USF Libraries
USF Digital Collections

Baker Act examination referrals among children and adolescents

MISSING IMAGE

Material Information

Title:
Baker Act examination referrals among children and adolescents an analysis of school related variables
Physical Description:
Book
Language:
English
Creator:
Beam, Bradley Scott
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla
Publication Date:

Subjects

Subjects / Keywords:
Involuntary
Examination
Psychopathology
District
Demographic
Dissertations, Academic -- School Psychology -- Doctoral -- USF   ( lcsh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Abstract:
ABSTRACT: The Baker Act is legislation that provides civil rights for individuals suspected of having a mental illness and may be in need of involuntary psychiatric evaluation. Its intent is to prevent the indiscriminate placement of individuals in residential treatment facilities and other restrictive placements. Referrals for involuntary psychiatric evaluation under Baker Act statutes have increased among children and adolescents in recent years, raising concerns related its use with this population. The purpose of this study was to explore the relationship between school based factors and the use of the Baker Act among children and adolescents. Results indicate that the use of the Baker Act is more prevalent receiving among the high school population, females, Other/mixed children and adolescents, and White children and adolescents. Multiple regression analyses indicated that school district variables (e.g., minority enrollment, graduation rates, out-of-school suspension, etc.) contributed to a statistically significant proportion of the variance in Baker Act ER rates and repeat Baker Act ER rates among the 67 counties in the state of Florida. The percent of students in a district that graduated with a standard diploma was the variables that most consistently contribute to a unique proportion of the variance in Baker Act ER and repeat Baker Act ER rates. The findings from the study have implications for the design of mental health and behavioral support systems for children and adolescents. Additional research is necessary to more closely examine the relationship between demographics, school related variables, and the use of the Baker Act.
Thesis:
Dissertation (Ph.D.)--University of South Florida, 2007.
Bibliography:
Includes bibliographical references.
System Details:
System requirements: World Wide Web browser and PDF reader.
System Details:
Mode of access: World Wide Web.
Statement of Responsibility:
by Bradley Scott Beam.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 164 pages.
General Note:
Includes vita.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001926537
oclc - 191884751
usfldc doi - E14-SFE0001912
usfldc handle - e14.1912
System ID:
SFS0026230:00001


This item is only available as the following downloads:


Full Text
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 001926537
003 fts
005 20080208135030.0
006 m||||e|||d||||||||
007 cr mnu|||uuuuu
008 080208s2007 flu sbm 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0001912
035
(OCoLC)191884751
040
FHM
c FHM
049
FHMM
090
LB1051 (ONLINE)
1 100
Beam, Bradley Scott.
0 245
Baker Act examination referrals among children and adolescents :
b an analysis of school related variables
h [electronic resource] /
by Bradley Scott Beam.
260
[Tampa, Fla] :
University of South Florida,
2007.
3 520
ABSTRACT: The Baker Act is legislation that provides civil rights for individuals suspected of having a mental illness and may be in need of involuntary psychiatric evaluation. Its intent is to prevent the indiscriminate placement of individuals in residential treatment facilities and other restrictive placements. Referrals for involuntary psychiatric evaluation under Baker Act statutes have increased among children and adolescents in recent years, raising concerns related its use with this population. The purpose of this study was to explore the relationship between school based factors and the use of the Baker Act among children and adolescents. Results indicate that the use of the Baker Act is more prevalent receiving among the high school population, females, Other/mixed children and adolescents, and White children and adolescents. Multiple regression analyses indicated that school district variables (e.g., minority enrollment, graduation rates, out-of-school suspension, etc.) contributed to a statistically significant proportion of the variance in Baker Act ER rates and repeat Baker Act ER rates among the 67 counties in the state of Florida. The percent of students in a district that graduated with a standard diploma was the variables that most consistently contribute to a unique proportion of the variance in Baker Act ER and repeat Baker Act ER rates. The findings from the study have implications for the design of mental health and behavioral support systems for children and adolescents. Additional research is necessary to more closely examine the relationship between demographics, school related variables, and the use of the Baker Act.
502
Dissertation (Ph.D.)--University of South Florida, 2007.
504
Includes bibliographical references.
516
Text (Electronic dissertation) in PDF format.
538
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
500
Title from PDF of title page.
Document formatted into pages; contains 164 pages.
Includes vita.
590
Adviser: George M. Batsche, Ed.D.
653
Involuntary.
Examination.
Psychopathology.
District.
Demographic.
690
Dissertations, Academic
z USF
x School Psychology
Doctoral.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.1912



PAGE 1

School Related Variables by Bradley Scott Beam A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Psychological and Social Foundations College of Education University of South Florida Major Professor: George M. Batsche, Ed.D. Michael J. Curtis, Ph.D. Kathleen H. Armstrong, Ph.D. John M. Ferron, Ph.D. Annette Christy, Ph.D. Date of Approval: March 22, 2007 Keywords: involuntary, examination, psychopathology, district demographic Copyright 2007, Bradley Scott BeamBaker Act Examination Referrals Among Children and Adolescents: An Analysis of

PAGE 2

Dedication This dissertation is dedicated to my many wonderful friends, who have consistently provided the s upport and laughter necessary fo r persevering throughout this arduous process. My three wonde rful older sisters, Karla; Angie; and Kerry; who have had to deal with my personality, the cons equence of being the youngest and only male child in the family! To my beautiful mother, who has provided consistent, unconditional love and support throughout my college a nd graduate school studies. Without this support, I would not have been able to realize my professional goals, and more importantly, I would not have been able to de velop the life perspec tive that has allowed me to persevere during the most difficult time s. To my father, whose life and death has shaped my life in incalculable ways. You ar e always missed, never forgotten, and always in my heart.

PAGE 3

Acknowledgements I would like to thank the fo llowing individuals, all of wh om have contributed to the completion of this process. George Batsche, whose never ending insistence on perfection, competence, and precise analyses is responsible for shaping a work ethic that has allowed me to make important steps to wards attaining my professional goals. His guidance and supervision has he lped me to achieve an unde rstanding of and appreciation for the most important conceptual and theore tical foundations that guide the practice of school psychology. Michael Curtis, whose leader ship and stewardship of the profession has inspired me to reach for the highest st andards of practice in school psychology while maintaining a humble attitude. Kathy Arms trong, whose consistent support of my professional development and introduction to a clinic setting helped to enhance my skills and broader my professional goals. John Ferr on, whose teaching style is responsible for increasing my understanding and a ppreciation of research design principles and statistical analyses. Annette Christy, whose willingness to collaborate was essential in completing this project.

PAGE 4

i Table of Contents List of Tables iv List of Figures vi Abstract vii Chapter One 1 Exclusion from Schooling 2 Types of Exclusion 3 The Baker Act 3 Accountability and Reform 5 Students with EBD 6 Developmental Models and EBD 8 Prevention and Intervention 8 Rationale for the Study 9 Purpose of the Study 11 Research Hypotheses 12 Chapter Two 14 School Exclusion 14 Rates of Disciplinary Exclusion 15 Behaviors that Lead to Disciplinary Exclusion 17 Student Characteristics and Disciplinary Exclusio n 19 Academic Achievement and Exclusion 22 Special Education and Exclusion 23 Cumulative Effects of Demographic Variables 25 Baker Act Examination Referrals 27 Prevalence of Baker Act referrals 28 Baker Act Referrals and School Based Behavior 30 Models of Psychopathology 31 Medical Models 32 Developmental Models 34 Risk and Resilience Models 38 Risk and Correlated Constraints 40 Support Services 42 Tier Models of Service Delivery 42 Tier Models and Restrictive placement 45 Availability of Support Services 46

PAGE 5

ii School Based Support Services 47 Factors Associated with Restrictive Placement 55 Severity of Behavior 56 Availability and Exclusion 5 8 Discipline and Exclusion 63 Conclusion 68 Chapter Three 71 Participants 71 Research Design 74 Independent Variables 74 Dependent Variables 75 Archival Databases 75 Data Retrieval Procedures 76 Steps in Data Retrieval 7 7 Data Analysis 79 Chapter Four 84 Data Transfer and Procedural Integrity 86 Research Question 1 87 Age 89 Gender 91 Race/Ethnicity 95 Research Question 2 10 1 Research Questions 3 and 4 1 03 Research Questions 5 104 Descriptive Statistics 105 Collinearity 109 Assumptions of Multiple Regression 110 Outliers 110 Multiple Regression Analyses 1 11 Chapter Five 121 Research Questions One and Two 121 Research Questions Three and Four 124 Research Question Five 125 Limitations 134 Implications and Future Directions 137 Conclusion 142 References 143 Appendices 157 Appendix A: Baker Act Initiation Form 15 8

PAGE 6

iii Appendix B: Baker Act Cover Form 159 Appendix C: Coding of District Level Variables 160 Appendix D: Data Transfer and Procedural Integrity Form 162 Appendix E: Scatterplot Displaying Homoscedasticit y for Selected Multiple Regression 164 About the Author End Page

PAGE 7

iv List of Tables Table 1 Abbreviations for Relevant Terminology 85 Table 2 Distribution of Baker Act Examination Refer rals (ERs) By 88 Grade Level Table 3 Who Received Baker Act Examination Referrals (ERs ) 89 Table 4 Odds Ratios Comparisons between Grade Level s for Children and adolescents Who Received Baker Act Examination Referrals (ERs) 89 Table 5 Chi-Square Statistics for Comparison of Bak er Act Examination Referral Proportions between Grade Le vels 91 Table 6 Distribution of Baker Act Examination Refer rals among Males and Females 92 Table 7 Baker Act Examination Referrals (ERs) 93 Table 8 Odds Ratio Comparisons between Males and Fe males Who Received Baker Act Examination Referrals (ERs ) 93 Table 9 Chi-Square Statistics for Comparison of Bak er Act Examination Referral (ER) Proportions between Mal es and Females 95 Table 10 Distribution of Baker Act Examination Ref errals (ERs) for Race/Ethnicity 96 Table 11 Adolescents Who Received Baker Examination Referr als (ERs) 97 Odds Among Grade Levels for Children and Adolescents Odds Among Males and Females Who Received A Odds Among Race/Ethnicity Categories for Children and

PAGE 8

v Table 12 Odds Ratio Comparisons between Race/Ethnic ity Demographic Categories for Children and Adolesce nts Who Received Baker Act Examination Referrals (ERs ) 98 Table 13 Chi-Square Statistics for Comparison of Ba ker Act Examination Referral Proportions between Race/Eth nicity Demographic Categories 100 Table 14 Comparison of Proportion of African Americ an Children and Adolescents Who Received Baker Act Examinatio n Referrals (ERs) to Proportion of Enrollment Withi n Florida Public Schools 101 Table 15 Summary of Relevant Findings for Demograph ic Categories 102 Table 16 Baker Act Examination Referrals (ERs) 104 Table 17 Descriptive Statistics for the Dependent a nd Independent Variables 106 Table 18 Correlation Matrix for Dependent and Indep endent Variables 108 Table 19 Summary of Multiple Regression Analysis fo r Selected Variables Predicting PerBA ( N = 67) 112 Table 20 Summary of Multiple Regression Analysis fo r Selected Variables Predicting PerBA ( N = 67) 114 Table 21 Summary of Multiple Regression Analysis fo r Selected Variables Predicting PerRep ( N = 67) 116 Table 22 Summary of Multiple Regression Analysis fo r Selected Variables Predicting PerRep ( N = 67) 118 Table 23 Summary of Relevant Findings for Multiple Regression Analyses 119

PAGE 9

vi List of Figures Figure 1. Odds of Receiving a Baker Act Examination Referral (ER) 90 Figure 2. Odds of Receiving a Baker Act Examination Referral (ER) Among Males and Females 94 Figure 3. Odds of Receiving a Baker Act Examinatio n Referral (ER) Among Race/ethnicity categories 99 Among Age/Grade Level

PAGE 10

vii School Related Variables Bradley S. Beam ABSTRACT The Baker Act is legislation that provides civil ri ghts for individuals suspected of having a mental illness and may be in need of invol untary psychiatric evaluation. Its intent is to prevent the indiscriminate placement o f individuals in residential treatment facilities and other restrictive placements. Referr als for involuntary psychiatric evaluation under Baker Act statutes have increased among child ren and adolescents in recent years, raising concerns related its use with this populati on. The purpose of this study was to explore the relationship between school based facto rs and the use of the Baker Act among children and adolescents. Results indicate that the use of the Baker Act is more prevalent receiving among the high school population, females Other/mixed children and adolescents, and White children and adolescents. Mu ltiple regression analyses indicated that school district variables (e.g., minority enro llment, graduation rates, out-of-school suspension, etc.) contributed to a statistically si gnificant proportion of the variance in Baker Act ER rates and repeat Baker Act ER rates am ong the 67 counties in the state of Florida. The percent of students in a district that graduated with a standard diploma was the variables that most consistently contribute to a unique proportion of the variance in Baker Act ER and repeat Baker Act ER rates. The fin dings from the study have Baker Act Examination Referrals Among Children and Adolescents: An Analysis of

PAGE 11

viii implications for the design of mental health and be havioral support systems for children and adolescents. Additional research is necessary t o more closely examine the relationship between demographics, school related v ariables, and the use of the Baker Act.

PAGE 12

1 Chapter One Introduction Incidents of school violence have brought the issue of school safety to the forefront of public debate in the past decade. Scho ol shootings like Columbine and Jonesboro sent shockwaves across the United States as Americans tried to comprehend how such events could unfold in one of our nations most valued and basic institutions. These incidents prompted educators and policy maker s alike to ask the question: Just how safe are our children when they are in school and w hat can be done to prevent future incidents of school violence (United States Secret Service and United States Department of Education, 2002). At the core of the issue are b eliefs about the proper environments to raise and teach children. From a practical standpoi nt, it is difficult to imagine any social or academic growth occurring in school environments where children feel unsafe and threatened. From a moral perspective, the thought o f children being denied access to the safest, most nurturing environments violates one of our most basic values as a society. So what happens when students engage in behavior t hat is disruptive to orderly school environments? Programs such as Safe and Drug Free Schools (U.S. Department of Education [USDOE], 2002) and reports like Early War ning, Timely Response (USDOE, 1998) have been developed to address growing concer ns about safety in schools. Additionally, law enforcement approaches have been implemented to confront disruptive behavior and drug use in schools. School security o fficers now roam the hallways in

PAGE 13

2 many schools. Random locker searchers, metal detect ors, school surveillance cameras, and other approaches have been utilized in a preven tative fashion (Skiba, 2000). Zero tolerance polices that increase the severity of con sequences for all disciplinary infractions also have become common in public schools across th e United States. A direct consequence of zero tolerance is the utilization of punitive disciplinary practices (e.g., suspension and expulsion) that often culminate in t he removal of disruptive students from school (Skiba, 2000). Exclusion from Schooling In the context of the current climate of zero toler ance, students who exhibit a continuous pattern of disruptive behavior are at ri sk for being excluded from schooling. These students sometimes are diagnosed with emotion al and behavioral disorders (EBD). Intensive behavioral and mental health services oft en are necessary to facilitate social development among the most troubled students. Howev er, students with EBD encounter significant barriers to obtaining these services, a nd consequently, represent an underserved population of youth (U.S. Department of Health and Human Services [USDHHS], 1999). Further, eligibility criteria for special education programs often restrict access to school based behavioral and ment al health services until students have experienced a protracted period of failure in schoo l (Morrison & D’Incau, 2000). For many students, unaddressed emotional and behavior p roblems escalate and are expressed as severe incidents of disruptive behavior that thr eaten the safety of the student as well as that of other students and school staff. It often b ecomes necessary to allocate even more intensive resources to address the needs of these s tudents, and these resources are unavailable in most schools (Wagner & Sumi, 2005). Consequently, exclusion becomes a

PAGE 14

3 likely response in environments without adequate su pport services, and a favored response among school personnel who are concerned a bout the quality of the learning environment. Types of exclusion. Exclusion from schooling can take many forms, rang ing from being sent to the office to involuntary psychi atric placement. Students who qualify for special education services for EBD are suspende d and expelled more often than students with other disabilities, and they are more likely than students with other disabilities to receive instruction in the most res trictive educational environments (USDOE; 2003; Wagner et al., 2005). Students with s evere EBD sometimes manifest behaviors that culminate in a referral to community based agencies that provide the most intensive mental health services. These placements can be voluntary (school or family placement) or involuntary. Involuntary evaluation a nd placement in hospitals, residential treatment facilities or special day schools remove children from schools, families, and communities. Consequently, most states have passed legislation to protect the rights of individuals suspected of having severe mental illne ss that necessitates involuntary evaluation and treatment. The Baker Act. In the state of Florida, The Baker Act (F.S. 394, Part I, 2005) provides due process and civil rights to indi viduals suspected of having severe mental illness who meet certain criteria for involu ntary placement and may be in need of emergency evaluation or treatment. The Florida Legi slature passed the Baker Act in 1971 in response to advocacy efforts targeted toward the indiscriminate admission of adults and children with severe mental illness into reside ntial treatment facilities. The Baker Act established a patients’ bill of rights and also pro hibited placements of persons with severe

PAGE 15

4 mental illness into jails, unless they had committe d criminal acts. Prior to the final passage of the Baker Act in 1971, placement in stat e residential treatment facilities was the primary treatment settings available to individ uals diagnosed with severe forms of mental illness (Florida’s Baker Act Website, 2002). In recent years, concerns have been raised about th e utilization of the Baker Act for children in the state of Florida. Recent resear ch (Christy, Stiles, & Shanmugam, 2003; Florida Senate, 2005) indicates that Baker Act exam ination referrals (ERs) for children have increased across the past several years. Furth er, Baker Act ERs are less likely in the summer than during the months when students are in attendance, suggesting that law enforcement personnel and mental health professiona l may initiate a meaningful proportion of Baker Act ERs via the public schools. It has been suggested that schools may be using Baker Act examination referrals to rem ove disruptive students from school environments instead of addressing these problems t hrough the use of school based behavioral and mental health support systems (Flori da Senate, 2005). If this is the case, then the Baker Act can be viewed as a crisis orient ed, reactive approach to addressing emotional and behavioral problems in schools. Baker Act examination referrals may be one avenue for accessing intensive resources that a re unavailable in most school settings. Additionally, zero tolerance policies related to th e use of weapons or substance abuse may impact the beliefs and practices of school pers onnel. Schools personnel that utilize Baker Act examination referrals may embrace a cultu re of exclusion, and view removal from the school environment as an appropriate respo nse to disruptive behavior that is manifested in schools (Florida Senate, 2005).

PAGE 16

5 Accountability and Reform The growing concerns over school safety and the re liance on educational practices that exclude students from schooling have occurred at a time when accountability for student outcomes has become the driving force behin d the school reform movements in the United States. These efforts are not new, and i n fact, have been supported by various reform advocates for several decades. In 1983, the United States Department of Education published “A Nation at Risk” (National Co mmission on Excellence in Education, 1983), which called attention to the sta tus of public education. The report described a disturbing decline in educational perfo rmance of students in the United States, and warned that American prosperity was in jeopardy due to inadequate educational practices: “Our nation is at risk. Our once unchallenged pree minence in commerce, industry, science, and technological innovation is being ove rtaken by competitors throughout the world. This report is concerned wit h only one of the many causes and dimensions of the problem, but it is the one t hat undergirds American prosperity, security, and civility. We report to t he American people that while we can take justifiable pride in what our schools and colleges have historically accomplished and contributed to the United States and the well-being of its people, the educational foundations of our society are presently being eroded by a rising tide of mediocrity that threatens our very future as a Nation and a people. What was unimaginable a generation ago has begun t o occur—others are matching and surpassing our educational attainment s.” (p.1)

PAGE 17

6 Twenty-three years later, similar concerns regardi ng the overall effectiveness of public schools in the United States remain at the f orefront of efforts to improve public education. A renewed focus on student outcomes has ushered in a new wave of school reform efforts, and accountability has become the c ornerstone of these educational policies. The stakes have never been higher for our nation’s schools. The No Child Left Behind Act (NCLB) of 2002 (No Child Left Behind Act 2002) was developed by a bipartisan effort in the United States congress to address underachievement in schools. Under No Child Left Behind, schools must demonstrat e that all students meet annual objectives and benchmarks for achievement by 2014. Schools that fail to make Adequate Yearly Progress (AYP) are identified, and correctiv e actions must be undertaken to improve student achievement. Students with EBD. The accountability movement presents an interesting conundrum for educators in an era of zero tolerance On the one hand, schools personnel are under intense pressure to maintain order and di scipline to facilitate optimal student outcomes. On the other hand, educators must assure that even the most difficult to teach are provided with effective, evidence based instruc tion, and to do this, students with EBD must be in school and academically engaged. Reform efforts intended to improve the effectiveness of public schools must address the ne eds of students with EBD or a significant number of students will be left behind. Students with disabilities now are included as one of the eight categories of students that must achieve Adequate Yearly Progress (AYP) to ensure that a school or district is not identified as “in need of improvement.” Consequently, schools will be held ac countable for the educational

PAGE 18

7 outcomes obtained by students with disabilities, in cluding those with EBD (USDOE, 2002). Educators face substantial challenges in meeting th e needs of this population. Students with EBD are at an increased risk for a my riad of negative academic outcomes throughout childhood (Hinshaw, 1992; Wagner, Kutash Duchnowski, Epstein, & Sumi, 2005; Willcutt & Pennington, 2000). Many students w ith EBD have reading and math scores that fall in the bottom quartile of performa nce. Consequently, 22% of elementary/middle school children with EBD and 37.7 % of secondary children with EBD have been retained at least once (Wagner, Kutash, D uchnowski, Epstein, & Sumi, 2005). Due to negative social and academic outcomes, child ren with EBD are more likely than students in other disability categories to drop out of school. According to the U.S. Department of Education (2001), 41.9% of students w ho qualify for special education services under the Emotional Disturbance category r eceived a standard diploma during the 1998-1999 school year. In comparison, 63.3% of students with Specific Learning Disabilities (SLD) graduated with a standard diplom a during the 1998-1999 school year. These negative experiences continue across childhoo d, adolescence, and into adulthood. Students who qualify under the Emotional Disturbanc e category are less likely than students in other disabilities categories, with the exception of Mental Retardation, to enroll in postsecondary education programs. Further they are more likely than students in other disability categories to be terminated from t heir jobs and to be arrested (Wagner, Newman, Cameto, & Levine, 2005).

PAGE 19

8 Developmental Models and EBD Developmental models can be utilized to understand the emergence of EBD across childhood and the social problems encountere d across the lifespan. According to developmental models, children develop social compe tencies as they interact with their environment. In educational environments, successfu l adaptation is dependent on the emergence of social competencies associated with th e attainment of academic skills. These competencies facilitate task engagement acros s a wide range of educational environments (Sroufe, 1997). Social skill deficits are likely to preclude the attainment of academic skills and children with EBD often engage in inappropriate behaviors maintained by classroom contingencies that allow th em to escape and avoid difficult academic tasks and situations. These interactions u nintentionally reinforce maladaptive behavior and lessen opportunities to develop prosoc ial and academic skills. This interaction between academic and social competence often repeats itself in a reciprocal pattern across childhood. The nature of the recipro cal relationship between academic and social competence demonstrates an important tenet o f development models: Failure in one domain of competence rapidly accelerates the ri sk of failure in other areas of competence (Sroufe, 1997). Prevention and intervention. Developmental models have important implications for the design of effective educational environment s that serve students with EBD. Prevention and intervention efforts require a timel y response to early signs of EBD. Adequate responses involve the creation of systems that enhance competence early in the developmental course of emotional disturbance (Mast en & Curtis, 2003). In the absence of these support systems, children encounter a wide r range of social problems as

PAGE 20

9 environmental expectations require an increasingly sophisticated repertoire of social and academic competencies. Children who fail to obtain a competency at one stage of development will encounter more pronounced emotiona l and behavior problems as the gap between their competence and environmental expe ctations widens (Sroufe, 1997). Support services that are more costly and resource intensive are reserved for children who do not respond to prevention programs and requi re individualized levels of support (Sugai, Sprague, Horner, & Walker, 2000; USDOE, 200 3). Most importantly, schools must provide support services that maximize academi c engagement or children with EBD will fail to acquire the skills needed to meet dist rict and state benchmarks. Rationale for This Study The increase in the number of Baker Act ERs among s chool aged children is problematic for two reasons. First, exclusion from schooling of any kind, whether it is through the use of suspension or referral for invol untary psychiatric treatment, is problematic because it removes students from academ ic instruction. The consequence is reduced opportunities for exposure to the core or r emedial curriculum within a district. Second, exclusion may be symptomatic of an educatio nal system that is unable to provide effective behavioral and mental health services to students with severe EBD. The implications associated with the failure to provide effective academic instruction and behavioral support services are of a heightened sig nificance in an era of accountability and reform. Currently, no explanation has been suggested for th e rapid increase in the use of the Baker Act with children and youth. In addition, no studies have investigated the relationship between the policies and practices tha t guide school-based decision-making

PAGE 21

10 regarding the delivery of mental health services an d the use of the Baker Act by school personnel. It is important that educators, communit y based clinicians, policy makers, and child advocates achieve a greater understanding of factors that result in the use or non-use of the Baker Act with school children and youth. In addition, it is important to know what happens to children and youth who receive Baker Act ERs following their return to schooling. These efforts will assist with the devel opment of policies intended to provide children with access to the most effective behavior and mental health support systems in the least restrictive environments. There is little information, other than the number of children who have received Baker Act ERs and the demographics of those childre n, to understand the conditions under which schools are likely to use this placemen t option. However, research has addressed the utilization of other school based pra ctices that culminate in exclusion. Demographic characteristics such as gender, race, a nd socioeconomic status are associated with higher rates of exclusionary discip linary practices (Costenbader & Markson, 1998; Florida Department of Education, 200 5; Kleiner, Porch, & Farris, 2005; McFadden, Marsh, Price, & Hwang,1992; National Cent er for Education Statistics (NCES), 2004; Raffaele-Mendez, Knoff, & Ferron, 200 2; Skiba, Peterson, & Williams, 1997; Skiba, Michael, Nardo, & Peterson, 2002;). St udents who receive special education services also are at an increased risk of disciplin ary exclusion (Wagner, Kutash, Duchnowski, Epstein, & Sumi, 2005). Further, school s that embrace behavior support systems as a method for addressing disruptive behav ior often have lower rates of disciplinary exclusion (Raffaele-Mendez et al., 200 2; Skiba et al., 2002). Research also indicates that placement decisions are influenced b y the availability of adequate resources

PAGE 22

11 within educational and community based setting (Bic kman, Foster, & Lambert, 1996; Hendrickson, Smith, & Frank; 1998; Rock, Rosenberg, & Cannan, 1994; Romansky, Lyons, Lehner, & West, 2003). It may be possible to apply the results of studies from other types of exclusion to answer questions about the utilization of Baker Act referrals among children. Children and adolescents may be removed from schooling from a relatively short period of time to longer periods depending on a number of factors ass ociated with Baker Act ERs. Regardless of the time period, the Baker Act involv es a process that culminates in the removal of children from mainstream environments an d thus represents a major form of exclusion. The over utilization of Baker Act referr als may be symptomatic of a larger systemic problem associated with the availability o f effective behavior and mental health support systems in schools, and conversely, the ove r reliance on exclusion as a method to remove disruptive students from educational environ ments. Purpose of the Study The purpose of the present study is to explore the relationship between schoolbased factors and frequency of Baker Act referrals among school age children in the state of Florida. The following research questions will b e addressed by this study: 1. What is the distribution of demographic variables ( e.g., age/grade level, gender, and race/ethnicity) for children who receiv ed Baker Act examination referrals? 2. Do African American children receive Baker Act exam ination referrals at rates that are disproportionate to their total stud ent enrollment within Florida public schools?

PAGE 23

12 3. What percentage of Baker Act examination referrals are repeat Baker Act examination referrals? 4. What percentage of children received more than one Baker Act examination referral? 5. What is the relationship between the size of the sc hool district, percent of students in a district who belong to an ethnic mino rity category, the percent of African American, the percent of students in a dist rict who received free and reduced lunch, the percent of students in a distric t who receive special education services, the percent of students in a di strict who receive special education services under the EH/SED category, the p ercent of students in a district who graduate with a standard diploma, the percent of students in a district who are retained, the percent of students in a district who obtained a Level 3 or higher on the FCAT, and total Baker Act examination referrals and repeat Baker Act examination referrals, school dist rict use of suspension and expulsion, and the ratio of mental health workers t o students in a school district, and total Baker Act examination referrals and repeat Baker Act examination referrals? Research Hypotheses 1. There will be significant differences between the p roportion of students in each demographic category who received a Baker Act examination referral and the proportion of all enrolled students in each demographic category in the Florida public schools.

PAGE 24

13 2. There will be no difference between the proportions of African American students who received a Baker Act examination refer ral and the proportion of enrolled students in the Florida public schools. 3. There will be no relationship between the size of t he school district, percent of students in a district who belong to an ethnic mino rity category, the percent of African American students in a district, the percen t of students in a district who receive free and reduced lunch, the percent of students in a district who receive special education services, the percent of students in a district who receive special education services under the EH/SED category, the percent of students in a district who graduate with a standard diploma, the percent of students in a district who are retained, the percen t of students in a district who obtained a Level 3 or higher on the FCAT and total Baker Act examination referrals and repeat Baker Act examination referral s, school district use of suspension and expulsion, the ratio of mental healt h workers to students in a school district, and total Baker Act examination re ferrals and repeat Baker Act examination referrals?

PAGE 25

14 Chapter Two Literature Review The purpose of this chapter is, first, to review r esearch that investigates the range of procedures (e.g., suspension, expulsion, involu ntary psychiatric placements) used by public schools to address the behaviors of students with severe emotional and behavioral difficulties. Second, a program and data-based over view of Florida’s Baker Act will be presented. Third, models of psychopathology will be explored to provide possible explanations for severe behavior in students and to evaluate the appropriateness of school-based procedures designed to address those b ehaviors. Next, literature related to the availability of support services for children w ith severe emotional and behavioral disorders will be presented. Finally, literature th at explores the variables associated with the decision to choose exclusionary settings for tr eatment of children with severe emotional and behavioral disorders will be reviewed School Exclusion Public schools employ a wide range of procedures to address behaviors that are harmful to other students, staff, or to the student Most of the procedures used involve some level of exclusion from the general education setting. Exclusion can be conceptualized as occurring along a continuum from least restrictive to most restrictive placement. For purposes of this literature review, time out and office referrals are viewed as the least restrictive form of exclusion because removal from typical environments

PAGE 26

15 tends to be relatively brief. Special education pro grams that remove children from mainstream settings also are considered along a con tinuum of least to most restrictive placements. Suspension and alternative education re present the next most restrictive form of exclusion. Finally, involuntary placement in psy chiatric facilities represents a very restrictive placement because children are removed from school and community. Perhaps suspension and expulsion represent the most severe form of exclusion. In this case, the student is removed both from educational settings a nd from any school or community supported intervention program. Rates of disciplinary exclusion. As part of a broader project investigating the use of office referrals to inform school wide disci pline planning, Sugai and colleagues (2000) examined the prevalence of office referrals among 11 elementary schools and 9 middle/junior high schools across seven school dist ricts. Elementary schools averaged 567 disciplinary referrals per year with a mean of 0.5 discipline referrals per student per year, and a mean of 1.7 disciplinary referrals per school day. On average, 21% of students at the elementary school level received on e or more disciplinary referrals. Middle schools averaged 635 students per year with a mean 2.4 disciplinary referrals per student. Middle schools averaged 8.6 disciplinary r eferrals per school day. An average of 47.6% of middle school students received at least o ne office referral. These data suggest that office referrals are more likely during middle school. According to DeVoe, Peter, Noonan, Snyder, & Baum ( 2005), 54% of public schools used a serious disciplinary action such as a suspension or expulsion against at least one student during the 1999-2000 school year. Eighty three percent of office referrals led to suspensions lasting 5 or more days 11% led to removals with no services

PAGE 27

16 (i.e., expulsions), and 7% were transferred to spec ialized schools. During the 1999-2000 school year, 6.6% of enrolled students in the Unite d States were suspended and 0.21% of enrolled students were expelled (National Center fo r Educational Statistics [NCES], 2005). According to the Florida Department of Education (2 005), 1.84% of enrolled elementary school students, 16.72% of middle school students, and 15.98% of high school students received in-school suspensions duri ng the 2004-2005 school year. During the 2004-2005 school year, 2.99% of elementary scho ol students, 14.07% of middle school students, and 11.77% of high school students received an out-of-school suspension. Raffaele-Mendez (2000) examined the out -of-school suspension (OSS) percentages and rates for a large urban school dist rict in Florida. During the 1996-1997 school year, there were a total of 33,620 out-of-sc hool suspensions (duplicated count) and 16,204 unduplicated out-of-school suspensions. Data were reported as the percentage of students who received at least one suspension and t he rate of students who were suspended. The OSS percentage represents the number of students who were suspended at least once and is reported here as the mean perc entage across schools. The OSS percentage for elementary schools was 3.3% and the rate was 5.6 per 100 students. For middle schools, the OSS percentage was 23.68% and t he rate was 52.3 per 100 students. At the high school level, the OSS percentage was 20 .7% and the rate was 39.2 per 100 students. These data indicate that middle school st udents are more likely than elementary and high school students to receive an in-school an d out-of-school suspension. Students often are transferred to alternative educa tional programs following the term of expulsion from a regular public school. Alt ernative educational programs are

PAGE 28

17 designed to serve students who are at risk for educ ational failure due to poor grades, truancy, disruptive behavior, suspensions, pregnanc y, or similar factors associated with early withdrawal from school (Kleiner, Porch, & Far ris, 2002). Kleiner et al. (2002) indicated that 39% of public school districts in th e United States administered at least on alternative school or program for at-risk students during the 2000-01 school year. Fiftynine percent of alternative school programs were lo cated in a facility separate from a regular school, 4% were housed in juvenile detentio n centers, 3% were in community centers, and 1% were charter schools. As of October 1, 2000, 1.3% of all public school students were enrolled in alternative educational p rograms in the United States. These data indicate that although a substantial proportio n of school districts utilize alternative education programs, only a small percentage of stud ents within these districts actually attend alternative education programs. Behaviors that lead to disciplinary exclusion. Students are referred to the principal’s office for a wide range of behaviors in cluding: defiance, disobedience, physical contact and fighting, insubordination, and verbal abuse (Skiba, Peterson, & Williams, 1997). In addition, non-interpersonal beh aviors such as excessive tardiness and absences, leaving the classroom or building without permission, and failure to complete written work result in both inand out-of-school s uspensions (Costenbader & Markson, 1998; Raffaele-Mendez, Knoff, & Ferron, 2002; Skiba et al., 1997). More serious disciplinary infractions, such as possession of a w eapon or possession of drugs and alcohol account, for a higher percentage of out-ofschool suspensions than other less serious violations (Costenbader & Markson, 1998; De Voe et al., 2005). DeVoe et al. (2005) reported that physical attacks and fights ac counted for 35% of all suspensions and

PAGE 29

18 expulsions in the United States during the 1999-200 0 school year. Insubordination accounted for 18% of all serious disciplinary actio ns, threat or intimidation accounted for 22%, possession or use of alcohol or illegal drugs accounted for 20%, and possession of a weapon other than a firearm accounted for 19% of al l serious disciplinary actions. Use of a weapon other than a firearm accounted for 5% of s erious disciplinary actions, possession of a firearm/explosive device accounted for 4%, and use of a firearm/explosive device accounted for 2% of all se rious disciplinary actions. Other nonacademic infractions accounted for 14% of all se rious disciplinary actions (DeVoe et al., 2005). It should be noted that these are aggre gated data, and do not specify the disciplinary action taken (i.e., suspension vs. exp ulsion). Thus, it seems that the restrictiveness of the disciplinary action is relat ed to the seriousness of the infraction. Disciplinary actions that culminate in the most res trictive disciplinary actions are reserved for more serious offenses that threaten th e safety and well being of students and school staff. However, minor offenses such as insub ordination also can lead to removal from schooling if they occur on a continuous basis across time. According to Kleiner et al. (2002), approximately o ne-half of all districts with alternative programs report that the following disc iplinary incidents were the sole reason for transfers: possession of illegal substance, dis tribution or use of drugs (52%), physical attacks or fights (52%), chronic truancy (51%), use of a weapon other than a firearm (50%), continual academic failure (50%), disruptive verbal behavior (45%), and possession or use of a firearm (44%). Reported arre sts or involvement in the juvenile justice system were cited as a sufficient reason fo r transfer to an alternative school in 38% of districts surveyed. Mental health needs were least likely to be cited as the sole

PAGE 30

19 reason for transfer. The research did not address t he relationship between externalizing behaviors and underlying mental health issues in th e students placed in alternative programs. The distinction between externalizing beh aviors and underlying mental health issues may be a false dichotomy and may obscure a t horough examination of the mental health needs of students placed in alternative educ ation programs. A number of conclusions can be drawn from the lite rature cited. It appears that schools utilize office referrals, suspensions, out of school suspensions, expulsion, and transfers to alternative schools in response to a w ide range of student behavior. Similar behaviors can lead to less exclusionary disciplinar y practices such as office referrals or they can be followed by more serious disciplinary a ctions such as suspension and expulsion. For example, noncompliance can lead to a n office referral, and in some cases be followed by internal and external suspension. Mo re serious disciplinary offense such as physical fighting and weapons possession tend to lead to greater levels of exclusion. Thus, there appears to be a relationship between th e severity of student behavior and the degree to which schools utilize the most exclusiona ry practices. However, the relationship is imperfect, and less severe behavior s often are followed by disciplinary actions that exclude students from school. Student characteristics and disciplinary exclusion Certain demographic characteristics (gender, race, and socioeconomic st atus) are associated with greater use of exclusionary disciplinary practices. The literatur e reviewed suggests that males are more likely to be excluded than females (Costenbader & M arkson, 1998; Florida Department of Education, 2005; Raffaele-Mendez, et al., 2002; Skiba et al., 1997). The National Center for Education Statistics (2003) reported tha t 9.2% of enrolled males and 3.9% of

PAGE 31

20 enrolled females were suspended during the 1999-200 0 school year. Additionally, 0.31% of enrolled males and 0.10% of enrolled females wer e expelled during the 1999-2000 school year (NCES, 2004). Skiba, Michael, Nardo, an d Peterson (2002) found that although males represented 51.8% of enrolled studen ts in a large urban school district, they accounted for 67.2% of suspensions and 83.7% o f expulsions. Females were underrepresented on all measures of school discipli ne. According to the Florida Department of Education (2005), 2.76% of males and 0.86% of females received an inschool suspension at the elementary school level du ring the 2004-2005 school year. At the elementary school level, 4.60% of males and 1.2 7% of females received an out-ofschool suspension. These trends were found across the middle school and high school levels (Florida Department of Education, 2005). Ma les are about 3.2 times more likely to receive an internal suspension at the elementary sc hool level, 1.74 times more likely at the middle school level, and 1.48 times more likely to be suspended in high school. For out of school suspensions, males are 3.62 times mor e likely to receive an out-of-school suspension in elementary school, 1.94 times more li kely to receive an out-of-school suspension in middle school, and 1.72 times more li kely at the high school level. Thus, although rates of disciplinary exclusion increase i n middle school among both males and females, the disparity between the genders is great est at the elementary school level. Race and socioeconomic status are associated with r ates of discipline practices that exclude students from school (Costenbader & Ma rkson, 1994, 1998; McFadden, Marsh, Price, & Hwang, 1992; Raffaele-Mendez, 2003; Skiba et al., 1997). Data from the Florida Department of Education (2005) demonstrates that 7.20% of enrolled African American elementary school students, 25.84% of Afri can American middle school

PAGE 32

21 students, and 19.74% of African American high schoo l students received an out-ofschool suspension during the 2004-2005 school year. In comparison, 1.71% of White elementary school students, 10.07% of White middle school students, and 9.24% of White high school students received at out-of-schoo l suspension (Florida Department of Education, 2005). These data indicate that compared to White students, African Americans are 2.34 times more likely to receive an internal suspension at the elementary school level, 2.02 times more likely in middle scho ol, and 1.76 times more likely in high school to receive an internal suspension. Compared to White students, African Americans are 4.2 times more likely to receive an out-of-scho ol suspension at the elementary school level, 2.56 times more likely at the middle school level, and 2.13 times more likely at the high school level. Skiba et al. (2002) found that African American stu dents received office referrals, suspensions, and expulsion at rates disproportionat e to their total enrollment while Caucasian students were underrepresented. African A merican students accounted for 56% of all middle schools students in a large, urba n school district, but they accounted for 66.1% of office referrals, 68.5% of suspensions and 80.9% of expulsions. Significant racial disparities remained after controlling for s ocioeconomic status (Skiba et al., 2002). Raffaele-Mendez et al. (2002) found that schools se rving a higher percentage of minority students from low socioeconomic backgrounds were mo re likely to have higher out-ofschool suspension (OSS) rates. The association betw een race, socioeconomic status, and OSS rates was strongest at the elementary school le vel. However, there were schools that served higher percentages of African American stude nts from low socioeconomic

PAGE 33

22 backgrounds and had lower suspension rates, suggest ing that other variables may moderate this relationship. Race and socioeconomic status also are associated w ith presence of alternative school programs in school districts (Kleiner et al. 2005). School districts with less than 5% minority enrollment were less likely to have alt ernative school programs than districts with 6-20%, 21-50%, and more than 50% minority scho ol enrollment during the 20002001 school year. Additionally, school districts wi th 10% or less of students at or below the poverty level were less likely to have alternat ive school programs than districts with 11-20%, and more than 20% of students at or below t he poverty level. Academic achievement and exclusion. Christle, Jolivette, and Nelson (2005) examined the relationship between school level vari ables and suspension rates among middle schools in Kentucky. Multivariate analysis i ndicated that schools with higher suspension rates had significantly higher dropout r ates, board of education and law violations, percentages of students from low socioe conomic backgrounds, and per pupil expenditures compared to middle schools with lower suspension rates. Schools with lower suspension rates had higher mean scores on st andardized achievement tests, greater attendance rates, and a higher percentage of Caucas ian students. Retention rate, enrollment, average years of teaching experience fo r the staff, student/teacher ratio, and the percentage of enrolled boys were not related to suspension rates. In a related study, these authors found that schools with higher drop o ut rates had higher percentage of students from low socioeconomic backgrounds, retent ion rate, suspension rate, and board of education violation rate compared to schools wit h low drop out rates. Rodney, Crafter, Rodney, and Mupier (1999) found that the number of suspensions received by students

PAGE 34

23 was the strongest predictor of grade retention amon g African American male teenagers. These results indicate that school level academic p erformance variables such as mean scores on norm referenced achievement tests and gra duation rates are associated with disciplinary exclusion. There also appears to be an association between grade retention and suspension. Special education and exclusion. The Least Restrictive Environment (LRE) clause of the Individual’s with Disabilities Educat ion Improvement Act (IDEIA) (U.S. Department of Education [USDOE], 2004) mandates tha t students with disabilities receive academic instruction in environments with n on-disabled peers to the maximum extent that is appropriate given their individual n eeds. Students with disabilities should receive academic instruction in classes or school s eparate from non-disabled peers only if available support services are insufficient to prov ide an appropriate education. The clause states: …to the maximum extent appropriate, children with d isabilities, including children in public or private institutions or other care facilities, are educated with children who are not disabled, and special classes, separate schooling, or other removal of children with disabilities from the regu lar educational environment occurs when the nature and severity of the disabili ty of the child is such that education in regular classes with the use of supple mentary aids and services cannot be achieved satisfactorily (Section 612, (a) (A) IDEA, 2004). Students who receive services under the Emotional D isturbance (ED) category represent 0.85% of the school age population, indic ating that a small percentage of students receive special education services under t his category (Wagner, Kutash,

PAGE 35

24 Duchnowski, Epstein, & Sumi, 2005). According to th e Twenty-Fifth Annual Report to Congress on the Implementation of IDEA (U.S. Depart ment of Education, 2003), 8.1% of students who are eligible for special education rec eived services under the ED category. Among these students, 16% of students ages 6-12 and 22% of student ages 13-17 were included in the regular education classroom 100% of the time during the 1999-2000 school year. Among all students with ED, 31.8% serv ed more than 60% of their school day outside the regular classroom, and 18.1% of stu dents with ED were served in a separate facility (USDOE, 2005). Thus, a substantia l proportion of students with ED receive academic instruction in environments separa te from their non-disabled peers. Some of these placements (i.e., separate facilities ) represent the most restrictive educational environments. In addition to exclusion through placement in restr ictive educational settings, students with EBD are subjected to disciplinary met hods that lead to exclusion. Punitive disciplinary methods such as in and out-of-school s uspension may be a more severe method of exclusion because behavior support system s are not provided while these students are absent from educational settings. Wagn er, Kutash, Duchnowski, Epstein, & Sumi, 2005 (2005) examined disciplinary exclusion r ates among students receiving special education services for an Emotional Disturb ance (ED) using data from the Special Education Elementary Longitudinal Study and the Nat ional Longitudinal Transition Study-2. They found that among elementary school st udents, 47.7% of the ED population had been suspended or expelled compared to 11.7% of students with other disabilities. Further, 72.9% of secondary students with ED were s uspended or expelled compared to 27.6% of students with other disabilities. These fi ndings parallel the overall trend found

PAGE 36

25 in rates of disciplinary exclusion, and indicate th at secondary students with ED are more likely than elementary school students with ED to r eceive a suspension or expulsion. Skiba et al. (1997) also found that students who re ceived special education services under the Emotionally Handicapped (EH) label were more li kely than students in general and special education to receive both an office referra l and a suspension. Students classified as learning disabled or mildly mentally retarded al so were more likely than students in general education to receive a suspension. These fi ndings suggest that enrollment in special education, especially in programs for stude nts with emotional/behavioral disorders, is a risk factor for receiving a discipl inary referral. Cumulative effects of demographic variables. The literature reviewed thus far suggests that race, socioeconomic status, and g ender are associated with disciplinary practices that exclude student from school. Researc h also suggests that as membership in one or more of these demographic categories increas e, so to does the odds of receiving a disciplinary referral that leads to removal from sc hool. Raffaele-Mendez (2003) conducted a longitudinal study based on data from a cohort of 8, 268 students who entered kindergarten in 1989. During the 1995-1996 school year, there was an overrepresentation of black males who were enrolled in special education, receiving free or reduced priced lunch, and had experienced an out -of-school suspensions (OSS). Students who fit in this demographic group represen ted less than 5% of the total student population. However, they accounted for 24% of all students suspended three to five times, 34% of all students suspended six to eight t imes, 48% of all students suspended nine to eleven times, and 56% of all students suspe nded 12-14 times. In sixth grade, 66.27% black males receiving free or reduced priced lunch and enrolled in special

PAGE 37

26 education were suspended at least once. In comparis on, 44.12% of all white males receiving free and reduced priced lunch and special education services were suspended at least once in sixth grade. Among black males who pa id for their lunch and received special education services, 13.60% were suspended c ompared to 54.29% of white males who paid for their lunch and were enrolled in speci al education. Black females also were more likely to be suspended regardless of whether o r not they received special education services. Thus, the risk of receiving an out-of-sch ool suspension was greatest among African American males enrolled in special educatio n and receiving free and reduced priced lunch. In summary, office referrals, time out, special edu cation placement in settings separate from the general education classroom, susp ensions and expulsions represent educational practices that exclude students from ma instream environments. The risk of disciplinary exclusion is greatest among males (Cos tenbader & Markson, 1998; Florida Department of Education, 2005; NCES, 2003; Raffaele -Mendez et al., 2002; Skiba et al., 1997;), African American students ( Florida Departm ent of Education, 2005; Kleiner et al., 2005; Raffaele-Mendez et al., 2002; Skiba et a l., 2002) and students from low socioeconomic backgrounds (Kleiner et al., 2002; Mc Fadden, et al., 1992; RaffaeleMendez et al., 2002; Skiba et al., 2002). Student w ho receive special education services under the Emotional Disturbance (ED) category also are at an increased risk of exclusion due to placements in the most restrictive education al environments (U.S. Department of Education, 2005) and disciplinary exclusion (Wagner Kutash, Duchnowski, Epstein, & Sumi, 2005). Further, the risk of being excluded in creases as the number of demographic risk factors increase within a population. African American males who receive free and

PAGE 38

27 reduced priced lunch and special education services appear to be at the greatest risk of disciplinary exclusion (Raffaele-Mendez et al., 200 3). Baker Act Examination Referrals. Children often engage in high intensity behaviors t hat threaten the safety and well being of themselves and others in their environment Schools may react to these incidents with a disciplinary referral for a suspension, expu lsion, and placement in an alternative education program. In many cases, these behaviors a re so severe that schools may choose to initiate an involuntary psychiatric examination if there also is evidence of a mental illness. In the state of Florida, the Baker Act (F. S. 394, Part I, 2005) provides due process rights for individuals suspected of having a mental illness and in need of an involuntary examination. Individuals with severe mental illness must meet certain criteria for involuntary examination and placement in a resident ial treatment facility. Several revisions to the Baker Act have had importa nt implications for children. In 1982, the Florida legislature directed the Depar tment of Health and Rehabilitative Services to draft a plan for the elimination of pla cement in state hospitals by developing community alternatives to treatment. The legislatur e also added language that encouraged exploration of less intrusive treatment options for children. In 1998, the Comprehensive Child and Adolescent Mental Health Services Act in Part III of Chapter 394 led to language specific to the creation of a system of ca re for children and adolescents with emotional and behavioral disorders. It directed the Department of Children and Family Services to establish a continuum of treatment opti ons for children and adolescents. Crisis stabilization, including Baker Act ERs was i ncluded as one option available to children. An amendment to the Baker Act in 2000 pro hibited admission of children and

PAGE 39

28 adolescents to state owned or operated mental healt h treatment facilities under any circumstance. However, the amendment authorized adm ission to crisis stabilization units, residential treatment facilities, or a licensed hos pital pursuant to an involuntary or voluntary admission process. Additionally, the amen dment required treatment in the least restrictive environment (Florida Senate, 2005). Baker Act ERs occur in receiving facilities that re ceiving funding from private and public sources. Receiving facilities that recei ve public funding are called crisis stabilization units. A “crisis stabilization unit” as defined by Part IV of Chapter 394, F.S., the Community Substance Abuse and Mental Heal th Services Act, is “a program that provides an alternative to inpatient hospitali zation and that provides brief, intensive services 24 hours a day, 7 days a week, for mentall y ill individuals who are in an acutely disturbed state.” In Florida, ten facilities are sp ecifically designated as children’s units. However, children and adolescents can be taken to a ny available receiving facility that can perform the evaluation procedures associated wi th Baker Act ERs. Children are taken to the nearest receiving facility if a referral is initiated in a county without a designated receiving facility. Crisis stabilization units were developed to provide short term acute mental health care and identification of the most a ppropriate and least restrictive community setting available (Florida Senate, 2005). Prevalence of Baker Act referrals. The Baker Act Reporting Center at the Louis de la Parte Florida Mental Health Institute i s responsible for maintaining a data base on Baker Act examination referrals (ERs). In 2 002, the center received 105,062 Baker Act Initiation Forms. The report indicated th at 16% of Baker Act ERs were for individuals 17 years and younger and 84% were for a dults. Additional analysis indicated

PAGE 40

29 that 21% of children experienced more than one exam ination between 2002 through most of 2004. Children who had one or more Baker Act ERs accounted for more than 44% of all examinations (Christy, 2005). Further, there wa s a 34% increase in Baker Act ERs for children and adolescents aged 4-17 from 2001-2005 a nd only a 7% increase in the population of children and adolescents in that age range (Christy, 2005). These findings indicate that the use of the Baker Act has increase d substantially among children and adolescents in recent years and that the increase c annot be attributed to in growth in the population. Males accounted for 50.10% of Baker Act ERs between the time period of 19992002 in a large urban county in Florida. During the same time period, the average age of receiving an initial Baker Act ER was 13.60 years o f age for those children and adolescents under 18 years of age. In 88% of report ed cases, the justification for a Baker Act ER was related to the finding that the individu al would likely cause serious bodily harm to self or others in the future. Evidence sugg esting that the individual would likely suffer from neglect resulting in real and present t hreat of substantial harm accounted for 5.34% of Baker Act ERs among children. Harm and neg lect accounted for 4.72% of Baker Act examination referrals among children and evidence was missing from 1.56% of reported cases (Christy, Petrila, Hudacek, Hayne s, Wedekind, & Pulley, 2005). These data indicate that males and females are at an equa l risk for a Baker Act ER. Behaviors that threaten the physical safety of the individual or others within the individual’s environment are the most common reasons cited for a Baker Act ER. It is difficult to make a statement regarding the age that represents the greatest risk period for receiving a Baker Act ER. Although the average age of receiving a Baker Act ER is at the onset of

PAGE 41

30 adolescence, differences in the distribution of Bak er Act ERs across childhood and adolescents may present challenges to making a vali d conclusion regarding the greatest period of risk. Baker Act referrals and school-based behavior. Several concerns related to the process by which children are referred to Baker Act receiving facilities have arisen in recent years. First, the increase in the number of Baker Act ERs suggests that systems responsible for serving children are utilizing the most restrictive option in response to intense behavioral incident. These findings may ind icate that fewer options are available for serving children with severe EBD in less restri ctive environments. It has been argued that involuntary examinations have been utilized in appropriately with children and adolescents. Anecdotal reports suggest that childre n are receiving involuntary examinations without exploration of less intrusive options (Florida Senate, 2005). A second concern stems from evidence suggesting tha t schools may be utilizing Baker Act ERs in response to zero tolerance policie s, which often lead to other disciplinary strategies that remove children from s chool. Schools may be using Baker Act ERs to remove children with severe emotional and be havioral disorders from school settings rather than implementing less intrusive sc hool based services (Florida Senate, 2005). Christy et al. (2005) examined factors assoc iated with acute mental health care (Baker Act examination referrals) within a large co unty in Florida. Information was collected from statewide and county Medicaid servic e utilization claims, the Florida Department of Children and Families Integrated Data System, the Baker Act database, and the Child Welfare and Emergency Management Serv ices. Results indicated that

PAGE 42

31 Baker Act ERs for children were less likely during the summer months, suggesting that schools may be involved in the initiation of many B aker Act ERs. Christy (2005) found evidence suggesting that the a vailability of crisis stabilization units (CSU) within a region may be as sociated with repeat Baker Act ERs. The Sun Coast Region in Florida had 25% of children having two or more Baker Act examinations between 2002 and most of 2004. The reg ion also had the highest population of children of any district/region and had more chi ldren’s crisis stabilization units than any other district/region. The availability of cris is stabilization units may prompt schools to utilize this resource to address severe EBD. Thu s, crisis stabilization units may serve as a resource for schools faced with addressing the needs of students with severe EBD. Intensive behavior support services along with adeq uately trained personnel may be underutilized or unavailable in most school setting s, prompting a reliance on crisis stabilization units to address challenging behavior Schools are designed to address the needs of typica l students and those students with less intense emotional and behavioral problems When a student engages in a very intense behavioral incident, schools might initiate a Baker Act ER because it provides a mechanism by which students can access more intensi ve support services that are unavailable in most school settings. Additionally, the presence of beliefs that support the removal of disruptive students from school environm ents through any means possible may increase the likelihood of a Baker Act referral Models of Psychopathology Literature related to models of psychopathology wil l be presented prior to discussing the research base that has identified an association between the availability of

PAGE 43

32 mental health and behavioral support services and e xclusion from schooling and placement in the most restrictive environments. Exp lanations for disturbance are derived from models that specify the origin of human behavi or. Once a particular paradigm or theoretical orientation is adopted, incoming data a nd information are integrated into a framework that serves as a heuristic for assessment and intervention planning. Hypotheses about the etiology of severe childhood p sychopathology are developed during evaluation and treatment planning for childr en. The evaluation techniques and services provided to children are consistent with e xplanatory models of psychopathology, and these beliefs also guide the selection of the m ost appropriate interventions. Implicit in this process are underlying beliefs about the most appropriate environments and support services for children who manifest severe psychopat hology. Consequently, the adoption of a model of psychopathology can influence the deg ree to which support services are utilized within environments that provide education al and mental health services for children (Sroufe, 1997). Medical models. Traditionally, medical models have been emphasized as the primary explanatory model of severe psychopathology Medical models have a long tradition in the literature that has addressed the etiology of severe psychopathology. According to Sroufe (1997), the medical model of ps ychopathology emphasizes traits and conditions within the child. These traits are thoug ht to have an organic basis and represent pathological disturbance of the individua l. Human problems are grouped into taxonomic systems according to unique traits associ ated with membership in a category of disturbance. Characteristics common to each cate gory are analyzed to determine traits that discriminate groups from one another. These c ategories are given names and serve

PAGE 44

33 as systemic frameworks for conceptualizing the natu re of disturbance. Within these systems, disturbance and the names given to delinea te these categories are believed to be representative of enduring conditions and traits en dogenous to the individual. Following this line of reasoning, an organic basis for severe psychopathology is emphasized within a medical model (Sroufe, 1997). The organic pathological expression of disturbance can develop from genetic or environmental pathogens. Research and clinical trea tment is directed towards the identification of pathological symptoms and managem ent of the symptoms associated with disturbance. Consequently, assessment techniqu es are designed to identify symptoms that covary with a particular category of pathology and medical treatments often are emphasized to eliminate or reduce the imp act of neuropsychological deficits and excesses. Environmental manipulations also might be included in treatment. However, limited emphasis is directed towards enhancing the adaptational skills of the child. Similarly, modifications to environmental systems t hat facilitate adaptation are given less attention in research and treatment programs within a traditional medical model. Symptom management is emphasized during treatment o f mental illness within a medical model while early intervention and preventative app roaches are given less attention (Sroufe, 1997). Neurobiological theories of psychopathology are an example of an etiological explanation that can be subsumed under a medical mo del. Individual variations and differences in neurobiologically based processes ar e emphasized. These processes are influenced by evolutionary mechanisms, genetic infl uences, neurochemical characteristics, and the neuroanatomical makeup of the individual. The manifestation of

PAGE 45

34 severe psychopathology is influenced by dysfunction in one or several of these processes, which causes emotional and behavioral disturbance ( Mash & Dozois, 2003). Models that emphasize neuropsychological constructs have been u sed to explain various forms of childhood psychopathology. For example, Moffit (199 3) reviewed 47 studies that explored the neuropsychological constructs associat ed with Conduct Disorder. Deficits in language based skills and “executive” self-control functions were identified as two processes commonly impaired in antisocial children. Neuropsychological theories postulate that executive function deficits interfer e with a child’s ability to control his or her behavior. Consequently, the child has difficult y considering the future, long term consequences of present behavior. Childhood behavio ral disorders are influenced by deficits in executive functioning, which impair a c hild’s ability to adapt to social demands within his or her environment, setting the stage for the emergence of antisocial behavior. Developmental models. Developmental models of disturbance also have been developed to explain emotional and behavioral maladjustment. According to these models, the development of adaptive and maladaptive behaviors is best explained when conceptualized as an interaction between the indivi dual and the environmental context in which growth occurs. Endogenous neuropsychological differences are examined when attempting to explain the emergence of behavioral m aladjustment However, the endogenous characteristics of the individual and en vironmental influences are viewed as inseparable within a developmental model. Further, similar principles that govern normal development also govern the emergence of maladaptiv e behavior. Prior adjustment

PAGE 46

35 interacts with current environmental characteristic s in a reciprocal manner, influencing the manifestation of maladaptive behavior at each s tage of development (Sroufe, 1997). Psychopathology is reflected in a child’s failure t o meet developmental expectations. According to Mash and Dozois (2003), psychopathology is an expression of adaptational failure. Adaptational failure is descr ibed as “deviation from age-appropriate norms, an exaggeration of normal developmental tren ds, an interference in normal developmental progress, or a failure to obtain a sp ecific developmental function or mechanism” (2003, p. 22). Disturbance occurs when a child fails to develop a competency that facilitates the successful completi on of developmentally appropriate tasks. The child’s ability to meet age appropriate developmental expectations is diminished, and the emergence of maladaptive behavi or becomes more likely. Developmental expectations change across the lifesp an and across environments. Therefore, disturbance can be manifested in various ways depending on the expectations of the environment and the age of the child. Enviro nmental expectations influence which behaviors are judged to be deviant, leading to the perception of disturbance. Consequently, the manifestation of adaptational fai lure involves a certain level of social judgment from adults within a child’s environment. This judgment entails the perception that a child has failed to attain a developmental t ask necessary for successful environmental adaptation. The concept of developmental pathways is one of th e most important and influential aspects of developmental models. These pathways are a metaphor used to describe the relationship between manifestations of disturbance and various stages of the life span. When a child deviates from the normative developmental patterns associated

PAGE 47

36 with successful environmental adaptation, there is an increased probability of adaptational failure. Adaptational failure at any p oint in time places the child on a pathway that has the potential to lead to the manif estation of severe psychopathology across the lifespan. Different pathways can lead to the manifestation of similar patterns of disturbance. For example, the manifestation of soci al withdrawal and depression share similar features in terms of their expression. Howe ver, these conditions may have their origin in divergent developmental pathways such as alienation, anxiety or helplessness (Sroufe, 1997). The process by which divergent cond itions lead to similar expression of disturbance is known as equifinality (Mash & Dozois 2003). Further, similar pathways can place the child at an increased risk for variat ion in the expression of severe psychopathology. Children who begin on a similar pa thway of adaptational failure can manifest different patterns of disturbance across t he life span (Sroufe, 1997). This process is referred to as multifinality within the developm ental psychopathology framework (Mash & Dozois, 2003). Regardless of the pathway th at leads to disturbance, severe psychopathology is the result of successive deviati ons in normative developmental patterns over the course of childhood development. Each deviation is leads to a widening in the gap between the expectations of the environm ent and the skills necessary to achieve adaptation. Consequently, the risk of sever e psychopathology increases across childhood when skills deficits go unaddressed (Srou fe, 1997). Adaptational failure can be manifested across each stage of development, but its expression can vary based on contextual features of the environment. Developmental models have important implications for the design o f systems that are intended to enhance environmental adaptation. Early identificat ion and early intervention are

PAGE 48

37 emphasized with developmental frameworks because co ntextual features of the environment can prevent or reverse negative develop mental trajectories. Severe psychopathology may result when a child enters a ne gative developmental trajectory early in development, continues to experience succe ssive deviations in normative patterns of development, and fails to receive environmental supports that facilitate adaptation. The duration of a maladaptive pathway is associated wit h the responsiveness to environmental manipulations across development, and therefore, in terventions that interrupt negative pathways early in development have the greatest lik elihood of success (Sroufe, 1997). Traditional medical model explanations for severe childhood psychopathology are increasingly being integrated into developmental mo dels. The Moffit (1993) study presented in the preceding paragraphs identifies ne uropsychological deficits as a key variable affecting the emergence of antisocial beha vior in childhood. However, the model is integrated into a development framework whereby neuropsychological deficits set the stage for dysfunctional patterns of interactions wi th caregivers early in development. Consequently, the risk of failure in response to ad ult mediated developmental tasks increases as early patterns of relationships become dysfunctional. Environmental variables have the potential to exacerbate the affe cts of neuropsychological deficits. The emergence of Conduct Disorders is believed to be an interaction between neuropsychological deficits and social environments that increase the risk of adaptation failure and exacerbate existing behavioral maladjus tment (Moffit, 1993). Patterson (1986) has proposed a model to explain t he emergence of oppositional behavior in childhood based on a negative reinforce ment paradigm. In this model, the reactions of caregivers to child noncompliance unin tentionally increase the chances that

PAGE 49

38 the child will engage in future noncompliance. Pare nts counter their child’s initial noncompliant behavior with highly punitive, emotion ally laden responses. The child then responds by escalating their behavior. Eventually, the parents withdraw the initial request and withdraw from the aversive interaction with the ir child. The culminating effect of such interactions is a tendency for the child to es calate their behavior in response to task demands perceived as aversive. Further, the noncomp liant response precludes the development of prosocial responses to aversive situ ations. Coercive patterns of social interaction often generalize to other social enviro nments and individuals, leading to a pattern of oppositional behavior that often impairs the child’s ability to meet developmental expectations across a wide range of e nvironments and developmental periods. Risk and resilience models. The concepts of risk and resilience emerged in the literature along with development models. These mod els are similar, and together, help explain why some children develop social competenci es despite the presence of aversive environments (Masten & Curtis, 2003). According to risk and resilience models, children encounter various developmental tasks as they progr ess across the life span. Adaptational failure is associated with problems in that attainm ent of competencies necessary for accomplishing a developmental task. Any factor that increases the probability of adaptational failure can be conceptualized as a ris k factor. The accumulation of risk can interfere with the acquisition or performance of be haviors that facilitate attainment of developmental tasks of childhood. Despite the delet erious impact of risk factors, other variables can enhance a child’s capacity to develop competencies that facilitate adaptation. These variables are known in the risk a nd resiliency literature as protective

PAGE 50

39 factors because they increase the capacity for adap tation despite the presence of accumulated risk factors. The concept of resilience refers to “successfully coping with or overcoming risk and adversity or the development of competence in the face of severe stress and hardship” (Doll & Lyon, 1998, p. 348). R isk and protective factors interact across each stage of development and affect traject ories associated with negative adult outcomes (Sroufe, 1997). In one of the most influential investigations of r isk and protective factors, Werner (1989) identified several variables that impacted t he developmental course of high risk and resilient children. The Kauai Longitudinal Stud y represents one of the largest interdisciplinary investigations of high risk indiv iduals and protective factors. The original sample consisted of a multiracial cohort o f individuals born in 1955 on the island of Kauai, Hawaii. The study monitored the impact of various biological and psychosocial risk factors, stressful life events, and protective factors across childhood and into adulthood. One third of the sample was considered “ at risk” because they had experienced moderate to severe degrees of perinatal stress, poverty, and were raised by mothers with little formal education. The home envi ronments of the at-risk portion of the sample also were characterized by discord, desertio n, or divorce, and evidenced parental alcoholism or mental illness. The at-risk children who accumulated four or more of the risk factors by age two, developed severe learning or behavior problems by age ten, and had delinquency records, mental health problems, or teenage pregnancies by the age of 18 (Werner, 1989). However, several individuals emerged from these hig h risk environments to live more successful lifestyles. Temperamental character istics in infancy such as alertness and

PAGE 51

40 autonomy, communication skills, locomotion and self help skills were identified as protective factors. The resilient portion of the hi gh risk sample also developed healthy relationships with peers and had better reasoning a nd reading skills during elementary school. They had developed a positive self concept and internal locus of control by high school graduation, found emotional support beyond t heir families, and relied on informal social networks for support during times of crises (Werner, 1989). Distributional attributes of the individual, affectional ties with in the family that provide emotional support in times of stress, and external support sy stems such as school, work, or church emerged as the three types of protective factors as sociated with positive adult outcomes. Risk and correlated constraints. Farmer, Quinn, Hussey, and Holahan (2001) discuss the development of behavioral disorders in the context of correlated constraints. The issue can be integrated into developmental and risk models of psychopathology. Correlated constraints refer to the multiple factor s that contribute to the development of disruptive behavioral disorders. Behavioral, biophy sical, cognitive, contextual, emotional, and social interaction variables interac t to influence behavioral development. The presence of multiple risk factors influences th e emergence of disturbance. Various developmental factors constrain each other and prom ote stability across each developmental period. For example, the presence of a supportive family environment, the development of prosocial behavioral patterns, and t he presence of a supportive social network can prevent early learning problems from es calating into severe emotional and behavioral maladjustment. In this example, the pres ence of these protective factors constrains the potentially negative affects of earl y learning problems.

PAGE 52

41 Alternatively, negative outcomes can occur when co rrelated constraints interact to support the development of behavioral disturbance. The same child who experiences early learning problems will likely experience beha vioral disturbance over the course of development if protective factors in the environmen t are limited. The development of oppositional behavioral patterns, inappropriate sup ervision and discipline at home, coercive interactions with parents and teachers, as sociation with a deviant peer group, and the development of social roles that result in reinforcement of antisocial behavior can lead the development of behavioral disturbance (Far mer et al., 2001). Developmental models have important implications fo r the design of effective service delivery systems. If contextual factors aff ect the development of severe psychopathology, then environments can be designed to facilitate the development of social competence (Sroufe, 1997). Early identificat ion of negative developmental trajectories can lead to the delivery of early inte rvention services designed to increase social competence while preventing adaptational fai lure. These interventions services can address negative developmental trajectories before emotional and behavioral maladjustment escalates to a point where the most i ntensive services become necessary. Further, the presence of environmental support syst ems that increase protective factors and resiliency within at-risk children may greatly reduce the need for placement in settings separate from typical peers (Doll & Lyon, 1998). Efforts to address severe and emotional disturbance also must account for the pre sence of correlated constraints. Prevention and intervention efforts are likely to f ail if only one risk factor is addressed. The presence of other risk factors will continue to constrain prosocial development and influence the development of emotional and behavior al disturbance (Farmer et al., 2001).

PAGE 53

42 When support services are limited, deficits in soci al competence can lead to successive occurrences of adaptational failure. The risk of de veloping severe psychopathology increases with each experience of adaptational fail ure (Sroufe, 1997). Additional resources must be allocated if environments are unp repared to address the needs of children experiencing severe psychopathology. The a lternative approach is to exclude students from their typical environment or to look beyond those environments to the community for expertise in dealing with children wh o experience severe psychopathology (Doll & Lyon, 1998). Support Services Developmental models help explain exclusion in envi ronments with fewer support services. Children with multiple risk facto rs often experience higher rates of adaptational failure early in childhood (Doll & Lyo n, 1998). The development of severe psychopathology is more likely when negative develo pmental trajectories are unaddressed through competency based support system s. Successive occurrences of adaptational failure often culminate in emotional a nd behavioral disturbance that often prevents successful integration into environments i n non-disabled peers (Sroufe, 1997). Educational environment often encounter children th at manifest severe childhood psychopathology. In an effort to address the wide range of emotional and behavioral problems encountered in the school settings, variou s tiered level systems of behavior and emotional support have been proposes in the literat ure (Gresham 2004). A three tiered model of school based service delivery is described in this section. Tier models of service delivery. Gresham (2004) provided a summary of school based behavioral intervention support system s based on a three tiered model. The

PAGE 54

43 allocation of resources and time are directly propo rtional to the frequency and intensity of challenging behavior that is manifested within the environment. Additional resources and personnel who possess increasingly more expertise a re available at each level of support. Therefore, within three tiered model of behavioral supports, resource allocation is directly proportional to the intensity of the prese nting problem and the resources that are needed to effectively address emotional and behavio ral disturbance. Tier of models behavior support are considered beneficial because resource allocation is associated with the severity of the emotional/behavioral problems m anifested by children (Deno, 2002). As the name implies, three levels of intervention s upport levels around implemented in a three tiered model of behavior sup ports. These levels are universal interventions, selected interventions, and targeted intensive interventions. Universal interventions are consistent with a primary prevent ion focus and target all students within a district, school building, or classroom. Each stu dent receives a similar intensity of services at the universal level. It is estimated th at between 80-90% of Tier I, universal interventions will be effective in preventing adapt ational failure in academic and/or social domains (Gresham 2004, Walker & Shinn, 2002). Approximately 5-10% of students who receive univer sal interventions will continue to manifest adaptational failure. Selected interventions, sometimes referred to as secondary prevention efforts (Walker & Shinn, 2002) are intended for students who require resources beyond what is offered at the uni versal level of intervention. These students often require interventions that are desig ned to target specific competency deficits. At tier II, the goal of interventions is to increase the likelihood that at-risk

PAGE 55

44 students will respond to interventions delivered at the Tier I, universal level of prevention (Gresham, 2004; Walker & Shinn, 2002). Despite the implementation of interventions at the universal and secondary levels, approximately one to five percent of students will continue to require intensive services to address adaptational failure that manifests as s evere behavioral disturbance. These students may require resources and expertise typica lly unavailable in most school environments. Consequently, involvement of mental h ealth, juvenile justice, and social services often is necessary to address emotional an d behavioral problems that have failed to respond to less intensive systems of support. At tier III, the reduction of problem behaviors that threaten the safety and well being o f the target student and his or her peers becomes the goal of intervention. Community based r esources often are utilized to maximize the intensity and expertise of interventio n that are implemented for the child. Consistent with universal and selected levels of in tervention supports (Tier I & Tier II respectively), the goal remains the development of student academic and social competence that will facilitate adaptation to less restrictive environments (Gresham 2004). Interventions at Tier III or the targeted le vel may include wraparound services that involve collaboration between families, schools, an d communities (Walker & Shinn, 2002). Behavioral support systems based on levels of supp ort have important implications from a developmental perspective. Cons istent with developmental models, negative long term outcomes for children can be red uced by decreasing the duration of negative development trajectories. Risk increases a s a child remains on a negative trajectory (Masten & Curtis, 2003). Tier level supp ort systems facilitate the identification

PAGE 56

45 of students on negative developmental trajectories. Early identification, coupled with interventions designed to enhance protective factor s and competencies can significantly reduce the number of children who require intensive specialized treatment corresponding to Tier II and Tier III levels of support (Gresham, 2004). Tier models and restrictive placement. Behavioral support systems based on tiered models suggest that placement in the most re strictive settings may be associated with the failure to provide adequate resources in l ess restrictive environments. Tiered models of behavior support are associated with stre ngth oriented approaches to mental health service delivery. A strength oriented approa ch shifts the focus of service delivery to the identification of protective factors within the child’s environment, and resources are directed towards the implementation of competen cy based support systems. The three tiered model facilitates a proactive approach to se rvice delivery, and resources are allocated according to the needs of children who re ceive services within each tier. Program development is based on the needs of a targ eted population, and services are provided in environments consisting of healthy chil dren (Power, 2002). Movement along the tiers is based on the child’s response to inter ventions delivered within the context of each tier. Children will move into a higher level o f services if the resources necessary to obtain an adequate response are unavailable (Gresha m, 2002). Strength oriented approaches can be compared to a d eficit approach to service delivery. Services typically are delivered in respo nse to a referral, which often occurs in the context of a crisis situation. Deficit oriented approaches are associated with reactive responses to crisis situations. Consequently, more intensive services are necessary and can lead to placement in the most restrictive setti ngs. These settings typically consist of

PAGE 57

46 children who demonstrate the most severe examples o f adaptational failure and psychopathology. Systems without such methods for a llocating resources in proportion to the needs may be more likely to develop deficit ori ented approaches and adopt approaches which lead to placement in the most rest rictive settings (Power, 2002). Availability of Support Services The literature reviewed thus far suggests that ment al health and behavioral support services can facilitate the development of social competence among at-risk youth. Successive occurrences of adaptational failure can be prevented when support services are provided early in the course of a negative deve lopmental trajectory (Masten & Curtis, 2003; Sroufe, 1997). Unfortunately, when it comes t o mental health and behavior support services, children represent an underserved segment of the population. Many children with the most severe mental health problems fail to receive treatment that can potentially lead to improvements in their social functioning. A ccording to the United States Department of Health and Human Services (1999) appr oximately 20% of children and adolescents have an identifiable mental disorder. B etween five to nine percent of children have a serious emotional disturbance and extreme fu nctional impairment, and 9-13% have a serious emotional disturbance and substantia l functional impairment. However, only 21% of children and adolescents identified wit h mental health concerns receive treatment (United States Department of Health and H uman Services [US DHHS], 1999). Several barriers to treatment have been identified in the literatures. Parent reports indicate that schedule constraints, divorced status and the presence of three or more psychosocial stressors limit access to treatment (O wens et al., 2002). Barriers also include a lack of transportation, unavailability of services, poor quality of services,

PAGE 58

47 scheduling conflicts, the cost of obtaining service s, ineligibility for services, location of services, transportation limitations, and language/ communication conflicts (Wagner & Sumi, 2005). Schools often become a primary provide r of mental health services because of these barriers that limit access to treatment. School based support services. Adelman and Taylor (1999) reviewed several mental health projects implemented in schoo ls across the United States. School based mental health programs vary according to the nature and scope of services provided within educational settings. These service s may focus on psychosocial problems such as school adjustment and attendance problems, dropout prevention, physical and sexual abuse, substance abuse, relationship difficu lties, emotional problems, teen pregnancy, and delinquency and violence. Programs f ocus on improving the capacity of schools to address emergency situations and enhance emotional and social well being, resiliency, self-esteem, intrinsic motivation, empa thy, and prosocial skills among students. Services also involve efforts to promote mental health, minimize the impact of psychosocial problems, psychotropic medication mana gement, and participating in a system of care. Larger school districts typically o ffer a wider continuum of preventative and intervention based approaches. The location and ownership of these services varies and often involves services operated solely by scho ol districts, and may include collaboration efforts between schools and community based agencies. Mental health services may be provided to all students within a s chool, those in certain grades, or targeted towards at-risk populations. Interventions can be delivered in regular or special education programs, school-based health centers or family resource centers. These services may encompass efforts to collaborate with community based agencies in an

PAGE 59

48 effort to address the needs of children with severe emotional and behavioral disorders. Consistent with tier models of service delivery (Gr esham, 2004), the projects reviewed by these authors emphasized the importance of a contin uum of coordinated and accessible services. These programs include preventative and e arly intervention programs and treatment programs for severe emotional and behavio ral disturbance. According to Adelman and Taylor (1999), several fa ctors have hampered the development of school based mental health programs. Existing educational policies typically have assigned a low priority to school ba sed mental health services. Consequently, the development of a comprehensive, i ntegrated, and multifaceted approach has been limited. Piecemeal programs and s ervices based on categorical funding formulas often limit the scope and range of the services provided to students. School based mental health programs often are restr icted to children who manifest severe emotional and behavioral disturbance with limited e mphasis on early intervention programs. Foster et al. (2005) examined the types of services and supports delivered by public schools in the United States during the 2002 -2003 school year. The national survey collected data from 83,000 public elementary middle, and high schools and their associated school districts. Results indicated that student were eligible to receive mental health services in 87% of schools. Ten percent of s chools reported that students must have an Individualized Education Plan (IEP) to qual ify for services. Assessment for mental health problems, behavior management consult ation, and crisis intervention were available in 87% of school. Referrals to specialize d programs were available in 84% of schools. Individual counseling, case management, an d group counseling were provided in

PAGE 60

49 76%, 71%, and 68%of schools, respectively. Substanc e abuse counseling was provided in less than half of all schools, and medication/medic ation management was the least likely service to be provided. Seventy percent of schools reported that school counselors were available to provide mental health services, 69% re ported the use of school nurses, 68% utilized school psychologists, and 44% utilized soc ial workers as mental health service providers. Mental health counselors, substances abu se counselors, clinical psychologists, and psychiatrists were reported to be utilized in l ess than 20% of schools. Only 15% of school reported the use of school wide screening fo r emotional and behavioral problems. Prevention and pre-referral intervention programs w ere implemented in 63% of schools. Curriculum based programs were implemented in 59% o f schools and school wide strategies to promote safe and drug free schools we re provided by 78% of schools. Slade (2003) also examined the availability of sch ool based mental health services in the United States. Data were collected from the first wave of the ADDHealth study, which is a nationally representative survey of stud ents in grades 7 through 12. Administrators from 132 schools were asked about th e availability of health services either at school or at another school located withi n the district. Results indicated that approximately 50% of middle and high schools offer on site mental health counseling services and 11% have mental health counseling, phy sical examinations, and substance abuse counseling on site. Rural schools, schools in the Midwest and South regions, and smaller schools were least likely to offer mental h ealth counseling. Access to Medicaid funding for financing health services was related t o disparities between schools in the availability of mental health counseling.

PAGE 61

50 Students who exhibit significant emotional and beha vioral problems in school often qualify for special education services under the serious emotional disturbance category (SED). For many students, the special educ ation system becomes the primary mechanism for accessing behavioral and mental healt h services. The Individuals with Disabilities Education Improvement Act (IDEIA, 2004 ) mandates a free and appropriate education for students who are identified as having one of thirteen disabling conditions, including SED. Students who are eligible for specia l education must receive services to address their individual needs. A child is eligible under the serious emotional disturbance category (SED) if they have a disorder in at least one of five criterion areas. The five areas are (a) an inability to learn that cannot be explained by intellectual, sensory, or health factors; (b) an inability to build or mainta in satisfactory relationships with peers or teachers; (c) inappropriate types of behaviors or f eelings under normal circumstances; (d) a general pervasive mood of unhappiness or depressi on; and (e) a tendency to develop physical symptoms or fears associated with personal or school problems. Schizophrenia is included in these criteria. Eligibility criteria do not apply to children who are social maladjusted unless they also have an emotional hand icap. The presence of one of these conditions must cause adverse effects on school per formance. The intention of IDEIA is to provide a free and app ropriate education to students with disabilities in the least restrictive environm ent. The assumption underlying IDEIA is that children with disabilities require a specializ ed curriculum and related services tailored to the individual needs of students with d isabilities. Thus, IDEIA serves as a mechanism by which access to mental health services is provided for students whose behavior impedes learning. However, the literature reviewed reveals several concerns

PAGE 62

51 related to the implementation of mental health and behavioral support services for children in school. First, many students experience a protracted period of educational failure prior to receiving any services at all or m ay receive services limited to their area of eligibility. The current system of special educa tion and related eligibility criteria may inadvertently contribute to a narrowed conceptualiz ation of student needs and a restricted view of student functioning. These practices may pr eclude the implementation of support services for students with emotional and behavioral disorders until problems become so severe that they require the most intensive resourc es available. Further, students may first qualify for services under one category and fail to receive support services to address areas of concern that are considered secondary to t he primary category. For example, students may have academic and behavioral problems which impact their adjustment at school. However, student may qualify for services u nder the SLD label and fail to receive mental health and behavioral support services. Morrison and D’Incau (2000) utilized a qualitative study design, using a combination of ethnomethodology and case study anal ysis to examine the case histories of 41 students. The students involved in the analys is had received special education services and had been recommended for expulsion. An alysis of the service trajectories across childhood indicated that many students did n ot receive services despite evidence of pervasive problems in multiple domains of functi oning. Eligibility criteria associated with these programs appears to have prevented acces s to these services. These students eventually were identified as SLD or SED, and two t o three evaluations were conducted until services were rendered. Another group of stud ents received minimally intensive services such as speech, followed by resource speci alist programming, and eventually

PAGE 63

52 were placed in special day classes. Analysis also i ndicated a “jagged” service trajectory, characterized by students who received services und er different categories at various points in time. Duncan, Forness, and Hartsough (1995) generated sim ilar findings from a sample of students in a school based day treatment program Archival records were used to examine the diagnostic and treatment histories of 8 5 children and adolescents served in two school based day treatment programs. Data sugge sted that very few systematic interventions occurred between identification of th e problem and initial placement in day treatment. The implementation of interventions typi cally were initiated a little more than a year after problems were noticed, and placement i n a school based day treatment occurred four to six years later. Consequently, stu dents failed to receive any services until they had experienced several years of adaptational failure. The interventions that were implemented consisted primarily of medication treat ment, infrequent outpatient therapy unrelated to other interventions, and resource room consultation for learning problems. Analysis of the case histories provided limited evi dence of collaboration between school and mental health personnel prior to placement in t he day treatment center. Once students do qualify for special education, it is expected that they will receive a comprehensive array of behavioral and mental heal th services. Although national data indicate that a vast majority of public school repo rt the implementation of mental health services (Foster et al., 2005), students with ED ma y not receive adequate services to address their social-emotional needs. Wagner and Su mi (2005) examined the services provided to students with Emotional Disturbance (ED ) and found that very few school based services were implemented. Data from the Nati onal Longitudinal Transition Study-

PAGE 64

53 2 (NLTS2) indicated that 38% of students with ED in general education classrooms received no curricular modifications. Fewer than 25 % of students with ED received learning supports to address behavior and learning concerns, 23% had a behavior management plan, 21% received assistance with study skills and instruction in learning strategies, 15% have an adult tutor, and 15% have a peer tutor. Sixty-nine percent of children with ED received psychological/mental heal th services at school. Forty-four percent of students with ED received substance abus e prevention or substance abuse treatment at school, and 30% participated in confli ct resolution, anger management, or violence prevention programs. Among these students, 36% received school based psychological/mental health services. These data su ggest that even when students qualify for special education services under the ED categor y, they often fail to receive appropriate modifications to the curriculum and sup port services designed to enhance social competence. In addition to providing support services for child ren who qualify under the ED category, IDEIA provides due process rights and pro cedures for students with disabilities who are suspended or expelled. Students with disabi lities are protected from permanent expulsion because of regulations from IDEIA, 2004). According to IDEIA, schools are mandated to make a determination as to whether a vi olation of school disciplinary policy by a special education student is related to his or her disability if the disciplinary response results in a suspension of over ten days (or a susp ension that would, with previous suspensions that year, total more than ten days) or expulsion, which would constitute a change of placement. However, special education stu dents can be placed in an interim alternative educational setting for up to 45 days. These settings include: (1) a more

PAGE 65

54 restrictive special education setting, such as a se lf contained program, (2) an alternative school, (3) an age-appropriate mental health facili ty, (4) a condition specific medical facility, (5) homebound instruction, or (6) a court mandated correctional facility (IDEIA, 2004). A substantial proportion of students whose b ehavior impedes their adaptation fail to receive support services because they have been labeled as socially maladjusted. These students then become subjected to disciplinary prac tices that exclude students from school and may not receive support services designe d to enhance social competence. Thus, the social maladjustment clause may actually limit services to children who fail to meet the criteria for eligibility. One could reason able conclude that many youth in the United States become subjected to punitive discipli nary practices and fail to receive behavioral and mental health services because they have been labeled socially maladjusted. When students are expelled form school, regardless of whether or not they receive special education services, they often are placed i n alternative education programs. A recent study conducted in the state of Florida sugg ests that very few alternative programs provide academic assistance, behavioral interventio ns, and counseling services for students placed in alternative school programs. Ber gquist, Bigbie, Groves, & Richardson (2004) examined data relevant to alternatives to su spension collected from 50 of the 67 school districts in the state of Florida. Available funds and staffing resources were identified as the two most prevalent factors impact ing decisions made about service delivery models. Very few instances of academic ass istance were found across the districts. For example, a specific academic curricu lum was used in only 32% of alternative schools/off-site locations, 24% of in-s chool suspension programs, and 18% of

PAGE 66

55 prevention programs. Tutoring was provided in 24% p revention programs, 18% alternative schools/off-site locations, and 18% of in-school suspension programs. Counseling programs and behavioral interventions we re offered in very few instances. The research presented in this section suggests tha t children with EBD face substantial barriers in obtaining behavioral and me ntal health services (Owens et al., 2003). A large percentage of schools indicate that they have behavioral and mental health services available for students (Foster et al., 200 5). However, a closer examination of the literature suggests that students with the most int ensive needs often receive limited accommodations and support services (Wagner & Sumi, 2005). Further, students often experience several years of failure before they acc ess these services. The services that are provided often are limited in scope and target what is believed to the primary disability, while failing to address associated learning or emo tional/behavioral concerns (Morrison & D’Incau, 2000). Factors Associated with Restrictive Placement Due to the relative scarcity of mental health servi ces, children with EBD enter negative developmental trajectories and the stage i s set for a chronic pattern of maladjustment that often requires intensive service s only available in the most restrictive settings. Tier models predict that children will pa ss through Tier I and Tier II levels of support before their behavior problems escalate to a point where the most intensive resources are necessary at Tier III. The child’s re sponse to less intensive interventions should predict the delivery of support services at higher levels along the continuum of supports (Gresham, 2004). However, there are cases in which a child may bypass Tiers I and II, and require the most intensive support serv ices at Tier III. Thus, involvement at

PAGE 67

56 Tier III can result from an aggregated accumulation of behavioral disturbance that is unresponsive to less intensive interventions. Invo lvement at Tier IIII also can occur after a single behavioral event prompts educators to look beyond the school setting for the most intensive support services available. The mode l presented here predicts that referrals for the most intensive support services available c an be attributed to two factors. First, the most intensive and restrictive placement options sh ould be reserved for children who exhibit the most severe emotional and behavioral di sturbance. Second, less restrictive placement should be associated with the availabilit y of support services within those settings. Referrals to the most intensive services will be seen as an appropriate response to severe emotional and behavioral disturbance when adequate support services are unavailable in less restrictive environments. Severity of behavior. The severity of the students emotional and/or behav ior problem, as perceived by adults in the child’s envi ronment, appears to be related to placement in the most restrictive educational and t reatment settings. Hendrickson, Smith, Frank, and Merical (1998) examined records from 99 students with severe emotional and/or behavioral disorders (EBD) to investigate fa ctors related to placement in more restrictive educational settings. Forty-nine studen ts received educational services in regular schools and 50 students received services i n a segregated school. Data collection methods included records review of demographic char acteristics, educational achievement, data on the IEP decision making proces s, and information pertaining to the justification for current special education placeme nt. Telephone interviews with one IEP team also served as a source of data. Results indic ated that staffing team members perceived the students in segregated school setting s to display more intense, severe, and a

PAGE 68

57 longer duration of emotional and/or behaviors disor ders, posed a greater threat to others, had poor self-control, more problems with noncompli ance, parental support issues, and problems with large school environments. Muscott (1997) examined the characteristics of stud ents with emotional and behavioral disorders (EBD) across four sequential s pecial education placements. Participants were 473 students served in resource r ooms within regular schools, special classes, special schools, and residential schools. Aggression and disruptive behavior were the most common behavioral profile of students with EBD. Elementary school students in residential settings were perceived to be more mala daptive than peers in resource rooms and special classes. Teaches perceived younger elem entary age students as exhibiting more adaptive behavior than older peers in resource rooms, but less adaptive behavior than peers in residential schools. In particular, e lementary school students in resource rooms were rated as less aggressive and disruptive than peers in separate special schools and residential schools. There were no differences among secondary students across the range of available placements. The severity of emotional and behavioral disorders also appears to be associated with the most restrictive placements in community s ettings. McDermott, Mckelvey, Roberts, and Davies (2002) found evidence to suppor t this hypothesis. Intake data for 603 children ages 4 to 16 years of age who visited a me ntal health treatment facility located in a pediatric hospital were utilized in the analysis. Care options included an in-patient unit, a day treatment program, enrollment in outpatient t reatment, or tertiary consultation services. Results indicated that the most costly an d time intensive settings were reserved for children with more severe psychopathology and m ore severe family dysfunction.

PAGE 69

58 Children who were admitted to the in-patient progra m were rated by parents as having more severe total psychopathology than children in outpatient treatment and tertiary consultation. Parent reports indicated that childre n in the most restrictive settings such as in patient and day treatment programs had lower com petency ratings than children receiving outpatient and consultation based service s. Adolescents receiving in-patient services had higher internalizing scores than child ren receiving day treatment, outpatient treatment, and tertiary consultation. Competency sc ores for adolescents receiving treatment through the in-patient program were lower than scores from adolescents attending outpatient and tertiary consultation pati ents. Additionally, adolescents receiving in-patient treatment were more likely to have an ax is II diagnosis than adolescents receiving outpatient treatment and tertiary consult ation. Availability and exclusion. The Hendrickson et al. (1998) study cited above indicated that placement in the most restrictive ed ucational settings was partly associated with the severity of emotional and behavioral disor ders displayed by children. However, the decision to place students in the segregated sc hools also was based on the greater availability of resources in those settings. Staffi ng team members indicated that decisions to place students in more segregated settings were associated with the presence of smaller class sizes, more specially trained staff, psycholo gical therapies, a more structured and supervised environment, increased family therapy/in volvement, more flexible scheduling, access to crisis room/personnel, and reduced distra ctions within those settings. Students in the segregated settings were perceived to be mor e aggressive and in need of more intensive intervention available in more restrictiv e settings. The severity of the students’ behavior seemed to be evaluated in terms of the sup port services necessary to

PAGE 70

59 accommodate students in the classroom. Students wer e placed in less restrictive settings when the intensity of the services matched the perc eived intensity of the students’ mental health, behavioral, and educational needs. Other studies have shown that placement in more res trictive settings often is based on the perception that students will receive the most appropriate services in those settings. Coutinho & Oswald (1996) examined nationa l and state placement patterns of students with SED between 1988 and 1991. Data from a national data base of 5.3 million children who received special education services un der the Individuals with Disabilities Education Act (IDEA) were used in the analysis. The data base contained child count, placement, and exiting information for children who received special education services under the SED category. State placement data were m erged with a data base comprised of economic and demographic variables. Descriptive, co rrelational, and step wise regression analyses and analyses of variance (ANOVA) were cond ucted to address the research questions. Results indicated that states with the h ighest quartile for serving Caucasian students served more students in regular classes an d fewer students in separate classes compared to states in the lower three quartiles. Fu rther, states with higher per pupil revenues and per capita incomes served more student s in restrictive settings. These results suggest that states with more financial res ources offered a continuum of treatment options and placements to students, and offered mor e comprehensive services. States with less financial resources may have developed pr ograms that provided services within less restrictive environments because more intensiv e services were unavailable in other settings. These results suggest that schools will u tilize support services in more restrictive environments when those services and placements opt ions are available. However, when

PAGE 71

60 other options are unavailable, students tend to rem ain in school environments that are in close proximity to non-disabled peers. In addition to the availability of appropriate sup port services in more restrictive environments, school personnel seem to choose setti ngs where that have staff trained to address the needs of students with severe EBD. Rock Rosenberg, and Carran (1994) examined program demographics, reintegration orient ation, teacher reintegration training, and teacher attitude to identify factors associated with the reintegration of students with serious emotional disturbance (SED). Data were coll ected from 162 special education teachers and 31 administrators in restrictive educa tional settings in grades k-12. Hierarchical regression analysis was utilized to co mpare reintegration rates among students in the schools served by these educators. Results indicated that a positive reintegration orientation, certain demographic char acteristics of the educational programs, and SED teacher experience and training a ccounted for a significant amount of variance in reintegration rates. Variables associat ed with a positive reintegration orientation included the availability of multiple r eintegration options such as part day, trial, and transitional reintegration options, easy to implement reintegration procedures, and reintegration training for special teachers. Fu rther, several schools serving the most severe student populations in separate public and n onpublic facilities had a more positive reintegration orientation and higher rates of reint egration. Demographic characteristics associated with higher reintegration rates included an SED program located in a wing of a comprehensive school building, the program was zero to 1 mile from the reintegration site, and the availability of multiple reintegratio n sites. The location of a SED program in the wing of a comprehensive school building was ass ociated with reintegration rates

PAGE 72

61 regardless of the severity of the student populatio n. SED teacher training and experience variables associated with higher rates of reintegra tion included greater number of places a teacher received reintegration training, more years experience serving students with SED, advanced levels of teacher training and advanced ce rtification in special education. The relationship between the availability of a cont inuum of services and placement in the most restrictive settings has been addressed in community settings as well. The findings have implications for school bas ed service delivery because they document the influence of a continuum of option on placement decisions on placement in the most restrictive settings. Bickman, Foster, and Lambert (1996) provide evidence to support this argument. They compared hospitalizatio n rates for children receiving mental health services in two models of mental health serv ice delivery. At one site, a continuum of services were available including home based cou nseling, after school group treatment services, day treatment services, therapeutic homes specialized treatment homes, and a 24-hour crisis management team. Interdisciplinary t reatment teams provided case management, and children received a comprehensive i ntake evaluation to determine the most appropriate level of service. The comparison s ite provided mental health services according to a traditional health insurance model. Results indicated that children who received treatment through the continuum of care mo del were less likely to be hospitalized than children in the traditional syste m. The most severe cases were admitted to hospitals. The results suggest that clinicians u tilized the least restrictive alternatives to hospitalization when those alternatives were availa ble. Clinicians at the comparison site had three treatment options available: outpatient v isits, long term residential care, and

PAGE 73

62 short-term hospital care. Consequently, fewer optio ns were available in less restrictive settings. Children with severe emotional and behavioral distu rbance continue to experience substantial impairment after receiving intensive me ntal health services. HallidayBoykins, Henggeler, Rowland, & DeLucia (2004) exami ned the symptom trajectories of children, psychosocial factors, and placement outco mes following psychiatric crises. The children in the study demonstrated symptoms of suic idal ideation, homicidal ideation, psychosis, or threat of harm to self or others due to mental illness that warranted psychiatric hospitalization. Evidence of symptom el evation in the high to borderline clinical range through 16 month post crises were id entified for half of all children in the sample. These results suggest that a high percentag e of youth continue to demonstrate severe emotional and behavioral disturbance followi ng the most intensive psychiatric services and will continue to benefit from access t o mental health services. According to literature reviewed, the availability of mental health services should prevent additional referrals to more restrictive en vironments after children have been reintegrated into mainstream environments. Romansky Lyons, Lehner, & West (2003) addressed this issue by examining data pertaining t o 1,275 children ages 7 to 17 in custody of the Illinois Department of Child and Fam ily Services. The final sample consisted of 500 randomly selected children who had been hospitalized in a psychiatric facility. The rate of hospital readmission was 21% for the sample. Post hospital service hours and living arrangement were associated with h ospital readmission rates. Children who received more service hours from the Screening, Assessment, and Supportive Services (SASS) program were less likely to be read mitted. The SASS program provided

PAGE 74

63 progress monitoring services for children who requi re psychiatric care, deflection of services for children who remain in the community, and support services following discharge from an inpatient setting. The literature reviewed in this section suggests th at placement in the most restrictive settings are reserved for children with more severe forms of psychopathology. Children who are judged to be less competent and mo re maladaptive tend to receive services in the most restrictive settings (Hendrick son et al., 1998; McDermott et al., 2002; Muscott, 1997). However, decisions to place s tudents in the most restrictive settings also are associated with availability of s upport services within those settings. Educators and other clinicians choose placement in more restrictive settings when they perceive a greater availability of intensive servic es in those settings (Bickman et al., 1996; Hendrickson et al., 1998; Rock et al., 1994; Romansky et al., 2003). Thus, exclusion becomes a likely response among educators and community based clinicians who believe that the support services available in less restrictive setting are insufficient in terms of the intensity and expertise required to pr oduce desired outcomes. Discipline and exclusion. It is clear from the literature reviewed thus far that schools encounter many students who exhibit severe behavioral and mental health problems and require intensive support services tha t may be unavailable in the typical school setting. School personnel may choose to util ize behavioral support services to address behavioral maladjustment in school. Schools also may react to student behavior, regardless of whether a student receives special ed ucation services, through disciplinary exclusion such as suspension and expulsion. Discipl inary exclusion reflects a decision to bypass the system of support services available to children who demonstrate maladaptive

PAGE 75

64 behavior. School personnel who favor reactive appro aches to discipline often believe that removal of a student from the school environment is an appropriate response to maladaptive behavior. School personnel who utilize suspension and expulsion are concerned about how disruptive behavior interferes with the learning environment (Costenbader & Markson, 1994). They are more likely to endorse beliefs that focus on removing disruptive students to maintain order and increase the ability to teachers to provide effective instruction (Wu et al., 1982; Cos tenbader & Markson, 1994; RaffaeleMendez et al., 2002). The reactions and philosophies of school administra tors and school boards influence the degree to which schools rely on exclu sionary disciplinary practices. The implementation of intervention based support servic es in school settings is highly dependent on the guiding philosophy of the school. Based on state mandates, school boards determine district policies that govern the use of exclusionary practices. Districts often rely on codes of student conduct to determine appropriate responses to disciplinary infractions. Although these policies attempt to def ine standards that guide the use of exclusionary disciplinary practices, the idiosyncra tic interpretation and implementation of these policies often affects differential rates of exclusion (Morrison & Skiba, 2001). Ultimately, these beliefs and philosophies influenc e the decision to address maladaptive behavior by providing intervention services designe d to enhance academic and social competency. The alternative to competency based sup port services is to choose discipline methods that often culminate in the removal of a ch ild from mainstream environments via suspension and expulsion.

PAGE 76

65 Hyman and Perone (1998) argued that inaccurate perc eptions regarding the extent of school violence in schools have led to the devel opment of educational policy that favors punitive approaches to misbehavior. Intrusiv e procedures such as strip searches, under cover agents, and other law enforcement orien ted procedures have been implemented to address disruptive behavior in schoo l settings. Consequently, schools have been less willing to adopt educational models that emphasize prevention and developmentally appropriate remediation programs to address behavioral maladjustment. Exclusion often becomes the most likely response to maladaptive behavior in schools (Hyman & Perone, 1998). Zero tolerance policies continue to be developed by schools across the United States in response to the general perception that s chool violence is rampant in schools (Skiba, 2000). According the Advancement Project/Ci vil Rights Project (2000), forty-one states currently have laws establishing grounds for suspension and 49 states have guidelines for expulsion. All of these states requi re a recommendation for expulsion due to possession of firearms or other deadly weapons. Eighteen states cite possession, use, or distribution of drugs on school campuses as grounds for expulsion. According to the National Center for Education Statistics (Heaviside Rowand, Williams, & Farris, 1998) 90% of schools have zero tolerance policies for wea pons or firearms, almost 90% for alcohol and drugs, and 79% for violence or tobacco. Suspension and expulsion rates have increased along with the implementation of zero tolerance polices across the United States. Ze ro tolerance policies are intended to send a message that inappropriate behavior will not be tolerated by school personnel. Severe punishment is used following incidents of ma jor and minor disciplinary incidents

PAGE 77

66 to set a precedent regarding tolerance for behavior that interferes with learning. Schools relying on zero tolerance policies often address em otional and behavioral maladjustment with a reactive approach that excludes students fro m educational settings (Skiba, 2000). The use of punitive disciplinary practices preclude s the large scale implementation of student support services designed to address emotio nal and behavioral maladjustment (Hyman & Perone, 1998). Schools that endorse beliefs supportive of discipli nary exclusion may be less likely to provide support services for students dem onstrating emotional and behavioral maladjustment. Raffaele-Mendez et al. (2002) utiliz ed quantitative and qualitative methodology to examine variables associated with ou t-of-school suspension (OSS) in a large school district in Florida. Data collection p roceeded in three stages, and included an analysis of the school districts’ main database, su rvey data, and a comparison of schools with the highest and lowest suspension rates. Inter views with elementary school principals indicated that schools with low OSS were more likely than high OSS schools to incorporate positive reinforcement strategies in to school wide discipline plans. Low OSS schools also were more likely to use social ski lls training to teach acceptable behavior to students. Further, low OSS schools were more likely to develop school wide discipline plans that involved the input of parents and/or the school psychologist or guidance counselor. Conversely, schools with high O SS relied more on punishment paradigms for inappropriate behavior and were less likely to report home-school collaboration to encourage prosocial behavior. At t he middle school level, low OSS schools were more likely to emphasize staff develop ment and training to improve classroom behavior management, and teachers reporte dly received more support for

PAGE 78

67 resolving discipline problems. High schools with lo w OSS rates also were more likely to rely on school staff to increase parent involvement and to include parents in the development of school wide discipline plans. Skiba et al. (2003) examined school disciplinary p ractices that influenced the degree to which principals rely on disciplinary exc lusion to address challenging behavior. Principals completed an on-line survey intended to assess their perceptions about school discipline. Principals who reported negative percep tions about suspension and expulsion were more likely to report a greater prevalence of counseling and bullying prevention programs at their schools. Those who believed that discipline served as an opportunity to teach appropriate skills also were more likely to u tilize counseling and teacher classroom management in-service activities. Schools with lowe r suspension rates had principals who believed that suspension interfered with learni ng time, schools had a responsibility to teach appropriate behavior, and disciplinary pol icies should be adapted to meet the needs of student’s with disabilities. Lower suspens ion rates for the most serious and dangerous offenses were associated with the percept ion that students should receive incentives for appropriate behavior. Higher suspens ion rates were associated with principals who believed that suspension and zero to lerance policies were a necessary component of school discipline, special education d isciplinary provision interfered with the principals authority, the home situations was r esponsible for behavior problems, and that resources were limited and violence was increa sing. In summary, placement in the most restrictive setti ngs appears to be reserved for children with more severe forms of psychopathology. Children who are judged to be less competent and more maladaptive tend to receive serv ices in the most restrictive settings

PAGE 79

68 (Hendrickson et al., 1998; McDermott et al., 2002; Muscott, 1997). However, decisions to place students in the most restrictive settings also are associated with availability of support services within those settings. Educators a nd other clinicians choose placement in more restrictive settings when they perceive a grea ter availability of intensive services in those settings (Bickman et al., 1996; Hendrickson e t al., 1998; Rock et al., 1994; Romansky et al., 2003). Thus, exclusion becomes a l ikely response among educators and community based clinicians who believe that the sup port services available in less restrictive setting are insufficient in terms of th e intensity and expertise required to produce desired outcomes. Additionally, schools oft en choose to rely on punitive disciplinary approaches to address maladaptive beha vior problems in schools. Zero tolerance policies have lead to the implementation of punitive disciplinary practices rather than prevention and remedial programs (Hyman & Perone, 1998; Skiba, 2000). Schools that embrace prevention and behavioral supp ort services appear to have lower suspension rates (Raffaele-Mendez et al., 2002; Ski ba et al., 2003). The literature reviewed suggests that an interaction may exist bet ween the availability of behavioral support services, reliance on disciplinary approach es to behavioral maladjustment, and exclusion from schooling. Conclusion An increasing number of children in the state of F lorida are being referred to for involuntary examination under Baker Act statutes (C hristy, Petrila, Hudacek, Haynes, Wedekind, & Pulley, 2005). It has been suggested th at Baker Act ERs have been over utilized by schools and other community based agenc ies (Florida Senate, 2005). Baker Act examination referrals (ERs) are less common dur ing the summer months when

PAGE 80

69 students are home on summer vacation, suggesting th at a meaningful percentage of Baker Act ERs are being initiated by school personnel (Ch risty et al., 2005; Florida Senate, 2005). Schools may be using Baker Act ERs to remove disruptive students from educational settings in an effort to address severe emotional and behavioral disturbance (Florida Senate, 2005). Several variables may contribute to higher rates of Baker Act ERs. Research that has investigated factors associated with restrictiv e educational placements and disciplinary exclusion suggests several variables t hat may impact a schools decision to exclude students. Males, African American students, and students from low socioeconomic backgrounds are overrepresented in al l types of disciplinary exclusion (Costenbader & Markson, 1994, 1998; Florida Departm ent of Education, 2005; NCES, 2005; Raffaele-Mendez et al., 2002; Skiba et al., 1 997). Students who qualify for special education services under the SED category are more likely than students in other disability categories to be excluded from school (W agner et al., 2005). The literature reviewed suggests that a reciprocal interaction exi sts between the severity of emotional and behavior problems among students, philosophies supportive of exclusion, and the availability of behavioral and mental health servic es. Placement in the most restrictive educational and community based settings occurs whe n adequate resources are unavailable in less restrictive environments (Bickm an et al., 1996; Blanz & Schmidt, 2000; Hendrickson et al., 1998; McDermott et al., 2 002; Muscott, 1994; Rock et al., 1994). Further, the availability of adequate suppor t services can prevent additional placement in the most restrictive environments (Rom ansky et al., 2003). The philosophical inclinations of school personnel can influence the degree to which

PAGE 81

70 exclusion is embraced as an appropriate response to student misbehavior. Zero tolerance policies have contributed to an increasing reliance on punitive approaches such as suspension and expulsion (Hyman & Perone, 1998; Ski ba, 2000). The implementation of punitive disciplinary practices precludes the use o f preventative and positive mental health and behavior support services (Raffaele-Mend ez et al., 2002; Skiba, 2003). The research presented in this literature review ma y contribute to an explanation of Baker Act ER rates among geographic regions in t he state of Florida. School districts with higher rates of suspensions and expulsions may utilize fewer preventative and remedial based behavioral support services. Develop mental models of psychopathology (Sroufe, 1997; Masten & Curtis, 2003) predict that emotional and behavioral disturbance becomes more severe as children interact with avers ive and unsupportive environments. School districts with limited support services will be more likely to have students whose behavior escalates to the point where access to mor e intensive support services becomes necessary. Further, Baker Act ERs may be seen as an appropriate response in schools with fewer support services and discipline policies that embrace exclusion.

PAGE 82

71 Chapter Three Method The purpose of this study is to explore the relatio nship between school and student/school district related variables and Baker Act examination referrals (ERs) among school aged children and adolescents in the state o f Florida. In this chapter, a description of the method that was used to answer the research questions is provided. First, a description of the participants is presented. The d iscussion includes information pertaining to the process by which children and ado lescents are referred for involuntary examinations under the Baker Act. Second, the resea rch design is described. Included in this section is a delineation of the independent an d dependent variable(s). Next, the process used to select the participants for the stu dy and data retrieval procedures are presented. The chapter concludes with a summary of the data analysis plan for each research question. Participants Participants for this study were the school distri cts in the state of Florida and students in those districts who received an involun tary examination under the Baker Act statutes. The Florida public school system consists of 67 school districts and 7 special schools for non-tradition students. These special s chool districts include the Florida School for the Deaf and Blind, Dozier, Florida Virt ual School, Florida Atlantic University Laboratory School, Florida State Univers ity Laboratory School, Florida

PAGE 83

72 Atlantic Metropolitan University Laboratory School, and the University of Florida Laboratory School. During the fall 2005 survey coun t, the pre-k to grade 12 student enrollment was 2,673,563 for the 67 school district s and the special schools. Ethnic affiliation of the student population for the total pre-k to grade 12 enrollment was as follows: White, Non-Hispanic = 49.89%, Black, Non H ispanic = 37.19, Hispanic = 5.22%, Multiracial 4.00%, Asian/Pacific Islander 3. 50%, and American Indian/Alaskan Native 0.21% (Florida Department of Education, 2005 ). The participants for this study also were children and adolescents ages 5-18 who received an involuntary examination under Baker Act statutes in the 2005 calendar year. Florida’s mental health statute, or “Baker Act,” (F .S. 394, Part I, 2005) stipulates that involuntary examination of an individual can be ini tiated if there is reason to believe that the individual has a severe mental illness as defin ed in the law, and because of his or her mental illness: “The person has refused voluntary examination or is unable to determine whether examination is necessary; and Without care or treatment, the person is likely to suffer from neglect resulting in real and present threat of substantial harm that can’t be avoided through the help of others; or There is substantial likelihood that without care or treatment the person will cause serious bodily harm to self or others in the near future, as evidenced by recent behavior.” An involuntary examination may be initiated by any one of the following means:

PAGE 84

73 “A court may enter an ex parte order, based upon sw orn testimony, directing a law enforcement officer to escort an in dividual to the nearest Baker Act receiving facility. A law enforcement off icer may serve and execute an ex parte order on any day of the week, a t any time of the day or night and may use such reasonable physical force as is necessary to gain entry to take custody of the person. A law enforcement officer shall take a person who a ppears to meet the above criteria into custody and deliver the person to the nearest receiving facility. The officer shall execute a written repor t detailing the circumstances under which the person was taken into custody, and the report shall be made part of the patient’s clinical record. A physician, clinical psychologist, psychiatric nur se, or clinical social worker, each as defined in the statute, may execute a certificate stating that he or she has examined a person within the precedin g 48 hours and finds that the person appears to meet the criteria for in voluntary examination and stating the observations upon which that conclu sion is based. A law enforcement officer shall take the person into cust ody and deliver him or her to the nearest receiving facility and shall exe cute a written report detailing the circumstances under which the person was taken into custody.” During the 2005 calendar year, there were a total o f 125,571 Baker Act ER initiation forms received by the Agency for Health Care Administration (AHCA), which maintains a repository of all data related to Baker Act ERs in the state of Florida. Age

PAGE 85

74 was reported for 123,273 or 98.17% of the total for ms that were received by AHCA. A total of 22,547 or 18% of the total referrals were for children and adolescents between the ages of 5-18 years of age. There were 582 or 2.58% duplicate forms (two forms were generated for the same referral). Therefore, a tota l of 21,956 Baker Act ERs were initiated during the 2005 calendar year for childre n and adolescents between the ages of 5-18. Research Design The study utilized a correlational and causal compa rative design to investigate the relationship between the independent and dependent variable(s). The independent variables were entered into a regression model in a n attempt to account for variance in the dependent variable, and to identify the variabl es that contributed most to the prediction of the dependent variable. Archival reco rds from the Florida Department of Education, the Baker Act Reporting Center, and data obtained via a survey completed by district personnel in each of the 67 school distric ts served as the sources of data for the independent and dependent variable(s). Independent variables. The independent variables for the study were: (a) size of the school district (b) the ethnic makeup o f the students in the district, (c) the percent of the student population within a school d istrict that received free and reduced priced lunch, (d) the percent of the student popula tion within a school district that received any special education services under the I ndividuals with Disabilities Education Improvement Act (IDEIA) (U.S. Department of Educati on, 2004), (e) the percent of the student population within a school district that re ceived special education services specifically under the Emotionally Handicapped (EH) / Serious Emotional Disturbance

PAGE 86

75 (SED) category, (f) the percent of students in a di strict that received a suspension, (g) the percent of the student population within a school d istrict that received an expulsion, (h) the percent of the student population within a dist rict that graduated with a standard diploma, (i) the percent of the student population within a school district that were retained, (j) the percent of students in a school d istrict who obtained a Level 3 or higher on the reading and math sections of the Florida com prehensive Achievement Test (FCAT), (k) the ratio of mental health workers (so cial workers, guidance counselors, school psychologists) to students in a school distr ict, (l) the types of mental health services provided to students in a school district, and (m) the methods used to monitor students who returned to school following a Baker A ct examination referral. Dependent variables. The dependent variables for the study were an estim ate of the per capita Baker Act ER rate and the repeat Bak er Act ER rate for each of the 67 counties in the state of Florida. Archival databases. Two archival databases were utilized to obtain data for the independent and dependent variable(s). First, the F lorida Department of Education located in Tallahassee, Florida, served as one sour ce of data for the independent variables associated with salient student and district level variables. The Florida Department of Education maintains a comprehensive management info rmation system. Each school district is required to implement an automated stud ent and staff information system that reports and maintains data on student, staff, schoo l, and district level variables. Schools districts transmit data electronically to the Flori da Department of Education three times per year. These data are maintained by the Florida Department of Education in an automated data base. The data was retrieved from th e Florida Department of Education

PAGE 87

76 website: http://www.fldoe.org. These data are repor ted in html web page and Microsoft excel spreadsheet. It was not necessary to obtain a password or special permission to access these data because they are available to the public by accessing the Florida Department of Education website. Second, the Baker Act Reporting Center database at the Louis de la Parte Mental Health Institute (FMHI) at the University of South Florida in Tampa Florida was used to identify the number of Baker Act ERs initiated in e ach county during the 2005 calendar year. Statutory language in F.S. section 394.463 re quires Baker Act receiving facilities to document each involuntary examination to the Agency for Health Care Administration (AHCA) within one business day of the examination. Copies of each involuntary examination initiation form (e.g., reports of law e nforcement officers, certificates of mental health professionals, and court issued ex-pa rte orders) (Appendix A) are sent to the Baker Act Reporting Center at FMHI by way of an agreement with the AHCA. A cover sheet which includes demographic information and the name of the provider of services is attached to the involuntary examination initiation form (Appendix B). The Policy and Services Research Data Center at FMHI se rves as a repository for the initiation forms. Data from the involuntary examina tion initiation forms are entered into a database that maintains information pertaining to e ach examination at a Baker Act receiving facility. Data Retrieval Procedures Approval was obtained from the University of South Florida Institutional Review Board to conduct the study. Confidentiality was mai ntained as follows: Baker Act data are stored on a secured server that is protected by a firewall. User IDs and passwords are

PAGE 88

77 necessary to log onto the server. Baker Act data a re contained in a folder on the server to which only those who are given permissions may acce ss. Baker Act data on paper forms are stored in a secured room that can be accessed w ith a master key and to whom only those involved in the use of these data are given k eys. Data from the Baker Act Reporting Center were tabulated as aggregate number s. There were no personal identifiers used in the tabulation and transmission of these data. Data from the Florida Department of Education for individual students are stored in a data base in Tallahassee, Florida. The data that was used for this study are accessible to the general public through the Florida Department of Education website (http:/ /www.fldoe.org). These data are reported as aggregate numbers and there are no pers onal identifiers associated with the data. Paper copies of the survey were stored in a locked file cabinet in the school psychology program at the Department of Psychologic al and Social Foundations, College of Education, University of South Florida, to which only those involved in the study had access. Steps in data retrieval. Data retrieval occurred through the following stepwise procedure. Step One: Personnel who had access to the Baker Ac t Reporting Center Data based at the Florida Mental Health Institute tabula ted aggregate numbers for the relevant demographic categories and exam counts by county. T hese data were stored and displayed in Microsoft excel spreadsheets. Personne l at the Baker Act Reporting Center provided the Microsoft excel spreadsheets to the pr imary investigator upon completion of the relevant data tabulations. The data were displa yed as aggregate numbers with no personal identifiers displayed.

PAGE 89

78 Step Two: A separate data base for salient distric t level variables was developed for the 67 school districts. The following data was obtained from Florida Department of Education District website (www.fldoe.org): (a) tot al student enrollment, (b) percent of students who belong to an ethnic minority category, (c) percent of African American students, (d) percent of students who received free /reduced lunch, (e) percent of students who received special education services, (f) percen t of student who received special education services under the Emotionally Handicappe d (EH) / Severe Emotional Disturbance (SED) category, (g) percent of student who were suspended, (h) percent of students who were expelled, (i) percent of students who were retained, (j) percent of students who graduated from high school with a stan dard diploma, (k) the percent of students who obtained a Level 3 or higher on the re ading and math sections of the Florida Comprehensive Achievement Test (FCAT), and (l) the combined number of school social workers, guidance counselors, and school psychologi sts in the district. These variables were coded according to the criteria specified in A ppendix C. Step Four: An estimate of the per capita Baker Act ER rate (PerBA) for each district was calculated by dividing the total numbe r of Baker Act ER’s within each school district by the total student enrollment within eac h district. An estimate of the per capita repeat Baker Act ER rate (exams representing the 2nd or more exams for a child) (PerRep) was calculated by diving the number of rep eat Baker Act ER’s by the total student enrollment within each district. These data were calculated in this manner to provide an estimate of the use of the Baker Act amo ng school age children in each county.

PAGE 90

79 Step Five: District level data from the Florida De partment of Education were transferred to an electronic database for storage a nd data analysis. These data were initially stored in a Microsoft Excel spreadsheet a nd were later transferred to SPSS Version 15.0 spreadsheet for storage and data analy sis. A data coder from the University of South Florida, School Psychology Program was tra ined in the data transfer and entry process by the primary investigator and was used to assist in integrity checks. The accuracy of calculations and transfers for each of the independent and dependent variables was measured for 20% of the transfers. A data sheet for data transfer and procedural integrity checks was used for this proce ss (Appendix D). The procedural integrity checks were conducted according to the fo llowing steps. First, the random numbers function in SPSS-15.0 was used to generate 13 numbers between 1 and 67. The numbers that were selected corresponded to one of t he 67 districts. These districts were selected for data transfer and data integrity check s. Procedural integrity checks also occurred for the calculation and transfer of data t hat was used to generate demographic comparisons. Data analysis The procedures that were used to analyze the data a re presented in this section. Chi-Square analyses, odds ratios, and multiple regr ession analyses procedures were used to answer the research questions. To answer research question 1, data analyses for de mographic categories included two levels of comparison. First, the total number o f Baker Act ERs that were initiated for grade level, gender, and race/ethnicity were examin ed. These calculations included initial and repeat Baker Act ERs. Second, the distribution of children who received a Baker Act

PAGE 91

80 ER in 2005 was reported for each salient demographi c category. These data were reported as the number and percentage of children w ho received a Baker Act ER for each demographic category. These analyses included only the number children and adolescents who received one Baker Act ER during 20 05, and excluded repeat Baker Act ER. Odds ratios were calculated for each demographi c category. Chi-Square analyses were conducted to determine if there was a signific ant difference between the proportion of children and adolescents in each demographic cat egory who received a Baker Act ER and statewide membership in each demographic catego ry. Odds ratios also were calculated to compare the odds of receiving a Baker Act ER among children and adolescents who received a Baker Act ER. The SAS statistical software package was utilized when performing Chi-square analyses, the c alculation of odds ratios, and the calculation of odds ratio confidence intervals. For purposes of grade level analyses, elementary sc hool children and adolescents were considered to be in the age range of 5-11.99 y ears old, middle school children and adolescents were 12-14.99 years old, and high schoo l children were 15-18.99 years old. These age ranges were entered into a database formu la that generated the total number of Baker Act ERs that were initiated during 2005, and the number of children and adolescents who received at least one Baker Act ER during the 2005 school year for the three grade levels. Baker Act ER counts were disagg regated according to these age ranges. For grade level comparisons of total Baker Act ERs, 21, 965 Baker Act ERs were used in these analyses. Total exam counts included repeat Baker Act ER referrals. The total number of children and adolescents who receiv ed a Baker Act ER was tabulated for purposes of odds calculations. There were a total o f 14,039 children and adolescents who

PAGE 92

81 received a Baker Act ER during the 2005 calendar ye ar. A total of 14,201 Baker Act ERs were used for the grade level analyses because 162 or 1.1% of the actual number of children who received a Baker Act ER had their exam s counted in more than one grade level. Duplicate/repeat Baker Act ER were excluded from these analyses. To calculate odds for each grade level, enrollment calculations were based on Fall 2005 K-12 membership in Florida Public Schools. These figures included Pre-K and laboratory school enrollment. The total Fall 2005 enrollment w as 2,673,563 for elementary, middle, and high school age students. A total of 20,406 or 92.9% of the total 21,965 Bake r Act ERs initiated during 2005 were used for gender comparisons. Gender was u nknown for 1,559 Baker Act ERs. These figures included repeat Baker Act ERs. The to tal number of male and female children and adolescents who received one Baker Act ER was used to calculate odds ratios. These numbers did not include duplicate or repeat Baker Act ERs among these children and adolescents. A total of 13,558 Baker A ct ERs were used for gender comparisons. This figure represented 96.5% of the t otal number of children and adolescents who received a Baker Act ER during 2005 There were 481 Baker Act ERs that were excluded from these analyses because the reporting form did not report a social security number or gender for the child. To calcula te odds among males and females, enrollment calculations were based on Fall 2005 mem bership in Florida Public Schools including laboratory schools and excluding pre-k me mbership. The total male and female enrollment that was used for these comparisons was 2,626,535. Pre-k enrollment that was excluded from these comparisons was 47,028, which r epresented 1.8% of the total Florida public school enrollment.

PAGE 93

82 A total of 19,818 or 90.23% of the total 21,965 Bak er Act ERs were used for race/ethnicity comparisons. These figures included repeat Baker Act ERs. Race/ethnicity was not reported for 2,147 Baker Act ERs. The total number of children and adolescents in each race/ethnicity category who received at lea st one Baker Act ER was used to calculate odds ratios. These figures did not includ e duplicate or repeat Baker Act ERs among these children. A total of 13,452 Baker Act E Rs were used for race/ethnicity comparisons. This figure represented 95.8% of the t otal number of children and adolescents who received a Baker Act ER during 2005 There were 587 Baker Act ERs that were excluded from these analyses because the reporting form was missing a county of residence and/or race of the child. To calculate odds for each grade level, enrollment calculations were based on Fall 2005 K-12 membershi p in Florida Public Schools. These figures included Pre-K and laboratory school enroll ment. The total Fall 2005 enrollment was 2,673,563 students. The distribution of the population of children and adolescents who were African American and received a Baker Act ER in 2005 was ca lculated to answer research question 2. These were reported as percentages of A frican American children and adolescents who received a Baker Act ER. Chi-Square analyses were conducted to determine if there was statistically significant di fference between the proportion of African American children and adolescents who recei ved a Baker Act ER and total student enrollment of African American children and adolescents. To answer research question 3, the number of Baker Act ERs during 2005 that represented a child’s second (or greater) referral was calculated. These data were reported

PAGE 94

83 as the number and percentage of the total Baker Act examination referrals that were repeat Baker Act examination referrals during 2005. To answer research question 4, the number of childr en and adolescents who received more than one Baker Act ER during 2005 was calculated. These data were reported as the percentage of children and adolesce nts that received more than one Baker Act ER. Multiple regression analyses were utilized to answe r research question 5. The multiple regression analyses allowed for the predic tion of the dependent variable (an estimate of the percent of students in a district w ho received a Baker Act ER) from the set of predictor variables. The following independent v ariables were excluded from all analyses: (a) the types of mental health services p rovided to students within a school district, and (b) the methods used to monitor the p rogress of students who return to school following a Baker Act ER. An estimate of the per ca pita Baker Act ER and per capita repeat Baker Act ER rate for each district was regr essed on the remaining set of independent variables. In addition to calculating t he regression equation for each research question, the percent of explained variance was obt ained for each research question. The Statistical Package for the Social Sciences-Version 15.0 (SPSS-15.0) computer software program was utilized to conduct the necessary stati stical analyses for question 5. The assumptions for a valid multiple regression analysi s were examined before proceeding with development of the regression models associate d with the research questions.

PAGE 95

84 Chapter Four Results The purpose of this chapter is to provide the resul ts of data analyses that were conducted to answer the research questions. First, the results of data analyses for comparisons of Baker Act examination referrals (ER) among age/grade level, gender, and race/ethnicity are presented. Second, data from mul tiple regression analyses are discussed. To prevent redundancy and maximize cohes iveness, a summary of the descriptive statistics and assumptions of multiple regression analyses are provided separate from the discussion of findings from indiv idual multiple regression analyses. The chapter concludes by providing the statistics f or the each the multiple regression equations. Table 1 displays the acronyms used to id entify the relevant terms that are discussed throughout the chapter.

PAGE 96

85 Table 1 Abbreviations for Relevant Terminology Acronym Definition Baker Act ER Baker Act Examination Referral PerBA an estimate of the Baker Act ER rate for ea ch county PerRep an estimate of the repeat Baker Act ER rate for each county enroll total student enrollment within a dist rict PerMinority percent of the student population that belongs to a racial/ethnic minority PerAfricanAM percent of the student population that is African American PerFree percent of the student population that receives free and/or reduced lunch PerSPED percent of the student population that received special education services PerEHSED percent of the student population tha t received special education services under the Emotional ly Handicapped (EH) and Severe Emotional Disturbance (SED) categories PerRet percent of the student population that was retained one grade level PerGrad percent of the student population that received a 4year standard diploma Per3read percent of the student population that received a level 3 or higher on the reading section of th e Florida Comprehensive Achievement Test (FCAT)

PAGE 97

86 Table 1 (continued) Abbreviations for Relevant Terminology Acronyms Definition Per3math percent of the student population that received a level 3 or higher on the math section of the Florida Comp rehensive Achievement Test (FCAT) Per3comp percent of the student population that received a level 3 or higher on the read and math section of the Flo rida Comprehensive Achievement Test (FCAT)-composite variable PerISS percent of the student population that re ceived an In-school suspension PerOSS percent of the student population the rece ived an Out-ofschool suspension PerExp percent of the student population that re ceived an expulsion RatioMHStud ratio of mental health professionals (guidance counselors, social workers, school psychologists) to studen ts within the district Note. See Appendix C for descriptions of the formulas tha t were used to calculate each variable. Data transfer and procedural integrity Data transfer and integrity checks were performed o n 20% of all district level database transfers and 100% of transfers related to the calculation of demographic data (i.e., age/grade level, gender, and race/ethnicity) For the district level analysis, initial integrity checks revealed that 218 of the 221 (98.6 %) district level calculations and

PAGE 98

87 transfers were accurate. The three inaccurate data points were corrected before proceeding with final data analyses procedures. For the calculation of demographic variables, 20 of the 20 (100%) transfers were accur ate. It was appropriate to proceed with final data analyses procedures because these data i ndicated adequate procedural integrity. Research Question 1: What is the distribution of demographic variables ( e.g., age, gender, and race/ethnicity) for children who receiv ed Baker Act examination referrals during 2005? Age. Age/grade level was related to the use of the Baker Act among children and adolescents. Baker Act ERs were more prevalent among the high school population. Table 2 displays the distribution of total Baker Ac t ERs for each grade level. Analysis of these data indicates that the greatest number of Ba ker Act ERs occurred at the high school grade level. High School students had the greatest odds of recei ving a Baker Act ER, followed by middle school students, and elementary school st udents. The data suggest that the odds of receiving a Baker Act ER increase as studen ts become older and progress to higher grade levels. Odds were calculated by dividi ng the probability of an event occurring by the probability of an event not occurr ing. Odds ratios also were calculated to compare the odds of receiving a Baker Act ER betwee n the grade levels. Table 3 displays the odds of receiving a Baker Act ER for each grade Table 4 displays the odds ratios for comparisons between grade levels. Figure 1 display s a histogram of the odds among children in each grade level.

PAGE 99

88 Table 2 Distribution of Baker Act Examination Referrals (ER ) By Grade Level Elementarya Middleb High Schoolc Total Nd 2,990 6,672 12,303 21,965 Percentage of Total Exams 13.6 30.4 56.0 100.0 Note. aElementary school children were 5-11.99 years old. bMiddle school children were 12-14.99 year old. cHigh school children were 15-18.99 years old. dThese figures include initial and repeat Baker Act ERs.

PAGE 100

89 Table 3 Odds among Grade Levels for Children and Adolescent s Who Received Baker Act Examination Referrals (ERs) Received BA Total Enrolld Odds Elementarya 1,925 1,260,528 0.00153 Middleb 3,994 611,749 0.00653 Highc 8,282 801,286 0.01034 Note. aElementary school children were 5-11.99 years old. bMiddle school children were 12-14.99 years old. cHigh school children were 15-18.99 years old. dEnrollment calculations based on Fall 2005 PreK-12 membership in Florida Public Schools, including enrollment in laboratory schools. Table 4 Odds Ratios Comparisons between Grade Levels for Ch ildren and adolescents Who Received Baker Act Examination Referrals (ERs) Comparison Odds Ratio 95% Confidence Int erval High vs. Elementary 6.768 6.440 – 7.113 High vs. Middle 1.583 1.524 – 1.644 Middle vs. Elementary 4.275 4.048 – 4.514

PAGE 101

90 Figure 1 Odds of Receiving a Baker Act Examination Referrals (ERs) among Age/Grade Level 0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 ElementaryMiddleHigh Odds N = 1925N = 3,994 N = 8,282 Note. aElementary school children were 5-11.99 years old. bMiddle school children were 12-14.99 years old. cHigh school children were 15-18.99 years old. dEnrollment calculations based on Fall 2005 K-12 membership in Florida Public Schools, excluding membership in lab schools, virtual schools, and pre -k enrollment. A Chi-square analysis was conducted to compare the proportion of Baker Act ERs that were initiated within each grade level to grad e level enrollment proportions in Florida public schools. Results indicate that Baker Act ER rates differed among children and adolescents in each grade level (2 (2, N = 2,673,563) = 7416.9836, p < .0001.

PAGE 102

91 Table 5 Chi-Square Statistics for Comparison of Baker Act E xamination Referral (ER) Proportions between Grade Levels Statistic DF Value Probability Chi-square 2 7416.9836 <0.0001 Phi Coefficient 0.0527 Note. N = 2,673,563 Prek-12 Enrollment. The value of the Phi Coefficient suggests a weak positive relationship between age/grade level and Baker Act ERs. Examination of t he expected and actual frequencies of Baker Act ERs within these populations suggests that Baker Act ER proportions were greater than expected among the high school populat ion in comparison to the proportion of enrolled high school students. Baker Act ER prop ortions were less than expected among elementary school students. Baker Act ER prop ortions among middle school students were consistent with the proportion of mid dle school student enrollment in Florida’s public schools. Table 5 displays the ChiSquare statistics associated with the chi-square analysis. Gender. A greater number of Baker Act ERs were initiated a mong females than males. Table 6 displays the proportion of total Bak er Act ERs that were initiated for males and females. The odds of receiving a Baker A ct ER were greater among females than males. Tables 7 and 8 display the odds and odd s ratios for the comparison of Baker

PAGE 103

92 Act ER proportions among males and females. Figure 2 displays a histogram of the odds of receiving a Baker Act ER among males and females Table 6 Distribution of Baker Act Examination Referrals (ER s) among Males and Females Males Females Total Na 10, 139 10,267 20,406 Percentage of Total Exams 49.6 50.3 Na = Gender was known for 20,406 (92.9%) of the total 21,965 exams. Total exam counts include repeat Baker Act exams.

PAGE 104

93 Table 7 Odds among Males and Female Children and Adolescent s Who Received Baker Act Examination Referrals (ERs) Received BAa Total Enrollb Odds Male 6,721 1,346,704 .00501 Female 6,837 1,279,831 .00537 Note. aBaker Act ER numbers do not include 481 referrals b ecause a social security number was not listed on receiving forms. bEnrollment calculations based on Fall 2005 membership in Florida Public Schools, excluding pre -k membership. Table 8 Odds Ratio Comparison for Males and Female Children and Adolescents Who Received Baker Act Examination Referrals (ERs) Comparison Odds Ratio 95% Confidence I nterval Females vs. Male 1.071 1.035 – 1.1 08

PAGE 105

94 Figure 2 Odds of Receiving a Baker Act Examination Referral (ER) Among Males and Females Children and Adolescents 0 0.001 0.002 0.003 0.004 0.005 0.006 MalesFemales Gender Odds N = 6,721 N = 6,837 Note. Baker Act ER numbers do not include 481 referrals t hat did not include a social security number. Enrollment calculations based on F all 2005 membership in Florida Public Schools, excluding pre-k membership. A Chi-square analysis was conducted to compare Bake r Act ER proportions among males and females to enrollment proportions i n Florida public schools. The proportion of males and females who received a Bake r Act ER differed from enrollment proportions (2 (1, N = 2,626,535) = 15.7798, p < .0001. The value for the Phi Coefficient suggests a weak positive association be tween gender and Baker Act ERs.

PAGE 106

95 Table 9 Chi-Square Statistics for Comparison of Baker Act E xamination Referral (ER) Proportions between Males and Female Statistic DF Value Probability Chi-square 1 15.7798 <0.0001 Phi Coefficient 0.0025 Note. N = 2,626,535 K-12 population. Excludes pre-k member ship of 47,028. Examination of the expected and actual frequencies of Baker Act ERs between males and females suggests that Baker Act ER proportions were greater than expected among females in comparison to their enrollment proportio ns. Table 9 displays the Chi-Square statistics associated with the comparison of propor tions. Race/Ethnicity. Race/ethnicity was associated with the use of the Baker Act among children and adolescents. The greatest number of Baker Act ERs was initiated for White children and adolescents. Table 10 displays t he total number of Baker Act ERs that were initiated for each race/ethnicity category. Ho wever, children and adolescents who belonged to the Other/Mixed category had the greate st odds of receiving a Baker Act ER. Children and adolescents with membership in the Oth er/Mixed category had a 9.071 greater odds of receiving a Baker Act ER than Hispa nic children and adolescents, 8.32 greater odds of receiving a Baker Act ER than Asian children and adolescents, 1.52 greater odds of receiving a Baker Act ER than Afric an American children and

PAGE 107

96 Table 10 Distribution of Baker Act Examination Referrals (ER s) for Race/Ethnicity Na Percentage of Total Exams White 13,279 67.0% African American 4,789 24.2% Hispanic 751 3.8% Asian/Pacific Islander 74 0.4% Other/Multiracial 925 4.7% Total Exams 19,818 100.0% Total Minority 6,539 33.0% Note. aRace/ethnicity was known for 19,818 (90.23%) of the total 21,956 Baker Act ERs that were initiated during the 2005 calendar year. adolescents and a 1.07 greater odds of receiving a Baker Act ER than White children and adolescents. White children and adolescents had the second highest odds of receiving a Baker Act ER. White children and adolescents had a 1.41 greater odds of receiving a Baker Act ER than African American children and ado lescents, 8.44 greater odds of receiving a Baker Act ER than Hispanic children and adolescents, and a 7.751 greater odds of receiving a Baker Act ER than Asian childre n and adolescents. White children and adolescents also had a 2.26 greater odds of rec eiving a Baker Act ER than all

PAGE 108

97 children and adolescents with membership in a Minor ity demographic category. African American children and adolescents had the third hig hest odds of receiving a Baker Act ER. Tables 11 and 12, and Figure 3 display the odds of receiving a Baker Act ER for each race/ethnicity category and odds comparisons b etween these categories. Table 11 Odds among Race/Ethnicity Categories for Children a nd Adolescents Who Received Baker Examination Referrals (ERs) Demographic Category Received BAa Total Enroll b Odds Other/MultiracialC 667 86,090 .00775 White 9,089 1,267,489 .00717 African American 3,114 621,506 .00501 Hispanic 527 624,372 .00084 Asian/Pacific Islander 55 59,367 .00 093 Total Minority 4,363 1,391,335 .00314 Note. aCalculations do not include 587 Baker Act ERs or 4. 2% of the total count because the referrals were missing a county of resi dence and/or race was not reported on the receiving form. bEnrollment calculations based on PreK-12 Fall 2005 membership in Florida Public Schools. CIncludes children who are described as belonging to Native American/Alaskan Native and the Multiracial categor ies.

PAGE 109

98 Table 12 Odds Ratio Comparisons between Children and Adolesc ents from Different Race/Ethnicity Backgrounds Comparison Odds Ratio 95% Confidence Interva l Other vs. White 1.065 0.984 – 1.152 Other vs. African American 1.524 1.401 – 1.657 Other vs. Asian 8.240 6.258 – 10.850 Other vs. Hispanic 9.044 8.066 – 10. 141 White vs. African American 1.431 1.374 – 1.491 White vs. Asian 7.740 5.937 – 10.091 White vs. Hispanic 8.496 7.781 – 9.2 76 African American vs. Asian 5.408 4.1 42 – 7.062 African American vs. Hispanic 5.936 5. 412 – 6.511 Asian vs. Hispanic 1.089 0.825 – 1.4 38 White vs. Minority 2.289 2.208 – 2.3 73 Note. aCalculations do not include 587 Baker Act ERs or 4. 2% of the total count because the referrals were missing a county of resi dence and/or race was not indicated. bEnrollment calculations based on PreK-12 Fall 2005 membership in Florida Public Schools. CIncludes children who are described as belonging to Native American/Alaskan Native and the Multiracial categories.

PAGE 110

99 Figure 3 Odds of Receiving a Baker Act Examination Referral (ER) Among Race/ethnicity categories 0 0.001 0.002 0.003 0.004 0.005 0.006 0.007 0.008 0.009W ht e African Am e rican Hi sp an i c Asian / Pac I sl a n der O ther/Mult i racial M in o ri t y Race/Ethnicity CategoryODDS N = 9,089 N = 3,114N = 527 N = 55N = 667 N = 4,363 Note. Calculations do not include 587 Baker Act ERs or 4. 2% of the total count because the referrals were missing a county of residence an d/or race was not indicated on the referral. Enrollment calculations based on Fall 200 5 membership in Florida Public Schools, excluding pre-k membership. Includes child ren who are described as belonging to Native American/Alaskan Native and the Multiraci al categories. A chi-square analysis was conducted to compare the proportions of Baker Act ER initiated for children and adolescents who belonged to each race/ethnicity demographic

PAGE 111

100 Table 13 Chi-Square Statistics for Comparison of Baker Act E xamination Referral (ER) Proportions between Race/ethnicity Demographic Cate gories Statistic DF Value Probability Chi-square 4 3615.8239 <0.0001 Phi Coefficient 0.0368 Note. N = 2,673,563 in Prek-12 population. category to the proportion of enrolled students for each race/ethnicity demographic category. Baker Act ERs proportions among race/eth nicity demographics differed from the proportions of enrolled students (2 (4, N = 2,673,563) =3615.8239, p < .0001. Examination of the expected and actual frequencies of Baker Act ER proportions suggests that Baker Act ERs were greater than expec ted among White children and adolescents in comparison to their enrollment propo rtions. Table 13 displays the ChiSquare statistics associated with the comparison of proportions. Research Question 2: Do African American students receive Baker Act exam ination referrals at rates that are disproportionate to the ir total enrollment within the Florida public schools? African American students were not overrepresented in the population of children and adolescents who received a Baker Act ER compare d to their enrollment in Florida public schools. African American children and adole scents accounted for 23.15% of all children and adolescents who received a Baker Act E R. African American children and

PAGE 112

101 adolescents accounted for 23.36% of enrolled PreK-1 2 students in Florida Public schools. Table 14 displays the proportion comparisons betwee n African American children and adolescents who received a Baker Act ER to enrollme nt proportions. Table 14 Comparison of Proportion of African American Childr en and Adolescents Who Received a Baker Act Examination Referrals to Proportion of Enrollment Within Florida Public Schools # of Children % of Total Children #Enrolleda %Enrolled Who Received B.A Who Received B.A N 3,114 23% 609,152 23% Note. aEnrollment calculations based on Fall 2005 membersh ip in Florida Public Schools, excluding pre-k membership.

PAGE 113

102 Table 15 Summary of Relevant Findings for Demographic Compar isons Age/Grade Level o The odds of receiving a Baker Act ER increases amon g older children and adolescents in higher grades. o The odds of receiving a Baker Act ER are greatest a mong the high school population. o Middle school students have greater odds of receivi ng a Baker Act ER compared to elementary school children. o Baker Act ER rates were greater than expected among the high school population and less than expected among the element ary school population when taking into account the proportion of enrolled students in these grades. Gender o The odds of receiving a Baker Act ER are greater am ong females than males. Race/Ethnicity o The odds of receiving a Baker Act ER are greatest a mong children and adolescents who belong to the Other/Mixed category and White Children and adolescents. o The odds of receiving a Baker Act ER are lowest amo ng Hispanic children and adolescents. o The odds are greater among White children and adole scents than all minority children and adolescents. o Baker Act ER rates among Black children are proport ional to the rates of enrollment among Black children in the Florida publ ic schools. ___________________________________________________ _____________________

PAGE 114

103 Research Question 3 and 4: What percentage of total Baker Act examination referrals are repeat Baker Act examination referrals? What percentage of children and adolescents received more than one Baker Act examin ation referral? A total of 18.3% of the children and adolescents wh o received a Baker Act referral accounted received more than one referral. A total of 37.91% of the total Baker Act ERs were repeat Baker Act ERs. During the calendar year of 2005, there were 14,200 children and adolescents who received a Baker Act ER. Of those, 2,578 children a nd adolescents were repeat referrals in 2005. There were a total of 161 children and ado lescents who had exams that were counted in more than one county. Consequently, 14,0 39 were used for this analysis. For the repeating Baker Act ER analysis, 18, 459 were u sed in the analysis. Of those, 6,998 were repeat Baker Act ERs. Table 16 displays the to tal number of Baker Act ERs, the total number of repeat Baker Act ERs, the total num ber of children and adolescents who received a Baker Act ER, and the percentage of chil dren and adolescents who received more than one Baker Act ER.

PAGE 115

104 Table 16 Baker Act Examination Referrals (ERs) Total B.A.a Total Repeat % Repeat Total Repeat % Repeat Childrenb Children N 18,459 6,998 37.91% 14,039 2,578 18.36 Note. aThere were a total of 21, 965 Baker Act ER’s during 2005. The number used in the repeating analysis is lower because of missing county of residence or Social Security numbers. bThere were 14,043 children who were documented as r eceiving at least on Baker Act ER during 2005. The number used in the re peating analysis is lower because of children who had exams counted in more than one county. Research Questions 5: What is the relationship betw een the district level variables and Baker Act ERs and repeat Baker Act ERs? District level variables accounted for a statistic ally significant proportion of the variance for three of the four multiple regression analyses that were examined. Further, there were several district variables (e.g., gradua tion rates, minority enrollment, out-ofschool suspension) that accounted for a unique prop ortion of the variance in Baker Act ERs and repeat Baker Act ERs after controlling for the variance attributed to other district variables. The remainder of this chapter will discu ss the results of multiple regression analyses that were utilized to examine these relati onships.

PAGE 116

105 Descriptive Statistics. Examination of the skewness values for the dependen t and independent variables indicates extreme positiv e skewness for enrollment and percent of the student population who received spec ial education service under the EHSED category, and moderate positive skewness for Baker Act ERs, percent of students who received special education services, percent re tained, and percent expelled. The remaining skewness values range from slight skewnes s to normal. Examination of the kurtosis values suggests a leptokurtic distribution for enrollment, Baker Act ERs, percent EHSED, percent retained, percent out-of-school susp ension, and percent expelled. The kurtosis values indicate normal kurtosis values for the remaining variables. Table 17 displays the descriptive statistics for each of the variables associated with research questions 5 and 6.

PAGE 117

106 Table 17 Descriptive Statistics for the Dependent and Indepe ndent Variables N Mean Median SD Skewness Kurtosis PerBA 67 0.007 0.006 0.003 1.12 3 2.158 PerRep 67 0.001 0.001 0.001 0.844 0.080 enrollment 67 39,741 12,274 65,352 3. 040 10.552 PerMinority 67 0.365 0.332 0.197 0 .836 0.416 PerFree/reduced 67 0.491 0.496 0.112 -0.051 -0.419 PerSPED 67 0.172 0.165 0.035 1.071 1.520 PerEH/SED 67 0.016 0.014 0.009 2.536 10.416 PerRetained 67 0.075 0.073 0.027 1.034 2.421 PerGrad 67 0.750 0.751 0.086 -0.316 0.880 Per3read 67 0.525 0.530 0.078 -0. 382 0.597 Per3math 67 0.571 0.580 0.089 -0. 517 0.672 Per3Comp 67 0.548 0.560 0.083 -0. 444 0.641 PerOSS 67 0.105 0.102 0.049 0.8 88 2.539 PerISS 67 0.125 0.131 0.077 0.403 -0.601 PerExp 67 0.0008 0.0001 0.001 1. 972 3.050 ratioMH/Stud 67 0.002 0.002 0.001 -0.693 1.503

PAGE 118

107 Table 18 displays the correlation matrix for the da ta set. The matrix displays the dependent variables and each of the independent var iables that were entered into the regression equations for research questions 5. Exam ination of the zero order correlations between percent of students who received a Baker Ac t ER and the independent variables indicates that the highest zero order correlation w as between percent of students who received a Baker Act ER and percent free and reduce d lunch ( r = -.234). The lowest zero order correlation was between Baker Act ERs and the percent of students in a district who were expelled (r = -.018). Further, analysis of the zero order correlations indicates that the only significant correlation was between B aker Act ERs and percent free and reduced lunch ( r = -.234, p = .028). Examination of the zero order correlation s between repeat Baker Act ERs indicates that the highest zer o order correlation was between repeat Baker Act ERs and enrollment ( r = .294). Additional analysis indicates significant zero order correlations between repeat Baker Act ERs and enroll ( r = .294, p = 0.008), percent free and reduced lunch ( r = -.233, p = 0.029), percent graduation ( r = -.244, p = .023), and the ratio of mental health professional to stud ents ( r = .221, p = .036).

PAGE 119

108 Table 18 Correlation Matrix for Dependent and Independent Va riables 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 1. PerBA 1.00 2. PerRep .748 1.00 3. enrollb .140 .296 1.00 4. PerMinority -.126 .058 .491 1.00 5. PerAfricanAm -.085 -.091 .161 .704 1.00 6. PerFreeab -.234*-.233**-.131 .406 .443 1.00 7. PerSPED -.112 -.187 -.390 -.121 .19 0 .261 1.00 8. PerEHSED -.150 -.168 -.195 .021 .265 .210 .650 1.00 9. PerRet -.022 -.020 .045 .494 .697 .522 .323 .304 1.00 10. PerGradb -.157 -.244* -.208 -.636 -.503 -.540 .119 .009 -.497 1.00 11. Per3read .132 .058 .026 -.597 -.57 4 -.824 -.188 -.213 -.650 .722 1.00 12. Per3math .096 .075 .099 -.518 -.58 9 -.809 -.251 -.278 -.684 .650 .722 1.00 13. Per3comp .114 .068 .065 -.561 -.58 8 -.825 -.224 -.250 -.675 .691 .988 .991 1.00 14. PerOSS .068 .120 -.165 .354 .574 .354 .429 .249 .327 .539 .384 -.5 51 -.615 1.00 15. PerISS .044 .027 -.174 -.044 .088 .127 .060 .079 .147 -.119 -.241 -.276 .263 .307 1.00 16. PerExp -.018 .006 -.197 .230 .365 .337 .155 .238 .451 -.349 -.426 -.4 21 -.428 .459 .085 1.00 17. RatioMHStudb .188 .221* .094 .061 .110 -.020 -.073 .073 .068 -.173 -.060 -.044 -.052 .126 .036 .121 1.0 0 Note. aStatistically significant correlation with PerBA, bStatistically significant correlation with PerRep; p <.05; **p<.01

PAGE 120

109 Collinearity. Analysis of the zero order correlations indicates s ignificant collinearity between the following independent vari ables: (a) percent of students who received a level 3 or higher on the reading FCAT an d (b) percent of students who received a level 3 or higher on the math section of the FCAT ( r = 0. 957). Stevens (2002) suggest that variables with a Variance Inflation Fa ctor (VIF) of greater than 10 have a significant influence on the R2 value for regression equations that include these variables. The VIF for percent 3 or higher on the reading FCAT and percent 3 or higher on the math FCAT were greater than 10 for each multiple regress ion analysis that included these variables, indicating that these variables had a si gnificant influence on the R2 value when entered into the regression equation. An achievemen t composite variable was calculated to address the collinearity between two variables. The achievement composite variable was calculated b y adding the values for the percent of children and adolescents who obtained a level 3 or higher on the reading FCAT and the percent of children and adolescents wh o obtained a level 3 or higher on the math FCAT, and then dividing by 2. For purposes of this study, the achievement composite variable is labeled as the percent of stu dents in a district who achieved level 3 or higher on the reading and math sections of the F CAT (Per3comp). The achievement composite variable was calculated for each district and entered into the regression equation to replace the Per3read and Per3math varia bles. The equation to calculate this composite variable is presented below: Per 3 Comp = Per3read + Per3math 2

PAGE 121

110 Assumptions of multiple regression. An appropriate analysis of multiple regression equations is dependent primarily on two assumptions. First, a valid and reliable analysis of data from multiple regression is dependent on a linear relationship between the dependent and independent variables. Sc atterplots with the predicted value plotted on the x-axis and the standardized residual s plotted on the y-axis were examined for each regression equation that was conducted. Th e scatterplots indicated a linear relationship for each of the regression equations. Thus, the assumption of linearity was confirmed. Multiple regression analysis also is dependent on the homoscedasticity of errors assumption. Homoscedasticity refers to the equal va riance of prediction errors (residuals) around the regression line. A homoscedastic distrib ution of errors indicates that the regression equation predicts equally well at each p oint on the y and x-axis. Scatterplots with the residuals plotted on the y-axis and the pr edicted values plotted on the x-axis were examined for each regression equation. Analysi s of the scatterplots indicated a homoscedastic distribution of errors around the reg ression line for each equation. Thus, the homoscedasticity of errors assumptions was conf irmed. An example of a scatterplot displaying homoscedasticity for one multiple regres sion is displayed in Appendix E. Outliers. Outliers in the data set were examined for each reg ression equation. Studentized residuals were examined to id entify extreme values that could have a disproportionate affect on the predicted values f or each regression equation. Studentized residuals with absolute values greater than 2.0 were considered significant. There were several variables that had studentized r esidual values of greater than 2.0. Cook’s D value was examined to determine the influe nce that the outliers exerted on the

PAGE 122

111 predicted values for each regression equation. A C ook’s D value of greater than 1.0 is considered to be indicative of a high degree of lev erage on the regression equation. The largest Cook’s D value was .511, indicating that th e identified outliers exerted a minimal effect on the predicted values obtained in each of the regression equations. Therefore, it was appropriate to proceed with the multiple regres sion analysis with these variables included in the analyses. Multiple regression analyses. Multiple regression analyses were conducted for several sets of independent variables to examine th e proportion of explained variance for Baker Act ER rates and repeat Baker Act ER rates. F irst, Baker Act ER rates were regressed on the following set of independent varia bles: enrollment, percent minority enrollment, percent free and reduced lunch, percent of students who received special education services, percent EHSED, percent retained percent graduation, percent 3 or higher on the FCAT, percent out-of-school suspensio n, percent percent expelled, and the ratio of mental health professionals to student. An alysis of the R2 value indicated that the independent variables entered into the equation acc ounted for a significant proportion of the variance in Baker Act ERs ( R2 = .337, F (12, 54) = 2.288, p = .019; Adjusted R2 = .190). The R2 value indicates that approximately 34% of the variance in Baker Act ERs was attributed to the combination of independent va riables in the regression equation. The adjusted R2 value indicates some shrinkage due to the sample si ze and the number of independent variables in the equation. The adjusted R2 indicates that approximately 19% of the variance in Baker Act ERs within the populat ion can be attributed to the independent variables in the data set. Table 19 dis plays the regression statistics associated with the multiple regression equation for this data set.

PAGE 123

112 Table 19 Summary of Multiple Regression Analysis for Selecte d Variables Predicting PerBA (N = 67) ___ Correlations t Zero Order Semi-partial ___ enrollment 0.303 1.890 0.140 0.209 PerMinority -0.561** -2.944 -0.126 -0.326 PerFree -0.385 -1. 871 -0.234 -0.207 PerSPED -0.164 0.964 -0.112 0.107 PerEHSED -0.200 -1.304 -0.150 -0.145 PerRetained 0.038 0.221 -0.022 0.025 PerGrad -0.554** -3.043 -0.157 -0.337 Per3comp 0.016 0.057 0.114 0.006 PerOSS 0.291 1.864 0.068 0.207 PerISS -0.030 -0.242 0.044 -0.027 PerExp -0.023 -0.165 -0.018 -0.018 RatioMHStud 0.082 0.709 0.188 0.079 Note. R2 = 0.337; p = 0.019. Adjusted R2 = 0.190. p < .05; ** p < .01 Analysis of the t-tests for the standardized regres sion coefficients of the independent variables indicates that percent graduation ( = -.554, p = .004) and percent minority ( = -.561, p = .005) accounted for a significant unique proport ion of the variance in Baker

PAGE 124

113 Act ERs after controlling for the variance accounte d for by the other variables in the regression equation. The semi-partial correlation b etween percent graduation, percent minority, and Baker Act ERs was negative. A second regression equation was run with percent A frican American enrollment substituted for percent minority. Analysis of the R2 valued indicated that the independent variables accounted for a insignificant percentage of the variance Baker Act ERs ( R2 = 0.265, F ( 12,54) = 1.622, p = 0.113; Adjusted R2 = .102)). Table 20 displays a summary of the multiple regression analysis for this equati on.

PAGE 125

114 Table 20 Summary of Multiple Regression Analysis for Selecte d Variables Predicting PerBA (N = 67) Correlations t Zero Order Semi-partial enrollment 0.105 0.714 0.140 0.083 PerAfricAm -0.291 -1.587 -0.365 -0.185 PerFree -0.359 -1.660 -0.234 -0.194 PerSPED 0.201 1.124 -0.112 0.131 PerEHSED -0.184 -1. 138 -0.150 -0.133 PerRetained 0.106 0.544 -0.022 0.063 PerGrad -0.457 -2.440 -0.157 -0.285 Per3comp 0.198 0.681 0.114 0.079 PerOSS 0.291 1.708 0.06 8 0.199 PerISS 0.028 0.214 0.044 0.025 PerExp -0.031 -0.21 3 -0.018 -0.025 RatioMHStud 0.121 0.997 0 .188 0.116 Note. R2 = 0.265; p = 0.113. Adjusted R2 = 0.102. A third multiple regression equation was run with r epeat Baker Act ERs as the dependent variable. The following set of independen t variables were included in this analysis: enrollment, percent minority, percent fre e and reduced lunch, percent of

PAGE 126

115 students who received special education services, p ercent EHSED, percent retained, percent graduation, percent level 3 or higher on th e FCAT, percent out-of-school suspension, percent in-school suspension, percent e xpelled, and the ratio of mental health professionals to students. Analysis of the R2 indicated that the independent variables accounted for a significant percentage of the varia nce in repeat Baker Act ERs ( R2 = 0.390, F ( 12,54) = 2.878, p = 0.004; Adjusted R2 = .255). Approximately 39% of the variance in repeat Baker Act ERS could be attribute d to the independent variables in the regression equation. The adjusted R2 indicates some shrinkage, suggesting that 25% of variance in repeat Baker Act ERs within the populat ion can be attributed to the independent variables. Examination of the t-tests o f the standardized regression coefficients for the variables in the equation indi cated that percent graduation ( = -0.532; p = 0.004), percent free and reduced lunch, ( = -0.450; p = 0.27), enrollment ( = 0.390; p = 0.014), percent minority ( = -0.368; p = 0.049), and percent out-of-school suspension ( = 0.347; p = 0.024) accounted for a significant unique proport ion of the variance in repeat Baker Act ERs after controlling for the variance attributed to other variables in the equation. The semi-partial correla tions between percent graduation, percent free and reduced lunch, percent minority, a nd repeat Baker Act ERs were negative. The semi partial correlations between per cent out-of-school suspension, enrollment, and repeat Baker Act ERs were positive. Table 21 displays the statistics for this multiple regression.

PAGE 127

116 Table 21 Summary of Multiple Regression Analysis for Selecte d Variables Predicting PerRep (N = 67) ___ Correlations t Zero Order Semi-partial ___ enrollment 0.390** 2.538 0.296 0.270 PerMinority -0.368* -2.014 0.058 -0.214 PerFree -0.450* -2.281 -0.233 -0.242 PerSPED 0.174 1.068 -0.18 7 0.114 PerEHSED -0.198 -1.346 -0.168 -0.143 PerRetained -0.103 0.627 -0.020 0.067 PerGrad -0.532** -3.045 -0.244 -0.324 Per3comp -0.039 -0.142 0.068 0.015 PerOSS 0.347* 2.315 0.120 0.246 PerISS -0.028 -0.238 0.027 -0.025 PerExp 0.016 0.119 0.006 0.013 RatioMHStud 0.093 0.837 0.221 0.089 Note. R2 = 0.390; p = 0.004; Adjusted R2 = 0.255. p < .05; ** p < .01 A final multiple regression equation was run with r epeat Baker Act ERs as the dependent variable and the following set of indepen dent variables included in this analysis: enrollment, percent African American enro llment, percent free and reduced

PAGE 128

117 lunch, percent of students who received special edu cation services, percent EHSED, percent retained, percent graduation, percent level 3 or higher on the FCAT, percent outof-school suspension, percent in-school suspension, percent expelled, and the ratio of mental health professionals to students. Table 22 d isplays the statistics for this multiple regression. Analysis of the R2 indicated that the independent variables accounted for a significant percentage of the variance in repeat Ba ker Act ERs ( R2 = 0.424, F ( 12,54) = 3.308, p = 001; Adjusted R2 = .296). Approximately 42% of the variance in repe at Baker Act ERs could be attributed to the independent vari ables in the regression equation. The adjusted R2 indicates some shrinkage, suggesting that 29% of va riance of repeat Baker Act ERs within the population can be attributed to the independent variables. Examination of the t-tests of the standardized regr ession coefficients for the variables in the equation indicated that percent graduation ( = -0.494; p = 0.004), percent African American enrollment ( = -0.443; p = 0.009), percent free and reduced lunch ( = -0.434; p = 0.028), percent out-of-school suspension ( = 0.421; p = 0.007), and enrollment ( = 0.305; p = 0.023) accounted for a significant unique propor tion of the variance in repeat Baker Act ERs after controlling for the variance at tributed to other variables in the equation. The semi-partial correlations between per cent graduation, percent African American enrollment, percent free and reduced lunch and repeat Baker Act ERs were negative. The semi-partial correlations between per cent out-of-school suspension, enrollment, and repeat Baker Act ERs were positive.

PAGE 129

118 Table 22 Summary of Multiple Regression Analysis for Selecte d Variables Predicting PerRep (N = 67) ___ Correlations t Zero Order Semi-partial ___ enrollment 0.305* 2.345 0.294 0.236 PerAfricAm -0.443** -2.729 0.058 -0.277 PerFree -0.434* -2.265 -0.233 -0.193 PerSPED 0.203 1.282 -0.18 6 0.130 PerEHSED -0.174 -1.216 -0.168 -0.116 PerRetained 0.047 0.273 -0.020 0.062 PerGrad -0.494** -2.984 -0.244 -0.362 Per3comp 0.062 0.294 0.068 0.155 PerOSS 0.421** 2.790 0.120 0.305 PerISS -0.008 -0.072 0.027 0.019 PerExp 0.010 0.078 0.006 0.010 RatioMHStud 0.120 1.116 0.221 0.097 Note. R2 = 0.424; p = .001. Adjusted R2 = 0.296. p < .05; ** p < .01.

PAGE 130

119 Table 23 Summary of Relevant Findings for Multiple Regressio n Analysis Enrollment, percent minority enrollment, percent fr ee and reduced lunch, percent of students who received special education services percent EHSED, percent retained, percent graduation, percent level 3 or hi gher on the FCAT, percent inschool suspension, percent out-of-school suspension percent expelled, and the ratio of mental health professionals to students ac counted for a statistically significant proportion of the variance in Baker Act ERs and repeat Baker Act ERs. These variables predicted a greater proportion of the variance in repeat Baker Act ERs ( R2 = 0.424; R2 = 0.390) than Baker Act ERs ( R2 = 0.337; R2 = 0.265). When African American enrollment was substituted fo r minority enrollment in regression analyses, the variables did not account for a statistically significant proportion of the variance in Baker Act ERs ( R2 = 0.265). However, regression equations that included African American enrollment accounted for a statistically significant proportion of the variance in repeat Ba ker Act ERs (( R2 = 0.424). Percent free and reduced lunch accounted for a uniq ue proportion of the variance in repeat Baker Act ERs. This relationship was not found for Baker Act ERs.

PAGE 131

120 Table 23 (continued) Summary of Relevant Findings for Multiple Regressio n Analysis Percent graduation was the variable that most consi stently contributed to a unique proportion of the variance in Baker Act ERs and rep eat Baker Act ERs after controlling for the variance attribute to other var iables in multiple regression analyses. The relationship between these percent gr aduation, Baker Act ERs, and repeat Baker Act ERs was negative. Percent out-of-school suspension accounted for a un ique proportion of the variance in repeat Baker Act ERs. The relationship between these variables was positive. Percent out-of-school suspension did not account for a significant proportion of the variance in Baker Act ERs. Other measures of disciplinary exclusion (i.e., percent in-school suspension and p ercent expelled) did not contribute to a unique proportion of the variance i n Baker Act ERs or repeat Baker Act ERs. Enrollment accounted for a unique proportion of the variance in repeat Baker Act ERs. The relationship between these variables was p ositive. It did not account for a unique proportion of the variance in Baker Act ER s. ___________________________________________________ _____________________

PAGE 132

121 Chapter Five Discussion The purpose of this chapter is to provide a summary of the findings for this study, explanations for the findings, limitations, and to describe the practice and research implications. The chapter is organized by first pre senting responses to each research question. A summary of the findings for each resear ch questions is presented. Current findings are compared to literature that has examin ed the relationship between demographic variables, school district variables, a vailability of mental health services, and responses to child and adolescent psychopatholo gy that are similar to the Baker Act. Explanations for the research findings also are dis cussed in the first section. The second section highlights the major design and methodologi cal limitations. The chapter concludes with a discussion about the implications for practice and future directions associated with the findings from this study. Research Questions One and Two The findings from the present study suggest that Ba ker Act ERs are more prevalent among certain demographic groups (e.g., a ge/grade level, gender, race/ethnicity). The odds of receiving a Baker Act ER were greatest among the high school population, females, White children and adol escents, and children and adolescents from the other/mixed demographic. Furthermore, the proportion of children who

PAGE 133

122 received a Baker Act ER were greater than expected among these demographics in comparison to their enrollment proportions within t he Florida public school system. These findings are consistent with other research f indings that have examined the relationship between demographics and responses to psychopathology among the child and adolescent population. Other studies have found that practices resulting in the removal of children and adolescents from mainstream settings (e.g., suspension, expulsion) are used more frequently among older chi ldren and adolescents who are enrolled in secondary education (Florida Department of Education, 2005; RaffaeleMendez, 2000). Furthermore, research (Wu et al., 20 01) suggests that females are more likely than boys to seek mental health treatment fo r depression, and depression represents a condition that is under treated among African Ame rican children and adolescents. Evidence also suggests that African American childr en and adolescents and children from other minority backgrounds are less likely than Whi te children to utilize mental health services (Taylor, 2005; United States Department of Health and Human Services [USDHHS], 2003). However, minority children and ado lescents in a large city were found to be more likely than White children and ado lescents to receive emergency mental health services (Chow, Jafee, & Snowden, 2003). A number of factors could explain the relationship between demographics and the use of the Baker Act. The most parsimonious explana tion is that the incidence of severe psychopathology and impairment is greatest among ch ildren from these demographic groups and may be more likely to display behaviors that warrant a Baker Act ER. For example, research indicates that the prevalence of psychopathology is greater among older children and adolescents (Moffit, 1993; Rober ts, Attkisson, & Rosenblatt, 1998).

PAGE 134

123 Developmental models suggest that more severe forms of psychopathology are related to the time spent on a negative developmental pathway that begins in early childhood (Sroufe, 1997). Minor behavioral problems that emer ge during early childhood can escalate into more severe forms of maladjustment as children experience chronic adaptational failure. The environmental demands ass ociated with middle childhood and adolescence may be more complex and require greater levels of competence to navigate successfully than the demands placed on elementary age children. Adolescence is a developmental period in which the gap between biolo gical and social maturity can have a profound negative impact on social-emotional and be havioral functioning. The behavior of adolescents also can be motivated by attempts to establish independence and autonomy from parents and other adults. Consequently, adoles cents may engage in high risk behaviors (e.g., alcohol use) that further predispo se them to severe psychopathology that culminates in the use of the Baker Act (Moffit, 199 3). Another explanation of the relationship between age and the use of the Baker Act relates to the responses used by adults to address severe psychopathology among older children. School based professionals, family member s, and law enforcement personnel may prefer an approach for younger children that ut ilize resources within less restrictive settings. The Baker Act is most commonly used in re sponse to behaviors that threaten the physical safety of the individual or others within the individual’s environment (Christy, 2003). When adolescents display these behaviors and fail to respond to less intrusive crisis intervention procedure, their size and stren gth may require the involvement of professionals with training in physical restraint a nd seclusion techniques. Baker Act ERs may represent the final response in a chain that be gins when an adolescent fails to

PAGE 135

124 respond to less intrusive crisis intervention proce dures. Assistance from law enforcement personnel may be requested because they are perceiv ed as individuals who have appropriate training in techniques that can decreas e the threat of an older child or adolescent who displays severe psychopathology. Ano ther possible explanation for the relationship between age and use of the Baker Act i s related to language development. Children develop a greater capacity to express nega tive views of themselves, others, and the world as they progress from early childhood to adolescents (Burke, 1998). Older children may have a greater capacity to communicate the intent of future behavior. Consequently, there is a greater likelihood that ad ults will become aware of an older child’s intentions to do harm, and therefore, more likely to initiate a Baker Act ER. Research Questions Three and Four Eighteen percent of children and adolescents receiv ed more than one Baker Act ER during the 2005 calendar year. Additionally, 37% of all Baker Act ERs that were initiated during this time period were repeat refer rals. These findings suggest that a high percentage of children and adolescents continue to demonstrate severe psychopathology after receiving access to emergency mental health s ervices. This would call into question long term outcomes associated with the use of the B aker Act among children and adolescents. Significant risk factors associated w ith severe psychopathology appear to negatively influence the adjustment of children and adolescents after receiving Baker Act ERs. These findings are consistent with research th at has examined the experiences of children and adolescents after receiving intensive mental health services. Many children and adolescents continue to demonstrate a significa nt level of psychopathology after receiving these services (Halliday-Boykins, 2004). Readmission to psychiatric hospitals

PAGE 136

125 is a relatively common outcome for some children an d adolescents who display severe psychopathology (Romansky et al., 2003). Furthermor e, a high percentage of children and adolescents receive several in-school and out-o f-school suspensions while in school (Skiba et al., 1997). Research Question Five District level analysis indicated that the variable s selected for inclusion in multiple regression analysis accounted for a signif icant proportion of the variance in Baker Act ER rates and repeat Baker Act ER rates am ong the 67 school districts in Florida. These findings suggest that the combinatio n of district level demographics (e.g., minority enrollment, socioeconomic status), special education placement rates, academic competence (e.g., reading and math achievement, gra de retention, graduation rates), and responses to severe psychopathology (e.g., suspensi on, expulsion, availability of mental health professional) contributed to the use of the Baker Act among children and adolescents. Certain district level variables contributed to a u nique proportion of the variance in Baker Act ERs and repeat Baker Act ERs. The perc ent of students in a school district that graduated with a standard diploma was the vari able that most consistently contributed to a unique proportion of the variance in Baker Act ER. There was a negative relationship between graduation rates and the use o f the Baker Act, suggesting that there is a greater use of the Baker Act among school dist ricts that have lower graduation rates. These findings are consistent with other research t hat has identified a relationship between rates of disciplinary referrals and high sc hool drop out rates (Christle, Jolivette, and Nelson, 2005), and high school drop among stude nts with emotional and behavioral

PAGE 137

126 disorders (United States Department of Education, 2 001). Further, these findings also are consistent with research that has identified a rela tionship between academic competence and child psychopathology (Hinshaw, 1992; Wagner, K utash, Duchnowski, Epstein, & Sumi, 2005; Willcutt & Pennington, 2000). The negative relationship between academic competen ce (e.g., graduation rates) and the use of the Baker Act can be explained by de velopmental models of psychopathology. Competence refers to the ability t o successfully meet age appropriate development tasks (Masten & Coatsworth, 1998). Deve lopmental models of psychopathology suggest that failure in one area of competence often leads to failure in another area of competence (Sroufe, 1997). Two deve lopmental pathways have been proposed to explain the inverse relationship betwee n academic competence and psychopathology (Masten & Curtis, 2000). First, the behavioral expression of severe childhood psychopathology can reduce academic engag ement and the acquisition of academic skills (Hinshaw, 1992; McClelland, Morriso n, & Holmes, 2000). Consequently, children and adolescents with severe psychopatholog y fail to attain academic skills that facilitate success in school, and are therefore at a greater risk of dropping out of school. A second explanation is that early academic failure r epresents one example of adaptational failure. In these cases, academic failure increases the risk of psychopathology among children and adolescent. As children and adolescent s struggle to attain academic competence they also may develop problems with soci al-emotional and behavioral functioning (Kellam, Mayer, Rebok, & Hawkins, 1992; Kellam, Rebok, Mayer, Iaolongo, & Kalodner, 1994; Maughan et al., 2003). Another p ossible explanation is that academic problems may be associated with other risk factors that are associated with severe

PAGE 138

127 psychopathology. The inverse relationship between a cademic competence in adolescence and Baker Act ERs may be an artifact of other risk factors (e.g., socioeconomic status; familial factors) common to both outcomes (Masten & Coatsworth, 1998;Masten & Curtis, 2000;Werner, 1989; Willcutt & Pennington, 2 000). Another possible explanation for the negative relat ionship between graduation rates and the use of the Baker Act relates to the s ocial competence of adolescents who graduate from high school. Social skills such as ma intaining relationships with adults (e.g., teachers, administrators, support personnel) may facilitate graduation from high school even for students who struggle academically, and therefore, may be an important protective factor for children and adolescents who are at risk for high school drop out. Many children and adolescents with severe psychopat hology may have impaired relationships with adults (Wagner & Sumi, 2005). Co nsequently, they may be less likely than socially competent students to receive the aca demic and social support that is important for persisting through high school until graduation. Similarly, school districts with higher graduation rates may provide more acade mic and behavioral supports that are important for facilitating graduation among childre n and adolescents than districts with lower graduation rates. School districts with highe r graduation rates also may provide a more positive climate that encourages student engag ement in school. For children and adolescents who display severe psychopathology, the availability of academic and behavioral support services coupled with a positive school climate may facilitate adaptation to the school environment while encourag ing high school graduation. Further, the availability of supports services and a more po sitive school climate may be important factors in preventing Baker Act ERs among many chil dren and adolescents.

PAGE 139

128 The percentage of children and adolescents that me t proficiency standards on the reading and math sections of the Florida Comprehens ive Achievement Test (FCAT) was not associated with Baker Act ER rates among school districts. Furthermore, grade retention rates were not associated with Baker Act ER rates. These findings are somewhat unexpected given the inverse relationship between g raduation rates and Baker Act ER rates. One explanation for these contradictory find ings is related to the data sources used for these variables. District proficiency levels in reading and math were calculated as the aggregate percentage of all students in a district who achieved a level 3 or higher on the FCAT for grades three to ten. The aggregate percent age of students in grades three through ten who achieved proficiency standards may not accurately reflect the academic outcomes of older students in grades 11 and 12 who may be at a greater risk of receiving a Baker Act ER. Retention rates also may not reflec t the academic proficiency of students at higher grade levels. Graduation rates are based on the number of high school seniors who graduate with a standard diploma and may be a m ore accurate reflection of the achievement outcomes among older students in the hi gh school population. An analysis of FCAT distributions between grade levels may have identified an inverse relationship between middle and high school FCAT scores and the use of the Baker Act. Analysis of district level variables indicated that race/ethnicity and socioeconomic status contributed uniquely to the variance in Bake r Act ERs and repeat Baker Act ERs. The negative relationship between the use of the Ba ker Act and minority enrollment suggests that the use of the Baker Act was less com mon in districts with a greater proportion of minority students. African American e nrollment was not significantly correlated with Baker Act ER rates after accounting for other variables in the regression

PAGE 140

129 equations. However, there was a significant negativ e correlation between repeat Baker Act ERs and African American students after control ling for the influence of other variables, suggesting that districts with a greater proportion of African American students were less likely to have students who received more than one Baker Act ER. The inverse relationship between minority enrollmen t, African American enrollment, and the use of the Baker Act is consist ent with findings from demographic comparisons that were analyzed as part of this stud y. White children and adolescents had greater odds of receiving a Baker Act than minority and African American children and adolescents. In addition to these findings, there a lso was an inverse relationship between socioeconomic status and the use of the Baker Act, suggesting that districts with a greater proportion of students who received free and reduce d lunch were less likely to use Baker Act ERs. Furthermore, children and adolescents from minority backgrounds, and children and adolescents of lower socioeconomic backgrounds historically have been less likely to receive community based mental health and behaviora l support services (Hough et al., 2002; McMiller & Weisz, 1996; Snowden, 1999; Taylor 2005; United States Department of Health and Human Services [USDHHS], 2003). Cultural variables may account for the negative rel ationship between race/ethnicity, socioeconomic status, and use of the Baker Act. Cul tural factors associated with race/ethnicity and socioeconomic status may influen ce the decisions of parents, school personnel, law enforcement personnel, and mental he alth professionals to utilize the Baker Act among children and adolescents. The cultu ral background of an individual has historically been a factor in determining mental he alth diagnosis, assessment, and treatment (Taylor, 2005;USDHHS (2003). For example African American and Latinos

PAGE 141

130 seek help from mental health service providers less often than White families (McMiller & Weisz, 1996). Incidents that culminate in a Baker Act ER begin when child and adolescents display symptoms and/or behaviors assoc iated with severe psychopathology. Influential adults such as school personnel, parent s, relatives, community based mental health professionals, and law enforcement personnel can choose to utilize the Baker Act as one response to children and adolescents who eng age in maladaptive behavior. The decision to utilize the Baker Act could be mediated by adult attributions regarding the etiology of psychopathology, knowledge of available mental health and behavioral services. Additionally, beliefs that support the ut ilization of mental health services can influence service utilization (Wu et al., 2001). It has been argued that current mental health syste ms have failed to incorporate the traditions, values, beliefs, and languages of diver se groups of individuals living in the Unites States (United States Department of Health a nd Human Services [USDHHS], 1999). Furthermore, variables such as adequate tran sportation and funding limit access to quality mental health treatment for many children a nd their families. Cultural factors also influence mental health and behavioral support serv ice providers and their beliefs about the appropriate response to severe psychopathology. These systemic factors have unintentionally created barriers to mental health s ervices that discourage service utilization among families from minority and lower socioeconomic backgrounds (Wagner & Sumi, 2005). Factors that affect access to qualit y mental health treatment seem to account for the disparity in service utilization. O ne consequence of these barriers is that minorities may choose other approaches to address s evere psychopathology among children and adolescents (USDHHS, 2003).

PAGE 142

131 The juvenile justice system seems to be utilized m ore often among African American children and adolescent (Miller, 2005; UHHS, 2001). Furthermore, the use of school based discipline practices (e.g, suspension) also i s more common among African American children and adolescents, and children fro m lower socioeconomic backgrounds (Skiba et al., 1997; Raffaele-Mendez, et al., 2002) Although the use of school based disciplinary referrals is not necessarily synonymou s with the use of the Baker Act, suspension and expulsion represent one option for r esponding to the behavioral expression of severe psychopathology in school sett ings. Minority children and adolescents, and children and adolescents from lowe r socioeconomic backgrounds may be less likely to receive Baker Act ERs because oth er approaches are more commonly utilized to address severe psychopathology among th ese populations (Florida Department of Education, 2005; Raffaele-Mendez, 2000; Raffaele -Mendez, et al., 2002).. District use of out-of-school suspension did not ac count for a unique proportion of the variance in overall Baker Act ER rates, althoug h the values associated with these analyses did approach significance. Out-of-school s uspension (OSS) rates were a unique contributor to the variance in repeat Baker Act ER rates. The positive relationship between OSS and repeat Baker Act ERs, suggests that districts that used OSS more frequently among students also had greater rates of repeat Baker Act ERs. The positive relationship between OSS and repeat Baker Act ER is consistent with previous studies that have found a higher risk of receiving an exclu sionary disciplinary referral (e.g., suspension) among children and adolescents who have severe emotional and behavioral disorders (Wagner et al., 2005; Skiba et al., 1997)

PAGE 143

132 It is plausible that the most severe types of psych opathology also are expressed as disruptive behavior in the school environment. Cons equently, OSS may be utilized as a strategy for addressing disruptive behavior display ed by child and adolescents who received more than one Baker Act ER. The presence o f zero tolerance policies utilize OSS for children and adolescents who display behavi or that is considered threatening to school environments. School administrators may beli eve that there is no other option other than the use OSS when district policies manda te this response following an incident that violates the student code of conduct (Skiba, 2 000). School districts located in counties with a greater use of repeat Baker Act ERs among the school age population may less likely to offer intensive mental health an d behavioral support services for children with severe EBD. These counties also may h ave less community mental health services that can address the needs of children and adolescents with severe psychopathology. The use of disciplinary strategies that remove children from mainstream settings can preclude the implementation of services that are designed to enhance social competence. Other methods of disciplinary exclusion such as inschool suspension and expulsion were not unique contributors to Baker Act ER rates. One explanation for the apparent contradiction in these findings is that school pers onnel utilize OSS on a more frequent basis for children and adolescents who display more severe and intense behavioral disruption (Costenbader & Markson, 1998; DeVoe et a l., 2005). In comparison to children and adolescents who received just one Bake r Act ER, multiple Baker Act ERs may be an expression of the most severe types of ps ychopathology among children and adolescents.

PAGE 144

133 The alternative to the use of exclusionary discipli nary practices is to provide mental health and behavioral support services for children with severe EBD. The present study attempted to examine the relationship between the a vailability of mental health service providers in a school district and the use of the B aker Act. Findings from the study indicated that the ratio of student mental health s ervice providers (i.e., guidance counselors, school psychologists, social workers) t o students did not contribute to a unique proportion of the variance in Baker Act ERs. The inclusion of this variable was based on research that has identified an associatio n between the availability of mental health and behavioral support services and placemen t in more restrictive educational and treatment settings (Bickman et al., 1996; Blanz & S chmidt, 2000; Hendrickson et al., 1998; McDermott et al., 2002; Muscott, 1994; Rock e t al., 1994; Romansky et al., 2003). Furthermore, school personnel are less likely to ut ilize exclusionary discipline practices if they have beliefs that support the use of behaviora l support services for children who display disruptive behavior (Hyman & Perone, 1998; Raffaele-Mendez et al., 2002; Skiba et al., 2003). Although the method used to calculat e this variable identified the number of school personnel who could potentially provide thes e services, the variable did not necessary reflect the types of mental health servic es available to students within a school district. The roles and responsibilities of many sc hool based practitioners often involve activities that are unrelated to the provision of b ehavioral and mental health services, and often there is a great deal of variability between the services provided by these individuals across districts (Curtis, Lopez, Batsch e, & Smith, 2006). Furthermore, ratios that incorporate school psychologists, guidance cou nselors, and social workers failed to account for other mental health practitioners that might provide services within the

PAGE 145

134 district. For example, many school districts may pr ovide mental health services through contractual agreements with community based provide rs. Limitations Several characteristics associated with the design of the study and data collection methodology limit the interpretability of the resul t obtained from this study. The primary limitation is its use of a correlational and causal comparative methodology. Data from correlational and causal comparative studies provid es an estimate of the association between selected independent and dependent variable s of interest. Second, archival records and survey data served as the independent and dependent variables for the study. Experimental manipulation of the independent variables would have been unethical and unfeasible given the nature of referrals for evaluation under Baker Act statutes. The current design and data ana lysis procedures allowed for the generation of statements related to the association between the selected district level variables and rates of Baker Act ERs. Regression eq uations facilitated the development of statements related to the proportion of variance accounted for by the selected district level variables. Additionally, data analysis proced ures identified variables that were most associated with Baker Act ER rates after controllin g for other variables entered into multiple regression equations. However, it is not p ossible to confirm any directional hypothesis related to the influence of the district level variables on Baker Act ER rates. Therefore, statements of causality are inappropriat e given the design of the study. It is not impossible to fully attribute any variation in Bake r Act ERs among school aged children to the selected independent variables. The statisti cal analyses procedures utilized in this study also represent another design limitation. Mul tiple regression analyses did not allow

PAGE 146

135 for an examination of possible interactions between the independent variables that were included in these analyses. A third limitation is related to the variables ide ntified for inclusion in the analyses. To date there is a limited literature bas e that has identified variables associated with Baker Act ERs. The majority of the literature related to Baker Act ERs has identified information related to rates and demogra phic characteristics of children who are referred under Baker Act statutes. Little is kn own about other variables that operate to influence the use and non-use of Baker Act ERs. The refore, literature related to actions conceptualized as being similar to Baker Act ERs wa s utilized to select independent variables for the analyses. The district level vari ables selected as independent variables for the study were based on a thorough review of th e research literature related to other forms of school based exclusion, utilization of men tal health services in community settings, and barriers to mental health services. O ther variables that contribute to the explanation of prediction of Baker Act ER rates may have been unintentionally omitted from the analyses. The current study focused on the analyses of demographic and district level variables that contribute to Baker Act ER rat es. The design of the present study did not allow for the examination of other variables (e .g., school based factors, familial characteristics, characteristics of the community) that operate at a more micro, level to influence Baker Act ER. For example, previous resea rch has identified a relationship between individual school level variables (e.g., be liefs of school personnel) that contribute to the use of other forms of referrals f or intensive mental health services and exclusionary disciplinary practices (i.e., Raffaele -Mendez et al., 2002). Therefore, it is possible that Baker Act ER rates vary more as a fun ction of variables that operate at the

PAGE 147

136 school building level as opposed to the district le vel processes that were examined for the present study. Additionally, the present study omit ted variables related to community based factors, familial influences, and individual personality variables that might play a role in Baker Act ER rates. For example, there is r esearch to suggest that Baker Act ER rates are higher in regions that are closer in prox imity to children’s crisis centers (Christy, 2005). Schools and other community based practition ers may be more likely to utilize the Baker Act if receiving facilities are available and in close proximity to the location where the referral is initiated. Fourth, the process in which Baker Act ER rates wer e calculated for each county may be a limitation to this study. The total number of Bake Act ERs within each of the 67 counties in Florida was divided by the number of st udents enrolled in the school district located in those counties. Thus, the number used as the dependent variable for the regression analyses should be more accurately ident ified as an index of Baker Act ER events as opposed to the rate among the population of children in each county. For purposes of this study, Baker Act ERs were conceptu alized as an event that led to removal of children from mainstream settings and th e intent was to identify a metric that closely approximated the frequency with which the e vent occurs among the population of children between the ages of 5-18 years of age. Clo sely related to this limitation is the fact that the denominator used to calculate odds an d risk ratios and in the calculation of Baker Act ER rates was based on enrolled students w ithin a district. The calculation of Baker Act ER rates may have excluded children who d ropped out of school prior to age 18 or graduated high before the age of 18 years The se variations in the calculation of Baker Act ER rates may have exerted a minimal impac t on the calculation of odds and

PAGE 148

137 risk ratios, proportionality, and regression analys is given the large sample size used in these analyses and the small number of children tha t were omitted from calculation of Baker Act ER rate. However, it is appropriate to me ntion any data collection and/or analyses procedures that have impacted the accuracy of the data used in the analyses. A final limitation of this study is related to the manner in which Baker Act ERs were identified. Odds ratios and Baker Act ER rates were calculated using Baker Act ERs that were initiated in both school and community se ttings. The present study did not account for the source of the Baker Act ER. An exam ination of data from Baker Act ERs initiated in school and community settings could de lineate any differences between variables that are common to those settings. The so urce of Baker Act ERs (e.g., parents vs. law enforcement, mental health profession, educ ational or community settings) may moderate the relationship certain variables and the use of the Baker Act. Implications and future directions The current study examined the relationship between demographic characteristics, school district variables, the availability of ment al health services, and use of the Baker Act. Additional research should continue to examine this relationship by identifying mediating and moderating variables that affect the decision to use the Baker Act with children and adolescents. Research designs that inc orporate quantitative and qualitative methodology can be useful in clarifying factors tha t contribute to the greater use of the Baker Act among the adolescent population. Case stu dies can provide information that contributes to a better understanding of developmen tal trajectories that culminate in a Baker Act ER. These studies should include longitud inal designs in an attempt to understand the case history of children and adolesc ents who receive Baker Act ERs.

PAGE 149

138 Additional research is needed to better understand the experiences of children and adolescents after receiving a Baker Act ER. Future research should continue to examine risk factors associated with multiple Baker Act ERs among children and adolescents. Potential risk factors that could be explored inclu de experiences upon re-entry into school and the community, and mental health services provi ded after Baker Act ERs. Additionally, the interaction between child and fam ilial risk factors should be explored to examine the contribution of personality and cultura l factors that influence repeat Baker Act ERs. A better understanding of these risk facto rs can facilitate the development of after care services and procedures that reduce the likelihood of repeat Baker Act ERs. Academic underachievement may contribute to or exa cerbate severe psychopathology by widening the gap between compete ncies among children and adolescents and environmental expectations (Masten & Curtis, 2000). School reform initiatives have emphasized the importance of acade mic achievement among all students, including those who display severe psychopathology (NCLB, 2001). Consequently, academic intervention and curriculum modifications should be considered an essential component of mental health and behavioral support s ystems (Ialongo, Werthamer, Kellam, Brown, Wang, & Yin, 1999; Kellam, Mayer, Re bok, & Hawkins, 1996; Kellam, Rebok, Mayer, Ialongo, & Kalodner, 1994). The devel opment of academic competence may be an important protective factor for children who are at risk for Baker Act ERs (Masten & Coatsworth, 1998). Behavioral support and mental health systems should be designed with an appreciation for the interaction b etween academic competence and behavioral/emotional problems. Academic interventio ns can enhance the competency of children and adolescents at risk for severe psychop athology and create important

PAGE 150

139 protective factors for many children and adolescent s who are at risk for severe psychopathology. Future research may use data disag gregated by grade level may help to clarify the impact of academic achievement on the i ncidence of severe psychopathology and the use of the Baker Act among children and ado lescents. Future studies also should include an examination of the relationship between various outcome measures of academic achievement and the use of the Baker Act. It also may be beneficial to examine the relationship between types and number of academ ic supports available to students and Baker Act ERs. The identification of early risk factors associated with the use of the Baker Act can contribute to early detection methods for child ren and adolescents most at risk for emergency mental health services like Baker Act ERs Effective systems of early identification and intervention can reduce the risk of severe psychopathology that emerges across childhood and into adolescence, ther eby reducing the need to access the most intensive mental services as children grow old er (Doll & Lyon, 1998; USDHHS, 2003). The creation of behavioral and mental health services delivery systems based on tiered models can facilitate service delivery that is responsive to the needs of children who display a wide range of emotional/behavioral di sorders across childhood and adolescence. These systems include universal preven tion strategies, secondary intervention for children at risk for severe psycho pathology, and tertiary supports for children and adolescents who have developed emotion al and behavioral disorders that significantly impair daily functioning (Gresham, 20 04; Walker & Shinn, 2002). Additional resources can be allocated in a manner t hat is responsive to the local population when children fail to respond to prevent ion and secondary intervention efforts

PAGE 151

140 (Deno, 2004; Power, 2002). These services should be delivered early to facilitate the social competence of children and adolescents, and prevent negative developmental trajectories from escalating into severe psychopath ology as children grow older (Sroufe, 1997). Future research should examine the relationship bet ween systemic factors that influence the process by which children and adolesc ents are identified for behavioral support and mental health services and the use of t he Baker Act. It would be beneficial to more closely examine the impact of various service delivery models (e.g., systems based on tiered models of support) on Baker Act ER rates. The location of these services (e.g., school setting vs. community based) also could be a factor that influences the use of the Baker Act. Case studies and other qualitative metho ds could be used among a sample of children and adolescents to determine developmental trajectories associated with Baker Act ERs. Longitudinal designs also could be used to investigate the experiences of children and adolescents who receive Baker Act ERs. Mental health and behavioral support systems should be designed with close attention to cultural factors that influence service utilizat ion. Service delivery systems should be designed to minimize the impact of factors that lim it access to quality of mental health services for children and families from minority an d lower socioeconomic backgrounds. Furthermore, an emphasis on culturally competent se rvice delivery can increase the quality of services that are provided to children f rom diverse backgrounds. The implementation of culturally competent mental healt h services can increase service utilization among underserved children and adolesce nts, and prevent involvement in school disciplinary and juvenile justice systems (U SDHHS, 2003).It may be beneficial to

PAGE 152

141 examine more closely the responses that are chosen by adults when children and adolescents display disruptive behavior, and whethe r cultural variables influence the decision to utilize intensive mental health service s like involuntary psychiatric examination (i.e., the Baker Act.). Furthermore, re search into adult responses should identify the procedures that are used when respondi ng to children in crisis and the events leading up to the decision to use the Baker Act. Ma ny adults may harbor attitudes and beliefs that lead to disciplinary approaches for ch ildren and adolescents who display severe psychopathology. These variables may vary ac cording to the cultural background of the child and the adults who interact with child ren. A greater understanding of the relationship between cultural variables, service ut ilization, and the use of the Baker Act can facilitate the development of culturally compet ent service delivery models that reduce barriers to quality mental health services f or traditionally underserved populations. Schools often become the sole provider of behaviora l and mental health services for children and adolescents with severe psychopatholog y. Consequently, school based mental health services can serve as a valuable reso urce for children and adolescents who have been underserved by traditional mental health models. Behavioral and mental health programs can vary across districts in terms of the types and intensity of services that are provided to students (Foster et al., 2005; Slade, 2 003). The decision to rely more heavily on punishment based strategies like OSS may limit o pportunities to provide interventions designed to enhance the academic and social compete nce of students with severe emotional and behavioral disorders (Raffale-Mendez, et al., 2003). Alternatively, the implementation of evidenced based behavioral and me ntal health services can enhance

PAGE 153

142 protective factors, reduce risk, and lessen the imp act of developmental trajectories associated with negative outcomes for children with emotional and behavioral disorders (Masten & Curtis, 2003; Sroufe, 1997). Additional r esearch is necessary to further investigate the relationship between the availabili ty of behavioral and mental health services, disciplinary reactions to severe psychopa thology, and the use of the Baker Act. Conclusion The use of the Baker Act facilitates access to the most intensive mental health services available in the community. The present st udy sought to identify demographic variables and school district level variables that are related to the use of the Baker Act among children and adolescents. The odds of receivi ng a Baker Act among children and adolescents varied according to age, gender, and ra ce/ethnicity. District level variables are associated with Baker Act ER rates among childr en and adolescents. Research related to the use of the Baker Act should continue to iden tify factors associated with the use and non-use of the Baker Act with children and adolesce nts. These efforts should focus on the identification of mediating and moderating variable s that influence the use of the Baker Act for children and adolescents who display severe psychopathology. Furthermore, attention should be given to the interaction betwee n systemic level variables and individual child variables. Developmental models of psychopathology and risk/resilience models can serve as useful frameworks for guiding r esearch related to the use of the Baker Act. The identification of risk and protectiv e factors can more clearly delineate the characteristics of children and adolescents most at risk for receiving a Baker Act ER.

PAGE 154

143 References Adelman, H.S., & Taylor, L. (1999). Mental health i n schools and system restructuring. Clinical Psychology Review, 19 137-163. Advancement Project / Civil Rights Project. (2000). Opportunities suspended: The devastating consequences of zero tolerance and sch ool discipline policies. Cambridge, MA: Harvard University. American Psychiatric Association (1994). Diagnostic and statistical manual of mental disorders (4th ed.) Washington, DC: Author. Bickman, L., Foster, M.E., & Lambert, E.W. (1996). Who gets hospitalized in a continuum of care? Journal of the Academy of Child and Adolescent Psyc hiatry, 35 74-80. Berquist, C.C., Bigbie, C.L., Groves, L., & Richard son, G.H. (2004). Executive summary: Evaluation report for the study of altern atives to suspension. Evaluation Systems Design, Inc. Tallahassee: FL. Blanz, B., & Schmidt, M.H. (2000). Practitioner rev iew: Preconditions and outcomes of inpatient treatment in child and adolescent psychi atry. Journal of the Academy of Child and Adolescent Psychiatry, 41 703-712. Christle, C.A., Jolivette, K., & Nelson, C.M. (2005 ). Breaking the school to prison pipeline: Identifying school risk and protective f actors for youth delinquency. Exceptionality, 13 69-88.

PAGE 155

144 Christy, A. (2005). Special report of repeated Baker Act examinations o f children with special emphasis on Department of Children and Fam ilies districts Louis de la Parte Florida Mental Health Institute, University of South Florida: Tampa, FL. Christy, A., Petrila, J., Hudacek, K., Haynes, D., Wedekind, T., & Pulley, A. (2005). Acute mental health care services for children in Pinellas County Florida. In C. Newman, C.J. Liberton, K. Kutash, &, R.M. Fried man (Eds.), The 17th Annual Research Conference Proceedings: A System of Care for Children’s Mental Health: Expanding the Research Base (pp. 395-400). Tampa: The Louis de la Parte Florida Mental Health Institute, University of South Florida. Christy, A., Stiles, P.G., & Shanmugam, S. (2003). The Florida Mental Health Act (The Baker Act): 2002 Annual Report. Prepared for the F lorida Agency for Health Care Administration Louis de la Parte Florida Mental Health Institute Department of Mental Health Law & Policy. Policy a nd Services Research Data Center, Tampa: FL. Christy, A., McCranie, M. (2005). The Florida Mental Health Act (The Baker Act): 2005 Annual Report. Prepared for the Florida Agency for Health Care Administration Louis de la Parte Florida Mental Health Institute, Department of Mental Health Law & Policy. Policy and Services Research Data Ce nter, Tampa: FL. Chow, J.C., Jafee, K., & Snowden, L. (2003). Racial /ethnic disparities in the use of mental health services in poverty areas. American Journal of Public Health, 93, 792-797. Costenbader, V.K., & Markson, S. (1994). School sus pension: A survey of current polices and practices. NASSP Bulletin 103-110.

PAGE 156

145 Costenbader, V., & Markson, S. (1998). School suspe nsion: A study with secondary school students. Journal of School Psychology, 36 59-82. Coutino, M.J., & Oswald, D. (1996). Identification and placement of students with Serious Emotional Disturbance. Part II: National a nd state trends in the implementation of LRE. Journal of Emotional and Behavioral Disorders, 4, 4052. Curtis, M.J., Lopez, A.D., Batsche, G.M., & Smith, J.C. (2006), March. School Psychology: A national perspective. Paper presented at the annual conference of the National Association of School Psychologists, Anaheim, CA. Deno, S.L. (2002). Problem solving as “Best Practic e.” In A. Thomas & J. Grimes (Eds.), Best practices in school psychology (4th ed.): Vol. 2 Bethesda, MD: National Association of School Psychologists. DeVoe, J.F, Peter, K., Noonan, M., Snyder, T.D, & B aum, K. (2005). Indicators of school crime and safety: 2005 (NCES 2006-001/NCJ21 0697) U.S. Department of Education and Justice. Washington, DC: U.S. Gov ernment Printing Office. Retrieved February 21, 2006 from http://nces.ed.g ov. Doll, B., & Lyon, M.A. (1998). Risk and resilience. Implications for the delivery of educational and mental health services in schools. School Psychology Review, 27 348-363. Duncan, B.B., Forness, S.R., & Hartsough, C. (1995) Students identified as seriously emotionally disturbed in school-based day treatmen t: Cognitive, psychiatric, and special education characteristics. Behavioral Disorders, 20 238-252.

PAGE 157

146 Farmer, M.Z., Burns, B.J., Phillips, S.D., Angold, A., & Costello, E.J. (2003): Pathways into and through mental health services for childr en and adolescents. Psychiatric Services, 54 60-66. Farmer, T.W., Quinn, M.M., Hussey, W., & Holahan, T (2001). The development of disruptive behavioral disorders and correlated con straints: Implications for intervention. Behavioral Disorders, 26 117-130. Florida’s Baker Act Website (2002). Baker Act. Invo luntary Examination: Criteria, processes, and timeframes. State of Florida Depart ment of Children and Families Mental Health Program Office. Florida Senate, Committee on Children and Families (2005). Clarifying the Baker Act requirements as they relate to children’s receivin g and crisis stabilization units (Interim Project Report 2006-103). Tallahassee, FL : Author. Florida Department of Education (2005). School Environment Safety Incident Report (SESSIR)-District and State Reports Tallahasse, FL: Author. Retrieved February 14, 2006 from http://www.firn.edu/doe/besss/sesir/ sesir0405.htm. Florida Department of Education. (n.d.). Florida sc hool district data. Tallahassee, FL: Author. Retrieved March 6, 2006 from http://www.firn.edu/doe/eias/flmove/florida.htm. Florida Statutes Section 394, Part I (2005). Forness, S.R. (1992). Broadening the cultural-organ izational perspective in exclusion of youth with social maladjustment: First invited rea ction to the Maag and Howell Paper. Remedial and Special Education, 13 55-59.

PAGE 158

147 Foster, S., Rollefson, M., Doksum, T., Noonan, D., Robinson, G., & Teich, J. (2005). School mental health services in the United States, 2002-2003 DHHS Pub No. (SMA) 05-4068. Rockville, MD: Center for Mental He alth Services, Substance Abuse and Mental Health Services Administrator. Gresham, F. M. (2002). Current status and future di rections of school-based behavioral interventions. School Psychology Review, 33 326-343. Halliday-Boykins, C.A., Henggeler, S.W., Rowland, M .D., & DeLucia, C. (2004). Heterogeneity of youth symptom trajectories follow ing psychiatric crisis: Predictors and placement outcomes. Journal of Consulting and Clinical Psychology, 72 993-1003. Hendrickson, J.M., Smith, C.R., Frank, A.R., & Meri cal, C. (1998). Decision making factors associated with placement of students with emotional and behavioral disorders in restrictive educational settings. Education and Treatment of Children, 21 275-302. Hinshaw, S.P. (1992). Externalizing behavior proble ms and academic underachievement in childhood and adolescence: Causal relationships and underlying mechanisms. Psychological Bulletin, 111 127-155. Hyman, I.A., & Perone, D.C. (1998). The other side of school violence. Educator policies and practices that may contribute to student misbe havior. Journal of School Psychology, 36 7-27.

PAGE 159

148 Ialongo, N.S., Werthamer, L., Kellam, S.G., Brown, H., Wang, S., & Lin, Y. (1999). Proximal impact of two first-grade preventive inte rventions on the early risk behaviors for later substance abuse, depression, a nd antisocial behavior. American Journal of Community Psychology, 27 599-641. Individuals with Disabilities Education Improvement Act (IDEIA), Pub. L. No. 108-446. (2004). Kellam, S., Mayer, L., Rebok, G. and Hawkins, W. (1 996). The Effects of Improving Achievement on Aggressive Behavior and of Improvin g Aggressive Behavior on Achievement Through Two Preventive Interventions: An Investigation of Causal Paths. In Dohrenwend (ed.) Adversity, Stress and Psychopat hology : American Psychiatric Press, American Psychiatric Associatio n Kellam, S.G., Rebok, G.W., Mayer, L.S., Iaolongo, N ., & Kalodner, C.R. (1994). Depressive symptoms over 1st grade and their response to a developmental epedimoligcally based preventive trial aimed at im proving achievement. Development and Psychopathology, 6 463-481. Kleiner, B., Porch, R., & Farris, E. (2002). Public alternative schools and programs for students at risk for education failure: 2000-01 (N CES 2002-004) U.S. Department of Education. Washington, DC: National Center for Education Statistics. Retrieved January 24, 2006 from http// : www.nces.ed.gov. Maag, J.W., & Howell, K.W. (1992). Special educatio n and the exclusion of youth with social maladjustment: A cultural-organizational pe rspective. Remedial and Special Education, 13 47-54.

PAGE 160

149 Mash, E.J., & Dozois, D.A. (2003). Child psychopath ology: A developmental systems perspective. In R.A. Barkley, & E.J. Mash (Eds.), Child Psychopathology (2nd ed.). Guilford Masten, A.S., & Curtis, W.J. (2000). Integrating co mpetence and psychopathology: Pathways toward a comprehensive science of adaptat ion in development. Development and Psychopathology, 12 529-550. Masten, A.S., & Coatsworth, J.D. (1998). The develo pment of competence in favorable and unfavorable environments: Lessons from researc h on successful children. American Psychologist, 53 205-220. McClelland, M.M., Morrison, F.J., & Holmes, D.L. (2 000). Children at risk for early academic problems: The role of learning-related so cial skills. Early Childhood Research Quarterly, 15 307-329 McDermott, B.M., McKelvey, R., Roberts, L. & Davies L. (2002). Severity of children’s psychopathology and impairment and its relationshi p to treatment setting. Psychiatric Services, 53 57-62. McFadden, A.C., Marsh, G.E., Price, B.J., & Hwang, Y. (1992). A study of race and gender bias in the punishment of school children. Education and Treatment of Children, 15 140-146. McMiller, W.P., & Weisz, J.R. (1997). Help-seeking preceding mental health clinic intake among African-American, Latino, and Caucasi an youth. Journal of the American Academy of Child and Adolescent Psychiatr y, 36 443-445.

PAGE 161

150 Miller, T.L. (2005). Evidence of disparities in men tal health services for black youth. In C. Newman, C.J. Liberton, K. Kutash, &, R.M. Fried man (Eds.), The 17th Annual Research Conference Proceedings: A System of Care for Children’s Mental Health: Expanding the Research Base (pp.323-326). Tampa: The Louis de la Parte Florida Mental Health Institute, University of South Florida. Moffit, T.E. (1993). The neuropsychology of conduct disorder. Development and Psychopathology, 5 135-151. Morrison, G.M., & D’Incau, B. (2000). Developmental and service trajectories of students with disabilities recommended for expulsi on from school. Exceptional Children, 66 257-271. Morrison, G.M., & Skiba, R.J. (2001). Predicting violence from school misbehavior: Promises and Perils. Psychology in the Schools, 38 173-184. Muscott, H.S. (1997). Behavioral characteristics of elementary and secondary students with emotional/behavioral difficulties in four dif ferent cascade placements. Education and Treatment of Children, 20 336-356. National Commission on Excellence in Education. (19 83). A nation at risk: The imperative for education reform. Recommendations. Washington, DC: U.S. Department of Education. National Center for Educational Statistics (2003). Suspensions and expulsions of public elementary and secondary school students, by state sex, percent of enrollment: 2000. U.S. Department of Education, Office of Civil Rigt hs. Washington, DC: Author. Retrieved November 15, 2005 from http://ww w.nces.ed.gov. Neuendorf, K.A. (2002). The content analysis guidebook. Thousand Oaks, CA: Sage.

PAGE 162

151 No Child Left Behind Act of 2001, Public L. No. 07110. (2002). Owens, P.L., Hoagwood, K., Horwitz, S.M., Leaf, P.J ., Poduska, J.M., Kellam, S.G., & Ialongo, N.S. (2002). Barriers to children’s menta l health services. Journal of the American Academic of Child and Adolescent Psychiat ry, 41 731-738. Patterson, G.R. (1986). Performance models for anti social boys. American Psychologist, 41 432-444. Power, T.M. (2002). Promoting children’s mental hea lth: Reform through interdisciplinary and community partnerships. School Psychology Review, 32 316. Raffaele-Mendez, L.M. (2000). An analysis of out-of-school suspensions in Hillsbo rough County: An analysis commissioned by the Hillsborou gh Constituency for Children. Tampa, FL: Children’s Board of Hillsborough County Raffaele-Mendez, L.M. (2003). Predictors of suspens ion and negative school outcomes: A longitudinal investigation New Directions for Youth Development, 99 17-33. Raffaele-Mendez, L., Knoff, H.M., & Ferron, J.M. (2 002). School demographic variables and out-of-school suspension rates: A quantitative and qualitative analysis of a large, ethnically diverse school district. Psychology in the Schools, 39 259-277. Roberts, R.E., Attkisson, C.C., & Rosenblatt, A. (1 998). Prevalence of psychopathology among children and adolescents. American Journal of Psychiatry, 155 715-725. Rock, E.E., Rosenberg, M.S., & Carran, D.T. (1994). Variables affecting the reintegration rate of students with Serious Emotio nal Disturbance. Exceptional Children, 61 254-268.

PAGE 163

152 Rodney, L.W., Crafter, B., Rodney, H.E., & Mupier, R.M. (1999). Variables contributing to grade retention among African American adolesce nt males. The Journal of Educational Research, 92 185-190. Romansky, J.B., Lyons, J.S., Lehner, R.K., & West C.M. (2003). Factors related to psychiatric readmission among children and adolesc ents in state custody. Psychiatric Services, 54 356-362. Skiba, R.J. (2000). Zero tolerance, zero evidence: An analysis of school disciplinary practices (Report # RR-SRS2). Bloomington, IN: Indiana University, Education Policy Center. Skiba, R.J., Michael, R.S., Nardo, A.C., & Peterson R.L. (2002). The color of discipline: Sources of racial and gender disproportionality in school punishment. The Urban Review, 34 317-342. Skiba, R.J., & Peterson, R.L. (2000). School discip line at a crossroad: From zero tolerance to early response. Exceptional Children, 66 335-347. Skiba, R.J., Peterson, R.L., & Williams, T. (1997). Office referrals and suspension: Disciplinary intervention in middle schools. Education and Treatment of Children, 20 295-315. Slade, E. (2003). The relationship between school c haracteristics and the availability of mental health and related services in middle and h igh schools in the United States. The Journal of Behavioral Health Services and Resea rch, 30 382-392. Snowden, L.R. (1999). African American service use for mental health problems. Journal of Community Psychology, 27 303-313.

PAGE 164

153 Sroufe, L.A. (1997). Psychopathology as an outcome of development. Development and Psychopathology, 9 251-268. Stevens, J. (1999). Intermediate statistics: A modern approach (2nd ed.) Mahwah, NJ: Lawrence Erlbaum Associates. Sugai, G., Sprague, J.R., Horner, R.H., & Walker, H .M. (2000). Preventing school violence: The use of office discipline referrals t o assess and monitor school-wide discipline interventions. Journal of Emotional and Behavioral Disorders, 8 94101. Taylor, E.D. (2005). Afrocultural ethos and mental health services: Understanding behavior in African American children. In C. Newma n, C.J. Liberton, K. Kutash, &, R.M. Friedman (Eds.), The 17th Annual Research Conference Proceedings: A System of Care for Children’s Mental Health: Expan ding the Research Base (pp. 320-322). Tampa: The Louis de la Parte Florida Mental Health Institute, University of South Florida. The Civil Rights Project. (2000). Opportunities suspended: The devastating consequences of zero tolerance and school discipli ne policies [On-line]. Available: www.law.harvard.edu/civil rights. United States Department of Education. (1998). Early warning, timely response: A Guide to safe schools. Washington, DC: Author.

PAGE 165

154 United States Department of Education. National Cen ter for Education Statistics. Violence and Discipline Problems in U.S. Public Sch ools: 1996-97, NCES 98030, by Sheila Heaviside, Cassandra Rowand, Catrin a Williams, and Elizabeth Farris. Project Officers. Shelley Burns and Edith McArthur. Washington, DC: 1998. United States Department of Education Office of Spe cial Education Program (2001). Twenty-third Annual Report to Congress on the Imple mentation of the Individuals with Disabilities Education Act. U.S. Department of Education, Office of Special Education Programs (OSEP). United States Department of Education Office of Spe cial Education Program (2003). Twenty-fifth Annual Report to Congress on the Imple mentation of the Individuals with Disabilities Education Act. U.S. Department of Education, Office of Special Education Programs (OSEP). United States Public Health Services Office of the Surgeon General (1999). Mental health: A report of the Surgeon General. Rockville, MD: Department of Health and Human Services, U.S. Public Health Service. United States Secret Service and United States Depa rtment of Education (2002). The final report and findings from the safe school ini tiative: Implications for the prevention of school attacks in the United States Washington, D.C.: Authors. United States Department of Health and Human Servic es (2003). New freedom commission on mental health, Achieving the promise : Transforming mental health care in America. Final Report. DHHS Pub. No SMA-03-3832. Rockville, MD.

PAGE 166

155 Wagner, M., Newman, L., Cameto, R., & Levine, P. (2 005). Changes over time in the early postschool outcomes of youth with disabiliti es. A report of findings from the National Longitudinal Transition Study (NLTS) and the National Longitudinal Transition Study-2 (NLTS2) Menlo Park, CA: SRI International. Available at http://www.nlts2.org/pdfs/changetime_compreport.pd f. Wagner, M., Kutash, K., Duchnowski, A.J., Epstein, M.H., & Sumi, W.C. (2005). The children and youth we serve: A national picture of the characteristics of students with emotional disturbances receiving special educ ation. Journal of Emotional and Behavioral Disorders, 13 79-96. Wagner, M., & Sumi, W.C. (2005). NLTS2: A national look at school programs and services for students with Emotional Disturbance. In C. Newman, C.J. Liberton, K. Kutash, & R.M. Friedman (Eds.), The 17th Annual Research Conference Proceedings: A System of Care for Children’s Menta l Health: Expanding the Research Base (pp. 395-400) Tampa: The Louis de la Parte Florida Mental Health Institute, University of South Florida. Walker, H.M., & Shinn, M.R. (2002). Structuring sch ool based interventions to achieve integrated primary, secondary, and tertiary goals for safe and effective schools. In M.R. Shinn, H.M. Walker, & G. Stoner (Eds.)., Interventions for Academic and Behavior Problems II: Preventive and remedial appr oaches Bethesda, MD: National Association of School Psychologists. Werner, E.E. (1989). High-risk children in young ad ulthood: A longitudinal study from birth to 32 years. American Journal of Orthopsychiatry, 59 72-81

PAGE 167

156 Willcutt, E.G., & Pennington, B.F. (2000). Psychiat ric comorbidity in children and adolescents with reading disability. Journal of Child Psychology and Psychiatry and Allied Disciplines, 41 1039-1048. Wu, P., Hoven, C.W., Cohen, P., Liu, X., Moore, R.E ., Tiet, Q., Okezie, N., Wicks, J., & Bird, H.R. (2001). Factors associated with use of mental health services for depressed children and adolescents. Psychiatric Services, 52 189-195. Wu, S.C., Pink, W.T., Crain, R.L., & Moles, O. (198 2). Student suspension: a critical reappraisal. The Urban Review, 14 245-303.

PAGE 168

157 Appendices

PAGE 171

160 Appendix C: Coding of District Level Variables District DemographicVariables: 1. Total student enrollment within a district (Fall 20 05). These data were entered as a continuous variable : 0-300,000. 2. The percent of minority students within a district (Fall 2005. These data were entered as a continuous variable: 0-100 3. The percent of students in the district who receive d free and reduced lunch. These data were entered as a continuous variable: 0 -100. 4. The percent of students within the district who are African American. These data were entered as a continuous variable : 0-100. 4. The percent of student in the district who received special education services under the Individuals with Disabilities Education A ct. These data were entered as a continuous variable: 0-100. 5. The percent of students in the district who receive d special education services under the EH/SED category. These data were entered as a continuous variable: 0-100. 6. The percent of students in a district who were susp ended. These data were entered as a continuous variable: 0-100. 7. The percent of students in a district who were expe lled. These data were entered as a continuous variable: 0-100.

PAGE 172

161 Appendix C: (continued ) 8. The percent of students in a district who were reta ined. These data were entered as a continuous variable: 0-100. 9. The percent of students in a district who graduated high school with a standard diploma. These data were entered as a cont inuous variable: 0-100. 10. The percent of the students in a district who obtai ned a Level 3 or higher on the reading and math sections of the Florida Compre hensive Achievement Test (FCAT). These data were entered as a continuou s variable:0-100. 11. The number of mental health workers in each school district. For purposes of this study, school social workers, guidance counsel ors, and school psychologists were considered providers of mental h ealth services. The ratio of mental health workers to enrolled student was ca lculated by utilizing these data. These data were entered as a continuous varia ble: 0-2,000.

PAGE 173

162 Appendix D: Data Transfer and Procedural Integrity Form Baker Act Study Data Transfer Integrity Form Directions: For each randomly selected district, first calcul ate the appropriate statistic using data sheets fro m the Florida Department of Education, and Microsoft excel sheets Second, use the SPSS data sheet to check for accu racy in the data transfer. Third, place a (+) or (-) symbol in the c orresponding cell to indicate if the data was calcu lated and transferred accurately. Fourth, calculate the percentage of tot al transfers that were transferred accurately and p ut this number in the box below the table. Use the following formula to calcu late percent accurate transfers. District Enroll PerBA PerRep PerMin PerFree PerSped PerEH SED PerRet PerGrad Per3read Per3math Per3 comp PerOSS PerISS PerExp Ratio MHStud Per Afr 4 5 6 13 20 24 27 40 43 45 50 58 64 Total Accurate Transfers = /221 Total % Accurate Transfers = %

PAGE 174

163 Appendix D: (continued) Demographic Data: Directions: For each demographic category, use the Microsoft ex cel spreadsheet printouts to calculate the percentage of accurate transfers to the SAS dat a sheets. Put a (+) on the line if the data was tra nsferred correctly and place a (-) on the line if the tr ansfer was inaccurate. Age/Grade: BA Children Total BA Exam s Elementary: ___ ___ Middle: ___ ___ High: ___ ___ Gender: Male: ___ ___ Female : ___ ___ Race/ethnicity: White: ___ ___ Black: ___ ___ Hispanic: ___ ___ Asian: ___ ___ Multiracial/Other: ___ ___ Total: /10 /10

PAGE 175

164 Appendix E: Scatterplot Displaying Homoscedasticiy for Selected Multiple Regression Regression Standardized Predicted Value2 1 0 -1 -2 -3 Regression Standardized Residual43210 -1-2 Dependent Variable: perBa Note. This scatter plot shows homoscedasticity for one mu ltiple regression equation involving the following variables. Dependent variab le = percent Baker Act; Independent variables included: enrollment, percent minority en rollment, percent free and reduced lunch, percent received special education services, percent received EHSED services, percent graduation, percent retained, percent level 3 or higher on the FCAT, percent out of school suspension, percent in-school suspension, percent expelled, and ratio of mental health service providers to students.

PAGE 176

About the Author Bradley S. Beam was born on August 12, 1978 in Mec hanicsburg, Pennsylvania. He received a B.A. in psychology from Millersville University in 2001. In August of 2001, Brad entered the Ph.D. program in school psyc hology at the University of South Florida, earning a M.A. (2002) and an Ed.S. (2006) in school psychology. While attending the school psychology program at the Univ ersity of South Florida, he specialized in the design and implementation of beh avioral support and mental health services for children with emotional and/or behavio ral disorders. Brad also developed an interest for pediatric health issues, systemic fact ors that influence service delivery, and organizational development. He has experience worki ng with children and adolescents in both school and clinic settings. In his free time, Brad enjoys exercising, eating o ut, going to the movies, relaxing with friends, watching football, learning about his toric events that have shaped the world, and reading. He currently resides in Tampa, FL.