4.6 5.6 5.0 4.2 33.2 IAA 7.6 5.4 4.6 5.4 8.0 4.4 35.4 KYE 7.8 5.4 3.6 4.6 5.6 3.8 30.8 KYV 9.0 5.6 4.6 5.2 3.2 -0.8 26.8 MDE 6.4 4.0 3.0 2.8 5.8 4.4 26.4 MDW 10.6 3.0 4.0 4.0 6.8 3.8 32.2 VTE 7.4 5.4 5.4 6.4 5.6 3.0 33.2 VTM 6.2 4.0 4.8 6.2 4.2 4.6 30.0 WIL 8.8 7.2 4.4 6.4 -1.8 4.0 29.0 Mean (Percent of Total) 8.0 (67%) 5.4 (60%) 4.4 (73%) 5.2 (58%) 4.4 (37%) 3.5 (58%) 30.9 (57%) SD 1.3 1.4 0.7 1.1 2.8 1.6 2.9 Range 6.2-10.6 3.0-7.2 3.0-5.4 2.8-6.4 -1.8-8.0 -0.8-4.6 26.4-35.4 Note. For each part of the proposition for each school, the raters ratings were averaged and then summed to get a score for each proposition. The six scores for each proposition were summed to get the Total Restructuring Score. Urban Schools Schools in the urban study were selected as implementing both high and low levels of reform and restructuring. For schools in the urban study, these eight schools on average implemented 15 % of the reform activities measured by the SRRI. The area with the highest level of implementation was Pro-Social Discipline (32%). The next highest area of implementation was Governance (23%) and the areas with the lowest implementation rates were Includeness (1%) and Parent Involvement (1%).
119 Table 31 Scores for Urban Schools by School Reform and Restructuring Proposition School Governance Accountability Curriculum and Instruction Includedness Parent Involvement ProSocial Discipline Total Restructuring Score Range -12 to +12 -12 to +12 -12 to +12 -12 to +12 -12 to +12 -12 to +12 -72 to +72 Urban MDM1 3.6 1.6 6.8 3.4 -2.0 6.6 20.0 MDK1 4.2 6.6 1.2 -3.4 4.6 -0.4 12.8 MDT 1.6 1.0 -2.6 -2.2 0.8 0.4 -1.0 MDF -2.8 3.4 -2.8 -9.4 3.4 -4.0 -12.2 OHA1 1.4 1.2 8.8 9.0 -1.2 9.2 28.4 OHC 4.2 0.2 0.8 -0.8 -3.8 7.6 8.2 OHG1 5.0 2.4 8.0 8.0 1.8 7.6 32.8 OHW 4.2 -1.2 -0.4 -3.6 -2.6 3.2 -0.4 Mean (Percent of Total) 2.7 (23%) 1.9 (16%) 2.5 (21%) .13 (1%) .13 (1%) 3.8 (32%) 11.1 (15%) SD 2.6 2.4 4.7 6.3 3.0 4.7 15.5 Range -2.8-5.0 -1.2-6.6 -2.8-8.8 -9.4-9.0 -3.8-4.6 -4.0-9.2 -12.2-32.8 1Nominated as actively engaged in reform activities Note. For each part of the proposition for each school, the raters ratings were averaged and then summed to get a score for each proposition. The six scores for each proposition were summed to get the Total Restructuring Score. Conversion to Percentage Scores Although changes to the SRRI were minor, the result was a new scale. In order to interpret the SRRI scores from these two measures, scores for all the schools were recalculated based on the percentage of the SRRI model they were implementing (see Table 33). The percentage score represents how closely the schools reform and restructuring activities aligned with the SRRI model as measured by the SRRI and rated by the original study raters. In other words, how closely the schools were to meeting the goals of the six SRRI propositions. Average percentage scores for the suburban/rural
120 schools (" = 59%) were significantly higher than the mean for the urban schools (" = 18%). Those scores were expected due to the selection criteria of the schools for the original studies. Suburban/rural schools were nominated and selected for their high engagement in reform initiatives and their percentage scores reflected those criteria. In the urban areas, schools were nominated and selected based on a broader range of reform activity criteria. Table 32 Reform and Restructuring Data on Study Schools School Level of Reform and Restructuring Activities SRRI Total Score Percentage Score % Suburban/Rural GAM High 31.6 61 GAR High 33.2 59 IAA High 35.4 66 KYE High 30.8 57 KYV High 26.8 50 MDE High 26.4 49 MDW High 32.2 60 VTE High 33.2 61 VTM High 30.0 56 WIL High 29.0 54 Urban MDF Low -12.2 0 MDK High 20.0 18 MDM High 12.8 28 MDT Low -1.0 0 OHA High 28.4 39 OHC Low 8.2 11 OHG High 32.8 46 OHW Low -0.4 0
121 All of the eight urban schools were less actively engaged in reform and restructuring than the ten suburban/rural schools. While the urban schools were purposely selected for their varying degrees of engagement in reform and restructuring activities, those schools nominated for high levels of engagement were lower than any of the highly engaged suburban/rural schools. The highest level of alignment to the SRRI propositions for an urban school was 46% while the lowest level of alignment for a suburban/rural school was 49% (see Table 33). Table 33 Rank Order of School Reform and Restructuring Index (SRRI) Scores from Lowest to Highest Rank School Location SRRI Percentage Scores 1 (Lowest) MDF U 1 2 MDT U 1 3 OHW U 1 4 OHC U 11 5 MDK U 18 6 MDM U 28 7 OHA U 39 8 OHG U 46 9 MDE S/R 49 10 KYV S/R 50 11 WIL S/R 54 12 VTM S/R 56 13 GAR S/R 56 14 KYE S/R 57 15 MDW S/R 59 16 GAM S/R 61 17 VTE S/R 62 18 (Highest) IAA S/R 66
122 Schools and Study Sample Participants Of the 10 schools that participated in the School and Community study, only nine schools were represented in the matched participant sample. No similar urban participants were found for any of the suburban/rural participants from the VTM school on the matching criteria (e.g., gender, income, age). Removing VTM from the suburban/rural school sample did not effect the average percentage score (N = 10, M = 59; N = 9, M = 59). VTM remained in the study sample of suburban/rural schools. School Reform and Restructuring Interviews In addition to the numeric score indicating the level of engagement in reform and restructuring activities, the SRRI captured a description of the activities and the climate of the schools through semi-structured interviews using multiple informants. The interviews provided a more detailed picture of the schools performance across the six propositions. In this section, a sample interview question from each of the six themes and typical responses from respondents at suburban/rural schools are presented along with typical responses at urban schools. Governance Interviewees were asked, Does your school use a site base management approach? A principal in a suburban/rural school responded, We have a 15 member leadership forum that meets every three weeks. Recent discussions included a more flexible budget, the lunch room incentive program, and how to successfully include children with emotional disturbances in the after school program. An urban school principal responded, There is a School Improvement Team that meets once a week. Its membership includes teachers, administrators, parents, para-professionals,
123 support staff, and community members. Achievement was a topic of discussion at a recent meeting. There are no issues relating to special education children discussed because it is kept separate in the school. Accountability In response to the question How do you measure outcomes for your school? A suburban/rural respondent said, Our students are assessed at many stages to check their progress. As an example, all 10th grade students are required to take the PSAT, which is compared to their 8th grade standardized testing results. The Armed Services Vocational Battery is administered to every junior. Students in the 11th grade also take the SAT and the state mandated exit exam to determine graduation proficiency. An urban school respondent said, The state and district offices mandate our assessment measures. Currently, we use standardized tests, portfolios, and functional level assessments. Curriculum and Instruction To the question, Are there any particular models of instruction used at this school, e.g., block-scheduling, instructional teams, continuous progress, multi-age grouping, etc? The suburban/rural response was When block scheduling was implemented a few years ago, it was a mess. However, the principal stuck with it and with faculty input, the system was modified. Now it works real well. In an urban school the response was We use regular curriculum; no real models of instruction, a little team teaching goes on but not much. Includeness The question was asked, To what extent are a variety of educational environments and opportunities available to address the needs of students who have emotional disturbances? A suburban/rural staff member responded, The staff has
124 moved from segregating special education students to inclusion. Every child is everybodys responsibility. Regular education teachers are accepting of special education students and know they have the support of other staff. An urban staff member responded, Everybody is willing to try to meet the educational needs of students with emotional disturbances but we feel unprepared and have a lack of support and resources. Parent Involvement A staff member was asked To what extent do general education and special education parents support and attend school functions during the school year? At a suburban/rural school the response was Parent involvement increases every year. Several years ago there was very little participation but adaptations were made and 80 parents volunteered to be on the Parent Advisory Committee this year. At an urban elementary school the response was We try to have special projects to encourage their involvement but it doesnt always happen. More of the parents are working due to welfare changes and dont have as much time to attend school functions. Pro-Social Discipline Is there a school-wide program that promotes pro-social skills such as social skills curriculum, skill streaming, peer mediation, or conflict resolution? The response from the suburban/rural interviewee was We have peer mediation groups, anger management groups, and recovery groups for students on drugs or returning from a residential placement. Two years ago, a program was started for students who are at risk that teaches social skill and study habits. There is also a mentoring program in which regular education students are paired with special education
125 students in order to improve social skills. The response from an urban interviewee was No. The counselor is trying to do some of this, but not much has happened. The responses highlight the distinct differences between the suburban/rural schools and the urban schools. While there was variability between schools within each group, the responses were representative of the schools and their levels of engagement in school reform and restructuring activities. Research Question 5 The following section addresses the research question: What is the contribution of school reform activities in explaining the differences found in academic functioning, psychological functioning, and mental health service utilization? A series of multiple regression equations were conducted to describe the relationship between the predictor variables (IQ, gender, age, annual family income, SRRI percentage score) and the four outcome variables (reading achievement (WRAT Reading), math achievement (WRAT-Math), psychological functioning (CBCL Total T score) and mental health service use (a composite variable)). An investigation was undertaken between the predictor variables and the scores on each dependent measure. The intent of these analyses was to examine whether the independent variables could significantly predict student scores on the outcome measures of interest in the study. The independent variables were entered in two blocks, with the SRRI score entered separately in the second block. This technique was used to determine the unique contribution of the SRRI to the prediction while controlling for IQ, gender, age, and income. Each table presents
126 the unstandardized regression coefficient (B), standard error of B, the standardized regression coefficient (#), t value, and significance of t for each regression analysis. Reading Achievement A multiple regressions analysis was conducted between the variable, reading achievement, and the predictor variables, IQ, gender, age, income, and SRRI (see Table 34). Income contributed significantly to the prediction of reading achievement after adjusting for all other variables, (t = 2.54, p < .001). This indicated that, while controlling for the other variables in the equation, students whose parents had higher incomes had higher reading achievement scores. SRRI percentages were also a predictor of reading achievement. Schools that were more highly engaged in restructuring activities had students with emotional disturbances that had higher reading levels. IQ, gender, and age failed to significantly contribute additional variance to the explanation of reading achievement. The R2 of Model 1 was .226. In Model 2, with SRRI added to the prediction, R2 =. 324, resulting in a R2 change of .069. The R change was significantly different from zero, F (5,125) = 2.93, p = .015. Math Achievement A multiple regression analysis was conducted between the outcome variable, math achievement, and the predictor variables, IQ, gender, age, income and SRRI (see Table 35). Only IQ contributed significantly to the prediction of math achievement after adjusting for all the other variables, (t = 5.66, p<. 001). This indicates that, while controlling for the other variables in the equation, students with higher IQ scores had higher math achievement scores. Gender, age, income, nor SRRI significantly
127 contributed additional variance explaining math achievement. The R2 of Model 1 was .298. In Model 2, with SRRI added to the prediction, R2 = .313, resulting in a R2 change of .015. The R2 change was significantly different from zero, F (5,127) = 11.581, p = .000. CBCL Total Problem Score A multiple regression analysis was conducted between the outcome variable, CBCL Total T-score, and the predictor variables, IQ, gender, age, income, and SRRI (see Table 36). None of the variables significantly predicted CBCL Total Problem Score. The R2 of Model 1was .027. In Model 2, with SRRI added to the prediction, R2 remained constant at .027. Thus R2 change was not significantly different from zero, F (5,128) = .703, p = .622. Mental Health Service Utilization A multiple regression analysis was conducted between the outcome variable, mental health service utilization, and the predictor variables, IQ, gender, age, income, and SRRI (see Table 37). None of the variables significantly predicted mental health service use. The R2 of Model 1 was .031. In Model 2, R2 = .034, resulting in a R2 change of .003. The R change was not significantly different from zero F (5,128) = .900, p = .484.
128 Table 34 Multiple Regression Analysis for Reading Sc ores of Matched Participants (N = 132) Predictor B SE B $ T Sig. of T IQ 2.888E-03 .010 .026 .300 .765 Gender -.916 5.51 -.015 -.166 .868 Age .135 .505 .023 .267 .790 Income 2.534E-.04 .000 .189 2.083 .039* SRRI score .180 .065 .239 2.748 .007 Note R2 for Model 1 = .228, F(4,128) = 9.459, p = .000 Note R2 for Model 2 = .238, F (5,132) = 7.936, p = .000 *p < .05 Table 35 Multiple Regression Analysis for Math Sc ores of Matched Participants (N = 132) Predictor B SE B # T Sig. Of T IQ .447 .079 .451 5.660 .000* Gender -.227 4.279 -.004 -.053 .958 Age -.600 .378 -.118 -1.588 .115 Income 1.198E-04 .000 .108 1.342 .182 SRRI score 2.108 1.240 .134 1.699 .092 Note R2 for Model = .298, F(4, 128) = 13.555, p = .000 Note R2 for Model 2 = .313, F (5,132) = 11.581, p = .000 *p< .05
129 Table 36 Multiple Regression Analysis for CBCL Tota l Scores of Matched Participants (N = 132) Predictor B SE B # T Sig. Of T IQ 7.608E-02 .058 .124 1.312 .192 Gender .349 3.159 .010 .111 .912 Age -.228 .279 -.072 -.818 .415 Income -8.127E-05 .000 -.118 -1.241 .217 SRRI score 7.274E-02 .916 .007 .079 .937 Note R2 for Model 1= .027, F(4,129) = .884, p = .460 Note R2 for Model 2 = .027, F (5,128) = .703, p = .622 Table 37 Multiple Regression Analysis for Mental Health Service Utilization of Matched Participants (N = 132) Predictor B SE B # T Sig. Of T IQ -6.925E-03 .014 -.048 -.506 .614 Gender .351 .745 .044 .471 .639 Age .110 .066 .147 1.668 .098 Income 9.609E-06 .000 .059 .622 .535 SRRI score -.143 .216 -.062 -.664 .508 Note R2 for Model 1 = .031, F (4,129) = 1.019, p = .400 Note R for Model 2 = .034, F (5,133) = .900, p = .484
130 Summary The present study was a secondary analysis of data collected as part of two larger studies, the School and Community Study (suburban/rural) and the Urban School and Community Study (urban). The purpose of this study was to investigate the academic, social, and behavioral functioning of children with emotional disturbances and the relationship between this functioning and the level of school reform operating in suburban/rural versus urban communities. Multiple regression analyses were conducted to examine the relationship between school reform and restructuring and differences in academic and psychological functioning and mental health service use, while controlling for cognitive functioning, gender, age, and income. Findings indicated a significant difference in academic and psychological functioning and mental health service utilization for the matched sample of suburban/rural and urban students. Specifically, suburban/rural students had a higher level of academic achievement and functional impairment than their urban counterparts. Multiple regression analyses were conducted to assess the relative contribution of the predictor variables to academic achievement, emotional functioning, and mental health service use. Results indicated that higher IQ was significantly associated with higher scores in reading and math achievement; the school reform and restructuring score and family income significantly predicted reading achievement. None of the other variables significantly predicted differences in the criteria variables. In summary, the primary purpose of these analyses was to determine whether there were differences in school reform and restructuring activities operating in
131 suburban/rural and urban schools and the effects associated with these activities for a matched sample of suburban/rural students and urban students in special education classes due to emotional disturbances. IQ was a significant predictor of differences in academic functioning, while school reform and restructuring index score and family income predicted academic achievement. There were no significant relationships found with the other variables for students with emotional disturbances attending, suburban/rural schools and students with emotional disturbances attending urban schools.
132 CHAPTER FIVE DISCUSSION This chapter reviews the rationale, purpose, and methodology of the present study and discusses the results and limitations. The possible implications of the findings and areas for further research are also addressed. The chapter concludes with a summary of this information. Overview of the Study There are a vast number of empirical research studies investigating urbanicity, outcomes for children with emotional disturbances and school reform mechanisms. However, very few studies have studied those three variables collectively. Efforts to effect changes in student achievement through altering the manner in which schools operate have been countless. While extensive school reform efforts have been initiated across the nation, there is growing concern that these reforms are not benefiting all students, especially those with disabilities and those from culturally and geographically diverse groups (Patton & Edgar, 2002). Further, literature on school reform, though replete, is lacking a comprehensive theoretical framework linked to student outcomes and school reform initiatives that are inclusive of all students (Duchnowski & Kutash, 2003). In addition, there have only been a limited number of empirical reform studies conducted investigating the impact of school reform mechanisms involving a combination of multiple restructuring elements on student outcomes (Duchnowski, Townsend, Hocutt, &
133 McKinney, 1995) and the communities in which those students reside. Thus, while the school reform movement offers an opportunity to improve outcomes for children with emotional disturbances, there is a need to investigate the relationship between those outcomes and the urbanicity of the schools they attend. Review of the Method This study was a secondary analysis of data collected as part of two studies of children with emotional disturbances conducted by staff of the Research and Training Center for Childrens Mental Health, Louis de la Parte Mental Health Institute at the University of South Florida: the School and Community Study (Kutash, Duchnowski, Treder, Robbins, Kip, Oliveira, Greeson, Calvanese, & Black, 1999) and the Urban School and Community Study (Kutash, Duchnowski, Kip, Oliveira, Greeson, & Sheffield, 2001). Students from the School and Community Study were matched with students from the Urban School and Community Study on several variables (e.g., gender, annual family income, and age) to create the study sample (N=132). For the study participants, the following variables were examined: academic achievement, emotional functioning, and mental health service utilization outcome data along with school reform and restructuring data from their attending schools. The primary purpose of this study was to investigate the relationship between student outcomes and school reform activities. Another objective of this study was to compare students attending suburban/rural schools and a matched sample of students attending urban schools on academic achievement, emotional functioning, and mental health service utilization. Differences in the levels of engagement in reform and
134 restructuring activities between suburban/rural schools and urban schools were also explored. The last research question employed a series of multiple regression analyses to identify if the variation in student outcomes found in the previous research questions, could be accounted for by IQ, gender, age annual family income, and levels of engagement in reform and restructuring activities. Discussion of Findings The present study contributes to the empirical research base, the refinement of the school reform model, and the impact of school reform on students with emotional disturbances. This section presents a discussion of the findings to the study research questions on the differences between suburban/rural students with emotional disturbances and urban students with emotional disturbances in academic functioning, emotional functioning, an mental health services utilization. A discussion of the school reform and restructuring strategies employed by the schools these students attend and their impact on the students follows. Academic Functioning The first research question posed in this study was: Are there differences in academic functioning (i.e., reading and math achievement) between suburban/rural students in special education classrooms due to emotional disturbances and a matched sample of urban students in special education classrooms due to emotional disturbances? Research conducted by Lippman et al. (1996) and Duchnowski et al. (2000) found that suburban/rural students performed higher academically than urban students which were supported by the findings of the present study. Specifically, this study found differences
135 between the students with emotional disturbances attending suburban/rural schools and the students with emotional disturbances attending urban schools. The urban students scored significantly lower on the measures of academic achievement. However, both groups of students, suburban/rural and urban, performed significantly lower on the academic measures than students not enrolled in special education classes and all of the students with emotional disturbances were performing below grade level. These findings suggest that despite the No Child Left Behind Act, these students are, in fact, being left behind. Are they being left behind due to poverty? Although, poverty is a characteristic of urban areas, the urban students in this study were matched with the suburban/rural student on an income variable. Any discrepancies in the student populations based on income were reduced. Could the discrepancies in academic performance be attributed to the larger school and classroom sizes that plague urban areas? The schools selected to participate in the core studies were similar in school populations and the special education classes were similar in sizes. Perhaps, researchers should take a closer look at where students are during the school day to explain the academic differences between the suburban/rural students and the urban students. In this study, the urban students spent more time in special education setting than the suburban/rural students. Bradley et al. (2004) found that more time spent in regular education settings increased academic performance. Could this have been a contributing factor in the achievement results in this study?
136 While the findings of this study support the literature, it also underscores that schools are lagging in their goals to educate their students with emotional disturbances (Brown Center on Education Policy, 2001) despite reform and restructuring efforts. While there is some evidence that special education programs have embraced reform strategies, improvements have yet to be realized. Emotional Functioning Exploring the differences in the emotional functioning (i.e. symptomatology and functional impairment) between suburban/rural students in special education classrooms due to emotional disturbances and a matched sample of urban students in special education classrooms due to emotional disturbances was the focus of the second research question. The results of this study showed that the suburban/rural students were similar to the urban students in their emotional and behavioral functioning, as measured by the Child Behavior Checklist (CBCL). On average, both groups scored in the clinical range meaning that the behavioral problems and competencies of the children need professional intervention. Although students from urban environments are perceived to be more problematic and challenging (Lippman et al.,1996), the findings of this study indicate that the urban students did not exhibit more problem behaviors and were not more functionally impaired than their suburban/rural counterparts. In fact, the suburban/rural students were reported by their guardians as being more impaired than the urban students as reported by their guardians. The differences in levels of functionality may be attributed to the different assessment instruments used in the core studies. The Child and Adoliscent Functional Assessment Scale (CAFAS) and the Child and Adolescent Service
137 Assessment (CASA) measured similar constructs (e.g., home, community, relationships, and school) but in slightly different formats. The CAFAS, used in the suburban/rural study, is a longer, more complex instrument while the CASA is a shorter, more concise instrument. The differences between the instruments may have yielded the reported higher levels of disfunctionality in the suburban/rural students. However, neither measure of functional behavior emerged as a significant predicator of urbanicity or level of school reform and restructuring. Mental Health Service Utilization The third research question compared the suburban/rural students in special education classrooms due to emotional disturbances and a matched sample of urban students in special education classrooms due to emotional disturbances for differences in their mental health service utilization. Results of the School and Community study and the Urban School and Community study found that students attending schools that were actively engaged in restructuring activities received more services from agency professionals at school and more inpatient services than students attending less actively engaged schools (Duchnowski et al., 2001). Students in this study had a wide array of mental health services offered at school during the school day. The highest number of suburban/rural students utilized the group counseling and case management services from school personnel and other services (e.g., family counseling, probation officer, and group activities) from agency professionals, while the highest number of urban students utilized the individual and group counseling services from school personnel. The results of this study found that while the types of services differed between the suburban/rural school
138 setting and urban school setting, the number of students being served was constant. There were 58 students being served in both school settings. As schools are being held accountable for student achievement and mental health issues, staff with mental health expertise is being employed to work in conjunction with public mental health providers to offer a range of intervention (Lippmanet al., 1996; Duchnowski, et al. 2000). Consequently, students with emotional disturbances are receiving a wide array of mental health services both in educational settings and agency settings. There were also no significant differences between the two study populations and their mental health service utilization (e.g., inpatient services or outpatient services). In a study conducted by Marcenko, Keller, and Delaney (2001), urban caregiver respondents were asked to identify mental health services they felt were needed. Those services that impacted overall family functioning were identified as most desirable, such as recreational activities, counseling, and support services for the children. In this study, parents of both suburban/rural students and urban students reported high usage of inpatient hospital treatment and professional outpatient treatment. The parents of urban students reported a higher overall use of mental health services. These results suggest that there is a disconnect between the services parents want and the services students receive. Schools are focused on improving those behaviors that will increase academic functioning. While, parents of children who have emotional disturbances want information about strategies and services that will improve behavioral functioning outside of the school boundaries and they report making greater efforts to get services for their children than parents of children in other disability groups (Wagner, Kutash,
139 Duchnowski, Epstein, and Sumi, 2004). Clearly, school administrators and parents need to communicate about the services that will improve the overall functioning of the students. School Reform and Restructuring Levels Schools across the nation are implementing reform and restructuring programs in efforts to create learning environments that are responsive to a wider array of student learning needs (Lippman et al., 1996). While comparisons between suburban/rural schools and urban schools abound, few research studies have been conducted investigating these comparisons and their efforts to reform. The third research question posed in this study explored the differences in the school reform activities (e.g., governance, accountability, pro-social discipline, accountability, inclusion, family involvement, curriculum and instruction) between the suburban/rural schools attended by students with emotional disturbances and the urban schools attended by students with emotional disturbances. Using the School Reform and Restructuring Index (SRRI), schools in the present study were compared. Of the eight schools located in the urban areas, six schools were engaged in lower levels of reform and restructuring activities than all four of the schools located in suburban/rural areas. Interestingly, four of the eight urban schools had been identified by district administrators as highly engaged schools. Only two of those schools had SRRI scores on par with the suburban/rural schools. The suburban/rural schools scores could be reflective of more resources and community support or other factors, outside the scope of the ones measured by the SRRI. Clearly, the perception of levels of activity in reform and restructuring varies which emphasizes
140 the continued need for research to develop a measure by which to measure, empirically, school reform. Study Limitations A number of limitations must be considered when interpreting the above mentioned findings. This section begins by summarizing the limitations described in Chapter One. This is followed by a discussion of additional limitations discovered during the course of the study. This study is a secondary analysis of data from two research projects. While the advantages of using existing data sets include time efficiency, reduced cost, and the nonintrusive means of analysis, the researcher was limited by research design, variables of interest, instrumentation, participants, and data collection process. The inability to administer the same instruments to both study populations limited the level of conclusions that could be drawn by the results. The variances in data collection methods required the creation of new metrics and variables. Income data was gathered in different formats (categorical and continuous) in the two studies and a new metric was created to analyze the data. The studies also varied in the instruments used to measure levels of functional impairment and mental health service utilization. Schools were selected to participate in the core studies base on the core studies based on their levels of engagement in school and reform activities. All of the schools in the School and Community Study were selected based on their exemplary approaches to school reform while the schools selected to participate in the Urban School and
141 Community Study represented both highly active and less active approaches to school reform. This sampling strategy may reduce the widespread use of the findings. The generalizability of the results to other public schools may also be limited because the schools, in both studies, were purposively selected based on their urbanicity. Schools located in suburban/rural areas were attended by a majority of white students while the schools located in the urban areas were attended by a majority of black students. While race was not a variable of interest, the complex relationship between race and urbanicity may have effected the outcomes investigated in this study. The generalizability of the findings is further hampered by the matching process that was employed. Students from the School and Community Study were visually paired with students from the Urban School and Community Study on three key variables (e.g., gender, annual family income, and age). The process decreased the disparities between the groups and analyses were not affected by other variables. However, the matching procedure has limitations, other variables may have created different matches and suitable matches could not be made for al of the participants. Analyses were conducted comparing the study sample participants with the remaining participants and the two samples were similar on the study variables. Nevertheless, matching the study participants limits the application of these results to a wider population. The methods of the study were limited to correlational analyses. Because causal statements cannot be made, analyses were confined to the examination of the differences in student outcomes and the levels of engagement in school restructuring. In addition, other student and school variables not included in the analyses may have contributed to
142 the differences in outcomes. For example, the nature and quality of a students experience in the classroom may positively impact academic performance. Further, this study did not consider the impact of parallel reform efforts in the various child serving systems on student outcomes. In addition, no general education students were used, inhibiting the examination of the different ways in which special education students might respond to restructuring and reform efforts. Furthermore, the methodology utilized in the study limits the investigation to the examination of the aggregate effects of reform and restructuring on a specific disability group. It does not allow for the determination of how such efforts may differentially impact students with varying levels of academic and behavioral functioning, such as those of other disability groups. The education community may be assuming that reform efforts influence all students in the same way. Recommendations for Future Research With the above limitations in mind, the findings from this study do have methodological, theoretical, and practical implications. This study, while exploratory in nature, contributes to the limited empirical literature base examining the relationship between urbanicity, school reform, and students with emotional disturbances and opens several research doors for future exploration. The first door to be entered is that of developing research that incorporates general and special education students. Past research projects have investigated the issues of each of these student populations separately. Current reform and restructuring strategies were created for a whole school implementation. Exploring the effects of these strategies on both student populations
143 yields a clearer perspective on the overall quality of the reform efforts. Can a whole school reform philosophy benefit the whole school? Another research door to open requires the incorporation of all of the various disability groups in studies. The same pitfalls that emerge from looking at general education and special education students in isolation arise from investigating a single disability group. Educators would benefit from a better understanding of how reform strategies impact all children with disabilities. Future reform and restructuring research should also explore the attendance and discipline records of the students with disabilities. Investigating variables such as the number of days a student walks through the school doors along with where the student spends the school day (e.g., in-school suspension) would lead to the creation of a more informed explanation of achievement outcomes. Low attendance rates and high suspension rates signal poor educational outcomes for the students in this population. Clearly, studies that align these school rates with school reform strategies for analyses would benefit the educational field. Opening the research door to scrutinize teacher quality will lead to a more comprehensive study of urbanicity and the outcomes for children with emotional disturbances. The number of years of teaching experience, the level of training, and certification along with that of school administrators will guide researchers to r each more far reaching and compelling conclusions. Finally, the development of a standardized scale to measure school reform and restructuring is needed to successfully compare the various strategies implemented nationwide. A scale capable of assessing the reform strategies schools are implementing
144 for effectiveness in improving student outcomes in both general and special education classrooms. A scale designed to score the changes in schools by evaluating key themes. The scores would guide the schools efforts to improve in-service programs in specific areas, such as pro-social discipline (Duchnowski, Kutash & Olivera, 2004). The SRRI is one such scale that has identified themes in reform and restructuring models to assess and examine those activities schools are engaged in to improve academic outcomes. Summary There has been a great deal of literature related to the differences between schools located in urban and suburban/rural locations along with comparisons of the students that attend those schools. A common thread in these reports is the dissatisfaction with the educational outcomes for all of the students, which has spurred on the reform and restructuring movement. While the reforms have provided guidance to improve academic achievement, there has not been a system to measure overall success. This study compared students with emotional disturbances attending schools located in suburban/rural and urban areas on several variables. It was found that the urban students scored significantly lower academically, accessed different mental health services, and were also reported to be in the clinical range of psychological impairment. The schools that the students in both geographical areas attended were similar in their levels of reform and restructuring activities. It is imperative that a standardized scale be developed to measure the effectiveness of the school and reform models that schools are employing. The SRRI has laid the foundation for future researchers to build a universal measure. Once a
145 standardized measure has been created, educators and school administrators can make informed decisions about altering their strategies with the goal being to increase academic achievement. This scale will be an important tool for the improvement of the educational outcomes for the general education student population and for those students receiving services in special education classrooms.
146 REFERENCES Achenbach, T.M. (1991). Manual for the Child Behavior Checklist/4-18 and 1991 profile Burlington, VT: University of Vermont, Department of Psychiatry. Achenbach, T.M. & Edelbrock, C.S. (1981). Behavioral problems and competencies reported by parents of normal and disturbed children aged four through sixteen. Chicago, IL: University of Chicago Press for the Society for Research in Child Development. Artiles, A. J. & Trent, S.C. (1994). Overrepresentation of minority students in special education: A continuing debate. The Journal of Special Education, 27 410-437. Ascher, B.H., Farmer, E.M.Z., & Burns, B.J. (1996). The Child and Adolescent Services Assessment (CASA): Description and psychometrics. Journal of Emotional and Behavioral Disorders, 4 12-20. Balfanz, R. (2000). Why do so many urban public school students demonstrate so little academic achievement? In M.G. Sanders (Ed) Schooling students placed at risk New Jersey: Lawrence Erlbaum Associates, Publishers. Bird, H.R., Shaffer, D., Fisher, P., Gould, M.S., Staghezza, B., Chen, J.Y., & Hoven, C. (1993). The Columbia Impairment Scale (CIS): Pilot findings on a measure of
147 global impairment for children and adolescent. International Journal of Methods in Psychiatric Research, 3 167-176. Bower, E.M. (1982). Defining emotional disturbance: Public policy and research. Psychology in the Schools, 19 55-60. Burns, B.J. (1991). Mental health service use by adolescents in the 1970s and 1980s. Journal of the American Academy of Child and Adolescent Psychiatry, 30 (1), 144150. Burns, B.J., Anglod, A., Magruder-Habib, K., Costello, E.J., & Patrick, M.K.S. (1996). The Child and Adolescent Services Assessment (CASA). Durham, Duke University Medical Center, Developmental Epidemiology Program. Carlson, D. (1998). Education as a political issue (I). In H.S. Shapiro & D.E.Purpel (Eds.) Critical Social Issues in American Education Mahwah, NJ: Lawrence Erlbaum Associates, Publishers. Chinn, P.C. & Hughes, S. (1987). Representation of minority students in special education classes. Remedial and Special Education, 8 41-46. Conant, J.B. (1961). Slums and suburbs New York: McGraw-Hill. Crosby, E. (1999). Urban schools: Forced to fail. Phi Delta Kappan December, 298-307. Cuban, L. (2001). Why bad reforms wont give us good schools. American Prospect 12 46-8.
148 Cullinan, D., Epstein, M.H., & Sabornie, E.J. (1992). Selected characteristics of a national sample of serious emotionally disturbed adolescents. Behavior Disorders, 17 273-280. Danielson, L.C. & Malouf, D.B. (1994). Federal policy and educational reform: Achieving better outcomes for students with disabilities. Special Services in the Schools, 9 11-19. Dedrick, R.F., Greenbaum, P.E., Friedman, R.M., & Wetherington, C.M. (1997). Testing the structure of the Child Behavior Checklist/4-18 using confirmatory factor analysis Educational and Psychological Measurement, 57 306-313. Duchnowski, A., Hall, B., Kutash, K., & Friedman, R. M. (1998). The Alternatives to Residential Treatment Study (ARTS): 12 month outcomes. In M.H. Epstein, K. Kutash, & A.J. Duchnowski (Eds.), Outcomes for children with emotional and behavioral disorders and their families: Program and evaluation best practices (pp. 55-80). Austin, TX: Pro-Ed.RTS. Duchnowski. A.J., Kutash, K., & Oliveira, B. (2204). Systematically examining school improvement activities including special education. Remedial and Special Education 25(2) 117-129. Dunlap, G.E. & Childs, K.E. (1996). Intervention research in emotional and behavioral disorders: An analysis of studies from 1980-1993. Behavioral Disorders, 21 125-136.
149 Economic Research Service (1993). Measuring rurality: Rural-urban continuum codes U.S. Department of Agriculture. Washington, DC: US Government Printing Office. Education of the Handicapped Act. (1977). Federal Register, 42 p. 42478. Amended in Federal Register 1981, 46 p. 3866. Entwisle, D.R. & Alexander, K.L. (1995). A parents economic shadow: Family structure versus family resources as influences on early school achievement. Journal of Marriage and Family, 57 399-409. Farmer, E.M.Z., Angold, A., Burns, B. J., & Costello, E.J. (1994). Reliability of self-reported service use: Test-retest consistency of childrens responses to the Child and Adolescent Services Assessment (CASA). Journal of Child and Family Studies, 3 307325. Farmer, E.M.Z., Stangl, D.K., Burns, B.J., Costello, E.J., & Angold, A. (1999). Use, persistence, and intensity: Patterns of care for childrens mental health across one year. Community Mental Health Journal, 35 (1), 31-46. Federal Register (1983, February 10). 58(28), 1938. Washington, DC: U.S. Government Printing Office. Forness, S.R. (1997). Personal reflections: Definition.. In J. M. Kauffman (Ed.). Characteristics of Emotional and Behavioral Disorders of Children and Youth, pp. 3436 Upper Saddle River, NJ: Prentice-Hall, Inc.
150 Forness, S.R., Kavale, K.A. & Lopez, M. (1993). Conduct disorders in school: Special education eligibility and comorbidity. Journal of Emotional and Behavioral Disorders, 1 101-108. Forness, S. R. & Knitzer, J. (1992). A new proposed definition and terminology to replace serious emotional disturbance in individuals with disabilities education act. School Psychology Review, 21 12-20. Friedman, R.M. (1996). Child mental health policy. In B.L. Levin & J. Peetrila (Eds.), Mental Health Services: A Public Health Perspective 234-249. New York: Oxford University Press. Friedman, R.M., Katz-Leavy, J.W., Manderscheid, R.W., & Sondheimer, D.L. (1998). Prevalence of serious emotional disturbance in children and adolescents. An update. In R.W. Manderscheid & M.A. Sonnenschein (Eds.), Mental health, United States, 1998 Washington, DC: U.S. Government Printing Office. Friedman, R.M., Kutash, K., & Duchnowski, A.J. (1996). The population of concern: Defining the issues. In B.Stroul (Ed.), Childrens Mental Health: Crossing Systems of car in a changing society, (pp. 69-89). Baltimore, MD: Paul H. Brookes. Frechtling, J. (2000). Evaluating systemic educational reform: facing the methodological, practical, and political challenges. Arts Education Policy Review, 101 25-30. Fuchs, D., & Fuchs, L.S. (1994). Inclusive schools movement and the radicalization of special education reform. Exceptional Children, 60 294-309.
151 Goldberg-Glen, R., Sands, R.G., Cole, R.D., & Cristofalo, C. (1998). Multigeneraltional patterns and internal structures in families in which grandparents raise grandchildren. Families in Society: The Journal of Contemporary Human Services, 30 477-489. Goodall, C., Kafadar, K., & Tukey, J.K. (1998). Computing and using rural versus urban measures in statistical applications. The American Statistician, 52 (2), 101-1.11 Gottlieb, J., Gottlieb, B.W., & Wishner, J. (1994). Special education in urban America: Its not justifiable for many. Journal of Special Education, 27 453-465. Greenbaum, P.E., Dedrick, R.F., Friedman, R.M., Kutash, K., Brown, E.C., Lardieri, S., & Pugh, A.M. (1996). National Adolescent and Child Treatment (NACTS): Outcomes for children with serious emotional and behavioral disturbance Journal of Emotional and Behavioral Disorders, 4 130-146. Greenbaum, P.E., Dedrick, R.F., Friedman, R.M., Kutash, K., Brown, E.C., Lardieri, S., & Pugh, A.M. (1998). National Adolescent and Child Treatment (NACTS): Outcomes for children with serious emotional and behavioral disturbance. In M.H. Epstein, K. Kutash, and A. Duchnowski (Eds.), Outcomes for children and youth with behavioral and emotional disorders and thei r families: Programs and evaluation best practices (pp. 21-54). Austin, TX: PRO-ED. Harry, B. & Anderson, M.G. (1994). The disproportionate placement of African American males in special education programs: A critique of the process. Journal of Negro Education, 63 602-619.
152 Hess, F.M. (1998). As the policy churns: Toil and trouble for urban school reform. The Education Digest 64 8-14. Hoagwood, K., & Erwin, H.D. (1997). Effectiveness of school-based mental health services for children: A 10-year research review Journal of Child and Family Studies, 6 435-451. Hodges, K. (2000). Predicting service utilization with the Child and Adolescent Functional Assessment Scale in a sample of youths with serious emotional disturbance served by Center for Mental Health Services funded demonstrations. Journal of Behavioral Health Services & Research, 27 47-59 Hodges, K., Bickman, L., Kurtz, S., & Reiter, M. (1991). A multi-dimensional measure of level of functioning for children and adolescents. In A. Algarin & R.M. Friedman (Eds ), Fourth Annual Research Conference Proceedings. A System of Care for Childrens Mental Health: Expanding the Research Base (pp. 149-154). Tampa: University of South Florida, Research and Training Center for Childrens Mental Health, Louis de la Parte Mental Health Institute. Hodges, K., Doucette-Gates, A., & Liao, Q. (1999). The relationship between the Child and Adolescent Functional Scale (CAFAS) and indicators of functioning. Journal of Child and Family Studies, 8 109-122. Hodges, K. & Wong, M.M. (1996). Psychometric characteristics of a multidimensional measure to assess impairment: The Child and Adolescent Functional Assessment Scale. Journal of Child and Family Studies, 5 445-467.
153 Kauffman, J.M. (1993). How we might achieve the radical reform of special education. Exceptional Children, 60 6-16. Kauffman, J.M. (1997). Characteristics of Emotional and Behavioral Disorders of Children and Youth Upper Saddle River, NJ: Prentice-Hall, Inc. Knapp, M.S. (1995). How shall we study comprehensive, collaborative services for children and families? Educational Researcher, 24 5-16. Knitzer, J. (1996). Meeting the mental health needs of young children and their families. In B.A. Stroul (Ed.), (pp. 553-574). Baltimore, MD: Brookes. Koyangi, C., & Gaines, S. (1993). All systems failure: An examination of the results of neglecting the needs of children with serious emotional disturbance. Alexandria, VA: National Mental Health Association. Kretovics, J. & Nussel, E. J. (Eds.), (1994). Transforming urban education Boston, MA: Allyn & Bacon. Kutash, K., Duchnowski, A.J., Rivera, V.R., Oliveira, B., Kelly, P. (1997). School and community study. Technical report on phase I Tampa, FL: Research and Training Center for Childrens Mental Health, Louis de la Parte Florida Mental Health Institute, University of South Florida. Kutash, K., Duchnowski, A.J., Kip, S., Greeson, M., Sheffield, S., & Oliveira, B. (2001). Urban school and community study. Technical report volume II Tampa, FL: Research and Training Center for Childrens Mental Health, Louis de la Parte Florida Mental Health Institute, University of South Florida.
154 Kutash, K., Oliveira, B., & Rivera, V. R. (1997). The Child and Adolescent Functioning Assessment Scale Research Form (CAFAS Research). Tampa, FL: Louis de la Parte Mental Health Institute, University of South Florida. Lippman, L., Burns, S., McArthur, E., & National Center for Educational Statistics. (1996). Urban schools: The challenge of location and poverty http://nces.ed.gov/pubsearch/pubsinfo.asp?pubid=96184 Marcenko, M.O., Keller, T.E., & Delaney, M.A. (2001). Children with SED and their families in an urban public mental health system: Characteristics, needs, and expectations. Journal of Child and Family Studies, 10 (2) 213-226. Maxwell, S.E. & Delaney, H.D. (2000). Designing experiments and analyzing data. Mahwah, NJ: Lawrence Erlbaum Associates. McKinnon, J. & Humes, K. (2000). The black population in the United States: March 1999. U.S. Census Bureau (Current Population Reports, Series P20-530). Washington, DC: U.S. Government Printing Office. Mitchell, P.B. (1988). Tower of Babel or shared vision? Teaching Exceptional Children, 21 49-51. Mulkey, L.M., Crain, R.L., & Harrington, A.J.C. (1992). One-parent households and achievement: Economic and behavioral explanations of a small effect. Sociology of Education, 65 48-65.
155 Muscott, H.S. (1995). A process for facilitating the appropriate inclusion of students with emotional/behavioral disorders. Education and Treatment of Children, 18 369-386. National Center for Children in Poverty (2001). Child poverty fact sheet New York, NY: Joseph L. Mailman School of Public Health of Columbia University. National Center for Education Statistics (1997). Characteristics of small and rural school districts Washington, DC: U.S. Department of Education. National Commission on Excellence in Education (1983). A nation at risk Washington, DC: U.S. Department of Education. National Research Council (2002). Minority students in special and gifted education Washington, DC: National Academy Press. Oswald, D.P., Coutinho, M.J., & Best, A.M. (1998). Ethnicity in special education: The influence of school-related economic and demographic variables. The Journal of Special Education, 32 194-206. Patton, J.M. (1998). The disproportionate representation of African Americans in special education: Looking behind the curtain for understanding and solutions. Journal of Special Education, 32 25-31. Paul, J.L., & Rosselli, H. ( 1995). Integrating the parallel reforms in general and special education. In J.L. Paul, H. Rosselli, & D. Evans (Eds.), Integrating school restructuring and special education reform Forth Worth, TX: Harcourt Brace & Company.
156 Peacock Hill Working Group (1991). Problems and promises in special education and related services for children and youth with emotional or behavioral disorders. Behavioral Disorders, 16 299-313. Quinn, K.P., & Epstein, M.H. (1998). Characteristics of children, youth, and families served by local interagency systems of care. In M.H. Epstein, K. Kutash, and A. Duchnowski (Eds.), Outcomes for children and youth with behavioral and emotional disorders and their families: Programs and evaluation best practices (pp. 81-114). Austin, TX: PRO-ED. Robbins, V. (1999). Academic and behavioral functioning of students with emotional and behavioral disabilities served in restructuring public schools University of South Florida Dissertation Rodgers, A.Y. & Jones, R.L. (1999). Grandmothers who are caregivers: An overlooked population. Child and Adolescent Social Work Journal, 16, 455-460. Shouse, R.C. & Mussoline, L.J. (2000). High risk, low return: The achievement effects of restructuring in disadvantaged schools Social Psychology of Education, 3, 245-259. Silver, S.E., Unger, K.V., & Friedman, R.M. (1994). Transition to young adulthood among youth with emotional disturbance Unpublished manuscript, University of South Florida, Florida Mental Health Institute, Department of Child and Family Studies.
157 Silver, S.E., Duchnowski, A.J., Kutash, K., Friedman, R.M., Eisen, J., Prange, M., Brandenburg, N., & Greenbaum, P. (1992). A comparison of children with serious emotional disturbance served in residential and school settings. Journal of Child and Family Studies, 1, 43-59. Slavin, R. E. (2001). Putting the school back in school reform. Educational Leadership, 58 22-27. Smith, M., & ODay, J. (1991). Putting the pieces together: Systemic school reform. CPRE Policy Brief. New Brunswick, NJ: Eagleton Institute of Politics. Stevens, J. (1996). Applied multivariate statistics for the social sciences (3rd ed.) New Jersey: Lawrence Erlbaum Associates. Stiffman, A. R., Horwitz, S. M., Hoagwood, K., Compton, W., Cottler, L., Bean, D. L., Narrow, W. E., & Weisz, J. R. (2000). The service assessment for children and adolescents (SACA): Adult and child report. Journal of the American Academy of Child and Adolescent Psychiatry, 39 1032-1039. U.S. Census Bureau (2000). School enrollment social and economic characteristics of students. U.S. Department of Commerce. Washington, DC: U.S. Government Printing Office. U.S. Department of Education (1996). Eighteenth annual report to Congress on the implementation of the Individuals with Disabilities Education Act. Washington, DC: U. S. Government Printing Office.
158 U.S. Department of Education (2001). Office of the Secretary. No Child Left Behind Washington, DC: U.S. Government Printing Office. Vanderwood, M., McGrew, K.S., & Ysseldyke, J.E. (1998). Why we cant say much about students with disabilities during education reform. Exceptional Children, 64 359-370. Wagner, M.M. (1995). Outcomes for youths with serious emotional disturbance in secondary school and early adulthood. Critical Issues for Children and Youths, 5 90112. Wagner, M., Kutash, K., Duchnowski, A.J., Epstein, M.H., & Sumi, W.C. (in press). The children we serve: A national picture of the characteristics of students with emotional disturbances receiving special education. Journal of Emotional and Behavioral Disorders. Wagner, M., Newman, L., DAmico, R., Jay, E., Butler-Walum, P., Marder, C., & Cox, R. (1991). Youth with disabilities. How are they doing? The first comprehensive report from the National Longitudinal Transition Study of Special Education Students Menlo Park, CA: SRI International. Walker, R., & Brunson, T.D. (1995). After high school: The status of youth with emotional and behavioral disorders Career Development for Exceptional Individuals, 18, 97-107. Walrath, C.M., Nickerson, K.J., Crowel, R.L., & Leaf, P.J. (1998). Serving children with serious emotional disturbance in a system of care: Do mental health and
159 non-mental health agency referrals look the same? Journal of Emotional and Behavioral Disorders, 6 205-213. Weist, M.D., Paskewitz, D.A., Warner, B.S., & Flaherty, L.T. (1996). Treatment outcome of school-based mental health services for urban teenagers. Community Mental Helath Journal, 32 149-157. Wilkinson, G.S. (1993). The Wide Range Achievement Test, 3rd edition Wilmington, DE: Wide Range. Williams, V.S.L., Jones, L.V., & Tukey, J.W. (1999). Contro lling error in multiple comparisons with examples from state-to-state differences in educational achievement. Journal of Educational and Behavioral Statistics, 24 42-69. Yegidis, B.L., & Weinbach, R.W. (1991). Research methods for social workers White Plains, NY: Longman. Yell, M.L. & Shiner, J.G. (1997). The discipline of students with disabilities: Requirements of the IDEA Amendments of 1997. Education and Treatment of Children, 21 246-257. Yin, R.K. (1993). Case study research: Design and methods (rev.ed.). Beverly Hills, CA: Sage.
160 ABOUT THE AUTHOR Karen M. Harris has worked with children and families over the past twenty years. After completing her Bachelors in Economics from Duke University, Karen completed both a Master of Education Degree and Specialist in Education Degree at the University of Florida in Counselor Education, with an emphasis in mental health counseling. She worked as a Family Life Coordinator for the Army Social Works Services in Katterbach, Germany While completing her doctoral work, Karen worked as a research assistant at the Research and Training Center for Childrens Mental Health at the Louis de la Parte Florida Mental Health Institute. Karen conducts presentations locally and nationally and is active in several professional organizations.