<%BANNER%>

The Impact of School Choice on Student Reading Achievement in the 'No Child Left Behind' (NCLB) Era

Permanent Link: http://ufdc.ufl.edu/UFE0043647/00001

Material Information

Title: The Impact of School Choice on Student Reading Achievement in the 'No Child Left Behind' (NCLB) Era
Physical Description: 1 online resource (76 p.)
Language: english
Creator: Castellani, Brian
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2011

Subjects

Subjects / Keywords: choice -- reading -- school
Human Development and Organizational Studies in Education -- Dissertations, Academic -- UF
Genre: Educational Leadership thesis, Ed.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: When No Child Left Behind (NCLB) became law, it was clear that government leaders wanted to align high-stakes testing with school, district, and state accountability. The goal was for all students to be successful, as defined by NCLB. An important part of this law is for all students to have the choice to attend higher-performing schools if their zoned school is deemed underperforming. For the first time, legislation provided this option to everyone. Since the choice option was enacted, few studies have explored the relationship between school choice and reading achievement. The purpose of this study was to determine whether there is a relationship between school choice and student achievement for selected subgroups (gender, race, English Language Learner ELL) status, Exceptional Student Education status, economically disadvantaged status) and subtests (Overall Reading Score, Information subtest, Literature subtest, Words and Phrases in Context subtest, Main Idea, Plot, and Purpose subtest, Comparisons and Cause/Effect subtest, and Reference and Research subtest) on the fourth grade Reading Florida Comprehensive Achievement Test (FCAT). This study also determined which factors predicted reading achievement during the 2008 and 2009 FCAT administrations. 10 The third and fourth grade Reading FCAT Developmental Scale Scores of 953 students were collected from one Southwest Florida school district. Student data were separated into two groups (School Choice and non-School Choice) and further separated by subgroups. Data from subgroups were analyzed separately, using SAS, Version 9.2. Three research questions were addressed, using tests of comparison groups (chi-square tests for binary variables and categorical groups and t-tests for numerical variables) and bivariate testing and regression (t-tests for binary variables and ANOVA, with a follow-up Tukey-Kramer procedure for categorical predictors). The correlation between 2008 and 2009 reading scores also was determined. The results of this study indicate that there is no evidence to support the claim that school choice significantly affects reading achievement. Students in both the School Choice and non-School Choice groups made statistically comparable gains on the Reading FCAT from 2008-09. Additionally, economically disadvantaged students, ELL students, and students from certain racial groups (African American and Hispanic) transferred at significantly lower rates than did students from comparison groups (non-economically disadvantaged, non-ELL, and Caucasian). The findings show that students who took advantage of the opportunity to transfer included those who were the least disadvantaged.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Brian Castellani.
Thesis: Thesis (Ed.D.)--University of Florida, 2011.
Local: Adviser: Behar-Horenstein, Linda S.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2013-06-30

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2011
System ID: UFE0043647:00001

Permanent Link: http://ufdc.ufl.edu/UFE0043647/00001

Material Information

Title: The Impact of School Choice on Student Reading Achievement in the 'No Child Left Behind' (NCLB) Era
Physical Description: 1 online resource (76 p.)
Language: english
Creator: Castellani, Brian
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2011

Subjects

Subjects / Keywords: choice -- reading -- school
Human Development and Organizational Studies in Education -- Dissertations, Academic -- UF
Genre: Educational Leadership thesis, Ed.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: When No Child Left Behind (NCLB) became law, it was clear that government leaders wanted to align high-stakes testing with school, district, and state accountability. The goal was for all students to be successful, as defined by NCLB. An important part of this law is for all students to have the choice to attend higher-performing schools if their zoned school is deemed underperforming. For the first time, legislation provided this option to everyone. Since the choice option was enacted, few studies have explored the relationship between school choice and reading achievement. The purpose of this study was to determine whether there is a relationship between school choice and student achievement for selected subgroups (gender, race, English Language Learner ELL) status, Exceptional Student Education status, economically disadvantaged status) and subtests (Overall Reading Score, Information subtest, Literature subtest, Words and Phrases in Context subtest, Main Idea, Plot, and Purpose subtest, Comparisons and Cause/Effect subtest, and Reference and Research subtest) on the fourth grade Reading Florida Comprehensive Achievement Test (FCAT). This study also determined which factors predicted reading achievement during the 2008 and 2009 FCAT administrations. 10 The third and fourth grade Reading FCAT Developmental Scale Scores of 953 students were collected from one Southwest Florida school district. Student data were separated into two groups (School Choice and non-School Choice) and further separated by subgroups. Data from subgroups were analyzed separately, using SAS, Version 9.2. Three research questions were addressed, using tests of comparison groups (chi-square tests for binary variables and categorical groups and t-tests for numerical variables) and bivariate testing and regression (t-tests for binary variables and ANOVA, with a follow-up Tukey-Kramer procedure for categorical predictors). The correlation between 2008 and 2009 reading scores also was determined. The results of this study indicate that there is no evidence to support the claim that school choice significantly affects reading achievement. Students in both the School Choice and non-School Choice groups made statistically comparable gains on the Reading FCAT from 2008-09. Additionally, economically disadvantaged students, ELL students, and students from certain racial groups (African American and Hispanic) transferred at significantly lower rates than did students from comparison groups (non-economically disadvantaged, non-ELL, and Caucasian). The findings show that students who took advantage of the opportunity to transfer included those who were the least disadvantaged.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Brian Castellani.
Thesis: Thesis (Ed.D.)--University of Florida, 2011.
Local: Adviser: Behar-Horenstein, Linda S.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2013-06-30

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2011
System ID: UFE0043647:00001


This item has the following downloads:


Full Text

PAGE 1

1 THE IMPACT OF SCHOOL CHOICE ON STUDENT READING ACHIEVEMENT IN THE NO CHILD LEFT BEHIND (NCLB) ERA By BRIAN D OUGLAS CASTELLANI A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF EDUCATION UNIVERSITY OF FLORIDA 201 1

PAGE 2

2 201 1 B rian Douglas Castellani

PAGE 3

3 To my family for their constant support

PAGE 4

4 ACKNOWLEDGMENTS I am eternally grateful to my wife Jean, for allowing me to devote substantial time and energy to this pursuit. She was truly supportive from beginning to end. I appreciate my children, Kati and Colleen for supporting this endeavor and helping me to prepare for my comprehensive oral examinations I thank my parents, Roger and Kathleen, for always modeling the importance of hard work and academic achievement and my sister, Laurie, for demonstrating a superior work ethic few can match I am grateful to Dr. Linda Behar Horenstein whos e supervision and g uidance as my committee chair allowed me to complete this process I also thank Dr. Franc e s Vandiver and Dr. Thomas e n i a Adams for serving on my committee and providing support during each phase of the process I appreciated the patience shown by Dr. Sharon Bear as she assisted with the process of editing. I would like to show my gratitude to Dr. Mary Beth Thomas and Dr. Bruce Mousa for providing inspiration throughout the dissertation writing process. Finally, I am truly indebted to Dr. Cynthia Garvin for committing so much t ime to leading me through the process of data analysis

PAGE 5

5 TABLE OF CONTENTS p age ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 TABLE OF CONTENTS ................................ ................................ ................................ .. 5 LIST OF TABLES ................................ ................................ ................................ ............ 7 LIST OF ABBREVIATIONS ................................ ................................ ............................. 8 ABSTRACT ................................ ................................ ................................ ..................... 9 CHAPTER 1 I NTR O D UCTION ................................ ................................ ................................ ..... 12 Purpose of the Study ................................ ................................ .............................. 14 Significance of the Study ................................ ................................ ........................ 15 Definition of Terms ................................ ................................ ................................ .. 16 2 REVIEW OF THE LITERATURE ................................ ................................ ............. 19 Federal Interest in Education ................................ ................................ .................. 19 Ba ckground of No Child Left Behind (NCLB) ................................ .......................... 20 No Child Left Behind School Choice ................................ ................................ ....... 25 Characteristics of Choosers ................................ ................................ .................... 27 School Choice and Student Achievement ................................ ............................... 29 NCLB School Choice, Student Achievement, and Demographics .......................... 32 Florida Comprehensive Assessment Test (FCAT) ................................ .................. 34 Summary ................................ ................................ ................................ ................ 37 3 METHODOLOGY ................................ ................................ ................................ .... 39 Theoretical Fram ework for Data Collection and Analysis ................................ ........ 39 Research Methodology ................................ ................................ ........................... 40 Researcher Qualifications and Bias ................................ ................................ ........ 42 Participants ................................ ................................ ................................ ............. 42 Access and Entry ................................ ................................ ................................ .... 43 Variables of the Study ................................ ................................ ............................. 43 Instrumentation ................................ ................................ ................................ ....... 44 Validity and Reliability of the FCAT ................................ ................................ ......... 44 Validity ................................ ................................ ................................ .............. 44 Reliability ................................ ................................ ................................ .......... 45 Assumptions and Limitations ................................ ................................ .................. 45 Assumptions ................................ ................................ ................................ ..... 45

PAGE 6

6 Limitations ................................ ................................ ................................ ........ 46 Summary ................................ ................................ ................................ ................ 46 4 RESULTS AND ANALYSI S OF DATA ................................ ................................ .... 48 Descriptive Analysis ................................ ................................ ................................ 48 Data Analysis ................................ ................................ ................................ .......... 49 Research Question 1 ................................ ................................ ........................ 49 Research Question 2 ................................ ................................ ........................ 52 Research Que stion 3 ................................ ................................ ........................ 54 Summary ................................ ................................ ................................ ................ 57 5 CONCLUSIONS ................................ ................................ ................................ ...... 59 Summary of Results ................................ ................................ ................................ 59 Limitations of the Study ................................ ................................ ........................... 62 Implications ................................ ................................ ................................ ............. 63 Recommendations ................................ ................................ ................................ .. 64 Conclusions ................................ ................................ ................................ ............ 65 APPENDIX A U NIVERSITY OF FLORIDA INSTITUTIONAL REVIEW BOARD (UF IRB ) APPROVAL ................................ ................................ ................................ ............ 68 B SCHOOL DISTRICT APPROVAL ................................ ................................ ........... 69 LIST OF R EFERENCES ................................ ................................ ............................... 70 BIOGRAPHICAL SKETCH ................................ ................................ ............................ 76

PAGE 7

7 LIST OF TABLES Table P age 2 1 One Year Gains in Florida Comprehensive Achievement Test ( FCAT ) Developmental Scale Scores in Reading as set by the state of Florida .............. 36 2 2 Range of FCAT Developmental Scale Scores for each Achievement Level in Grade 4 Reading as set by the state of Florida ................................ .................. 36 4 3 2008 Comparing Groups on Reading Achievement ................................ ............ 52 4 4 Predictors of Reading Achievement in 2008 ................................ ....................... 53 4 5 Predictors of Reading Achievement in 2009 ................................ ....................... 55 4 6 2009 Comparison Groups for Reading Achievement ................................ .......... 57

PAGE 8

8 LIST OF ABBREVIATION S AYP Adequate Yearly Progress. An individual state's measure of progress toward reading/language arts and math proficiency. Technically it refers to the minimum level of proficiency that the state, school districts, and schools must achieve each year on annual tests and relat ed academic indicators (U.S. Department of Education, 2009 a ). FCAT Florida Com prehensive Achievement Test. of student achievement that is administered to students in grades 3 11 across the state. The FCAT measures student performan ce on the Sunshine State Standards in reading and mathematics (grades 3 10), science (grades 5, 8, and 11), and writing (grades 4, 8 and 10 ; Florida Department of Education, 2005 b ). LEA Local Education Agency. A public board of education, other public in stitution or agency that holds administrative cont r ol of and provides direction for a public school (Florida Department of Education, 2005 b ). NCLB No Child Left Behind The Elementary and Secondary Education Act (ESEA), reauthorized as the No Child Left Behind Act of 2002. This mandate is built on four principles : accountability for results, more choices for parents, greater local control and flexibility, and an emphasi s on doing what works based on scientific research (U.S. Department of Education, 2009 a ). SEA State Education Agency. The State Board of Education or other agency or officer primarily responsible for the supervision of public schools in a state (Florida Department of Education, 2005 b ). SSS Sunshine State Standards. Educational standards approved by the State Board of Education in 1996 for student achievement in Florida. The standards cover seven subject areas, with each area divided into four separate gr ade clusters (PreK 2, 3 5, 6 8, and 9 12 ; Florida Department of Education, 1996).

PAGE 9

9 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Docto r of Education THE IMPACT OF SCHOOL CHOICE ON STUDENT READING ACHIEVEMENT IN THE NO CHILD LEFT BEHIND (NCLB) ERA By Brian D ouglas Castellani December 2011 Chair: Linda Behar Horenstein Major: Educational Leadership When No Child Left Behind (NCL B) became law, it was clear that government leaders wanted to align high stakes testing with school, district, and state accountability. The goal was for all students to be successful as defined by NCLB An important part of this law is for all students t o h ave the choice to attend higher performing schools if their zoned school is deemed underperforming. For the first time, legislation provided this option to everyone. Since the choice option was enacted few studies have explored the relationship between school choice and reading achievement. The purpose of this study was to determine whether there is a relationship between school choice and student achievement for selected subgroups (gender, race, English Language Learner [ELL] status, Exceptional Studen t Education [ESE] status, economically disadvantaged status ) and subtests (Overall Reading Score, Information subtest, Literature subtest, Words and Phrases in Context subtest, Main Idea, Plot, and Purpose subtest, Comparisons and Cause/Effect subtest, and Reference and Research subtest) on the fourth grade Reading Florida Comprehensive Achievement Test (F CAT ) This study also determined which factors predicted reading achievement during the 2008 and 2009 FCAT administrations

PAGE 10

10 T he t hird and fourth grade Reading FCAT Developmental Scale Scores of 953 students w ere collected from one Southwest Florida school district. Student data w ere separated into tw o groups (School Choice and non School Choic e) and further separated by subgroups. Data from subgroups w er e analyzed separately using SAS V ersion 9.2. Three research questions were addressed, using tests of comparison groups (chi square test s for binary v ariables and categorical groups and t tests for numerical variables) and bi variate testing and regression ( t tests for binary variables and ANOVA with a follow up Tukey Kramer procedure for categorical predictors). The correlation between 2008 and 2009 reading scores also was determined The results of this study indicate that there is no evidence to support the claim that school choice significantly affe cts reading achievement. Students in both the School Choice and non School Choice groups made statistically comparable gains on the Reading FCAT (overall and subtests) from 2008 2009. School Choice students o utperformed non School Choice students at a statistically significant level on all subtests in both 2008 and 2009 Students classified as Caucasian, non economically disadvantaged, and non ELL scored at a statistically significant higher rate than economic ally disadvantaged, African American, Hispanic, Haitian Creole, and active ELL in both years. While all subgroups are making comparable gains it is clear that the achievement gap is not closing. Additionally, economically disadvantaged students ELL students, and students from certain racial groups ( African American and Hispanic) transferred at significantly lower rates than did students from comparison gro ups (non economically disadvantaged non ELL and Caucasian ). The findings show that

PAGE 11

11 students wh o took advantage of th e opportunity to transfer included those who were the least disadvantaged. This is in direct contrast to the original intent of NCLB.

PAGE 12

12 CHAPTER 1 I NTRODUCTION Theoretically, parents always have been able to choose which schools their children attend The pre dominant form of choice in the United States choosing their residence as a means to ensure that their children can attend a certain neighborhood school (Hoxby, 2002). This pattern of choice however, has resulted in an unequal system in which high income parents routinely have the opportunity to exercise more choice because they have the ability to reside in preferab le school zones (Hoxby, 1998). Moreover, a ffluent parents have the option to send their children to private schools and/or to provide home schooling if the zoned public school d oes not meet their needs (Hall, 2010 ). In January 2001, President George W. Bush signed No Child Left Behind (NCLB) into law. This legi slation attempted to align high stakes testing with specified accountability measures, the latter of which were designed to ensure that all students reach minimum proficiency targets (Smith, 2005). An i mportant component of NCLB is ensuring that all students have an opportunity to attend higher performing schools if their zoned school is designated as needing improvement (Simpson, LaCava, & Graner, 2004). T heoretically t his option provides parents with an opportunity to think about what at kind of education they want fo r their child (Neild, 2005). In this regard, Bast and Walberg (2004) contend that parents can do a better job of choosing schools for their children than can g overnment agencies. Prior to this legislation, this option was not available to all parents NCLB requires states and districts to offer public school choice at the same time t hat it notifies parents that a school had been identified for school improvemen t,

PAGE 13

13 corrective action, or restructuring (U.S. Department of Education, 2009 c ). This information is required to be in an easily understood fo home language so that all families understand their rights Additionally school dis tricts are responsible for providing or paying for transportation so that students can attend the school of their choice (U.S. Department of Education, 2009 d ). Thus if the neighborhood school is designated as underperforming, legislation provides all stud ents in that school zone have the opportunity to attend a higher performing school. T he mandate has caused market forces to be introduced into public education ( Hastings, Van Weelden, & Weinstein, 2007). NCLB also requires that schools are evaluated yearl y and that subgroups meet state proficiency targets. These subgroups included majo r ethnic/racial groups (African American, Asian/Pacific, Caucasian, Hispanic, and Native American) economically disadvantaged students, English Language Learners (ELL) and students with disabilities (U.S. Department of Education, 2002). The rationale for this mandate is that it offers schools that are in danger of failing an opportunity to imp rove their academic performance and to avoid political embarrassment and pote ntial revenue losses from student attrition (Gree n e 2001). When considering school choice from a policy perspective one needs to ask two important questions: First, how will school choice affect school quality ? (H astings, Kane, & Staiger, 2005). Second, how will school choice affect student achievement ? (Hoxby, 2002) R esearch on school choice has showed unexpected results. For example, not all studies show a correlation between school choice and improved student achievement. R esearchers have reported th at there is no significant difference in reading achievement between students who chose to transfer and those who stay in

PAGE 14

14 neighborhood schools designated by state and federal measures as needing to improve ( Hall, 2010; Kirkland, 2009; Phillips, Hausman, & Larsen, 2009; McCombs, 2007; U.S. Department of Education, 200 8 ). Additionally, Nicotera, Teasley, and Berends (2007) found that students who elected n ot to transfer to other schools initially outperformed those who transferred. Overall, researchers have c onsistently recommended that additional research is needed to determine the relationship between student achievement and school choice. Purpose of the Study The purpose of this study was to determine whether school choice affe cts student reading achievement. As noted NCLB provides the opportunity for students at failin g schools to transfer to higher performing schools ( U.S. Department of Education, 2003) T here is a gap in the literature however, related to the success of NCLB school choice prog rams as measured by student reading achievement. This study analyzed reading test scores for students who used the NCLB school choice option versus those who remained in their zoned schools. The findings from this study provide school and district decision makers with an opportunity to consider the effectiveness of school choice in relation to student reading achievement. Specifically, the purpose of this analysis was to assess the impact of the school choice option between 2007 to 2008 and 2008 to 2009 in one suburban Southwestern Florida district. The following research questions were addressed. What is the relationshi p between School Choice and non School Choice when compared by gender, race, English Language Learner ( ELL ) status, Exceptional Student Edu cation ( ESE ) status, economically disadvantaged status Information subtest, Literature subtest, Words and Phrases in Context subtest, Main Idea, Plot, and

PAGE 15

15 Purpose subtest, Comparisons and Cause/Effect subtest, and Reference and Research subtest of the Rea ding FCAT? What factors predict ed reading achievement in 2008? What factors predict ed reading achievement in 2009? The following null hypotheses guided the investigation: H o 1 : There will be no statistically significant relationships between Sc hool Choice and non School Choice when compared by gender, race, ELL status, ESE status economically disadvantaged status Information subtest, Literature subtest, Words and Phrases in Context subtest, Main Idea, Plot, and Purpose subtest, Comparisons and Cause/Effec t subtest, and Reference and Research subtest as measured by the FCAT D evelopmental Scale Score in 2008 and 2009. H o 2 : There will be no statistically significant relationships between the control variables of studen t gender, race/ethnicity, economically disadvantaged status, ES E status and ELL status and overall reading mean scores in 2008. H o 3 : There will be no statistically significant relationships between the control variables of studen t gender, race/ethnicity, economically disadvantage d status ESE status and ELL status and overall reading mean scores in 2009. Significance of the Study A priority of NCLB was to give parents choices about where their children attend school. Although 6.9 million students were eligible for school choice between 2002 and 2007, only 1% chose this option (U.S. Department of Education, 2009 d ). The relationship between NCLB school choice and its impact on student achievement has not been clearly established (Kirkland, 2009). The State of Florida, in particular appears

PAGE 16

16 to lack a system for evaluating the implementation and effectiveness of NCLB school choice (Hall, 2010). This study analyzed the results of the FCAT reading test for 4 th grade students who elected to use NCLB school choice and compared their scor es to those students who wer e eligible for school choice but remained at their zoned schools. These results are likely to allow school personnel, districts, and state departments to make more informed program decisions for schools and individual students a nd to begin exploring program effectiveness i n reading within choice and non choice schools. Definition of Terms To provide an understanding of the concepts and terms related to this study, the following definitions are provided: Adequate Yearly Progress (AYP) : An individual state's measure of progress toward reading/language arts and math proficiency. Technically it refers to the minimum level of proficiency that the state, school districts, and schools must achieve each year on annual tests and related academic indicators (U.S. Department of Education, 2009). Developmental scale score : This score is a measure of student learning as a student move s from one grade level to the next (Florida Department of Education, 2010). Economically d isadvantaged : Refe rs to students who are members of households that meet the income eligibility guidelines for free or reduced price meals under the National School Lunch Program (Wisconsin Department of Public Instruction, 2009). Florida Comprehensive Achievement Test (FCA T) : student achievement that is adminis tered to students in grades 3 11 across the state.

PAGE 17

17 The FCAT measures student performance on the Sunshine State Standards in reading and mathematics (grades 3 10), science (grades 5, 8, and 1 1), and writing (grades 4, 8, and 10 ; Florida Department of Education, 2005b ). Local Education Agency (LEA) : A public board of education, other public institution or agency that holds administrative control and direction of a public school (Florida Depart ment of Education, 2005 a ). LF : A code used by the state of Florida to identify a student who has exited an ELL program who is followed for a two year period beyond the program (Florida Department of Education, 1994). LY: A code used by the state of Florid a to identify a student classified as limited English proficient and who receives specific instruction services that are designed for ELL students, regardless of model/approach that is used to deliver these instructional services (Florida Department of Education, 1994). LZ: A code used by the state of Florida to identify a student who has completed the two year follow up period after exiting the ELL program (Florida Department of Education, 1994). No Child Left Behind (NCLB) : The Elementary and Secondary Education Act (ESEA), reauthorized as the No Child Left Behind Act of 2002. This legislative mandate is built on four principles: accountability for results, more choices for parents, greater local control and flexibility, and an emphasis on doing what works based on scientific research (U.S. Department of Education, 2009 a ). No Child Left Behind (NCLB) School Choice : Refers to schools that receive federal Title I funds that have not made state defined AYP for two consecutive schoo l

PAGE 18

18 years and are now designated as needing improvement. Students in these identified schools must be given the option to transfer to another public school (one that has not been identi fied as in need of improvement) and the school district is required to provide the transportation (U.S. Department of Education, 2009 a ). State Education Agency (SEA) : The State Board of Education or other agency or officer primarily responsible for the su pervision of public schools in a state (Florida Department of Education, 2005 a ). Sunshine State Standards (SSS) : Educational standards approved by the State Board of Education in 1996 that provide expectations for student achievement in Florida. The stand ards cover seven subject areas, with each area divided into four separate grade clusters (PreK 2, 3 5, 6 8, and 9 12 ; Florida Department of Education, 1996). Title 1 : Funding authorized under th e ESEA provision that all children have a fair, equal, and significant opportunity to obtain a high quality education (U.S. Department of Education, 2009 b ). ZZ: A code used by the state of Florida to identify a s tudent who has never received ELL services (F lorida Department of Education, 1994).

PAGE 19

19 CHAPTER 2 REVIEW OF THE LITERA TURE This review provides an overview of the literature that is relevant to this study including federal interest in education; background of No Child Left Behind ( NCLB ) ; NCLB school choice; characteristics of choosers; school choice and student achievement; NCLB school choice, student achievement and demography; and the Florida Comprehensive Achievement Test ( FCAT ) The review shows that the relationship between the bodies of literature supports the need for an analysis of whether the NCLB school choice option affe cts student achievement. Federal Interest in Education Throughout the educational history of the United States, public school systems ha ve been regulated by individual states and t he federal government has provided funding to states and local districts to motivate their participation in selected federal programs. One policy, Title 1 of the Elementary and Secondary Education Act ( ESEA ) was designed to motivate states and local school districts to focus on specific student groups and to increase the services e.g., pull out remedial programs and additional staff working directly with the identified students provided to them. Through this pol icy, which offered decision makers financial assistance if they undertook certain prescribed activities such as supplemental services in pull out classes the federal government strove to influenc e state and local programs (McDonnell, 2005). The federal go are driven by decades of stagnating or declining scores on two key nationally administered tests, the Scholastic Aptitude Test (SAT) and the National Assessment of Educational Prog ress (NAEP; Fusare lli, 2004). Between 1963 and 1980, the average verbal and

PAGE 20

20 math scores on the SAT declined by 50 and 40 points, respectively. These results were well below comparable scores of foreign students which contributed to federal s (Heise 1999). Additionally reading and mat h scores on NAEP have shown little improvement since 1975, despite a tenfold increase in federal spending on ESEA (Fusarelli, 2004). Moreover student performance on international comparative tests fell below that of o ther nations, reaching back as far as the First International Math ematics and Sc ience Study in the 1960 s. Lawmakers reacted to these results by increasing federal involvement in education. Since the passage of ESEA, there has been little accountability for di stricts and schools who receive federal funding. Historically, s chools earned funding based solely on numbers of eligible students, with little focus on whether the funding improved the performance of targeted populations (Fusarelli, 2004). To increase accountability for student achievement, recent elementary and secondary education reform proposals such as NCLB require school competition to raise test scores and to expand parental choic e. These reforms have created an expectation that schools will impro ve student performance rather than risk losing students or facing federal sanctions (Smith, 2005). Background of No Child Left Behind (NCLB) No Child Left Behind Act date back to the 1954 Brown v. Board of Education case when the U. S Supreme Court outlawed racial segregation in (U.S. Department of Education, 2004 p 13) The passage of this act evolved from over 50 years of related legislation. In 1964, President Lyndon Johnson formed the Gardner Commission. The intent of this federal commission was to make policy recommendations to the president in regard to federal educa tion aid to schools. The

PAGE 21

21 committee recommended ESEA, which was approved by Congress on April 11, 1965. The purpose of Title 1 of ESEA was to motivate states and local school districts to allocate resources to targeted groups and to increase the level and q uality of services provided to these students (McDonnell, 2005). Title 1 represented the largest financial component of the ESEA legislation (Thomas & Brady, 2005). The evolution of Title 1 followed three distinct phases over the next 45 years. The first p hase (1965 1980) followed Brown v. Board of Education During that time, Title 1 program funds were dedicated to economically disadvantaged students in need of services such as pull out remedial programs In most schools, these programs operated separately from the main instructional program (McDonnell, 2005). During the second phase (1981 1987), the Education Consolidation and Improvement Act of 1981 (ECIA) was passed. Not long after, A Nation at Risk (1983) published by the Reagan administration empha sized that educational standards needed to be raised and accountability increased (Smith, 2005) This publication promoted the belief that public schools were unsuccessfu l based on falling or stagnating scores on the SAT and NAEP ECIA, part of President Omnibus Budget Reconciliation Act of 198 7 reduced federal funding across most domestic policy areas, including ESEA. Under this act, Title 1 was renamed Chapter 1 (Thomas & Brady, 2005). Although compensatory educational services for economically disadvantaged students, significant reductions and relaxed regulatory requirements reduced student eligibility for services (Thomas & Brady, 2005).

PAGE 22

22 The third phase of the Title 1 evolution signified by the 1988 ESEA reauthorization, required states to define academic achievement criteria. State agencies were required to submit plans that detailed the use of high level content, student performance standards, state assessments and annual reports on the attainment of standards. In addition, plans for supporting teachers and students, aligned with updated curriculum standards and specified assessment instruments, were part of the documenting requirements (McDonnell, 2005). During this time, state adoption of accountability structures became more standard. By the late 1980s and early 1990 s, a majority of states began to use test results to hold schools accountable for student performance (Linn, 2005). Although the emphasis on raising standards could be t race d back to at least the 1980 s, federal interest in public school a ccountability, paired with high stakes testing and components of school choice, led to the creation and passage of NCLB (Smith, 2005). One piece that was missing from all versions of ESEA was direct accountability for student performance. This changed in January 2002, when President George W. Bush signed into law NCLB Public Law 107 110. This act linked high stakes testing with specified accountability measures to ensure that students in scho ols that received federal funding would show evidence of success (Smith, 2005). Although this a ct altered the requirements of ESEA, it continued the practice of distributing federal funds to school districts with lower economic and multicultural student po pulations (Arce, Luna, Borjian, & Conrad, 2005). NCLB guidelines stipulated that states use Adequate Yearly Progress ( AYP ) as the measure to hold schools accountable. In January 2003, as required, each state

PAGE 23

23 submitted to t he U.S. Department of Education a detailed plan that outlined the steps that they would take to ensure compliance with the statutes set forth under NCLB. By 2005, states were required to annually assess student performance in grades 3 to 8 in language arts, literacy, and mathematics. Stat es also had to indicate how individual schools and school districts would demonstrate AYP as well as make public their test results (Smith, 2005). AYP requires that all subgroups, including majo r ethnic/racial groups (African American, Asian/Pacific, Cauca sian, Hispanic, and Native American) economically disadvantaged students, ELL students, and ESE students reach specified proficiency levels (U.S. Department of Education, 2002). If all the AYP targets were not met, the entire school failed (Fusarelli, 20 04). States set criteria for the percentage of students who were expected to meet proficiency in math and reading each year. These percentages increased each year. By 2014, all students are expected to meet the standard (Florida Department of Education, 20 06). Although NCLB has expanded federal reporting requirements for states and local districts, states still retain control of many important decisions relat ed to NCLB compliance, such as selecting content standards, choosing assessments, and setting acade mic achievement levels. M ost states had state testing and accountability systems prior to 200 1 Moreover most states had worked for several years to coordinate curriculum, teacher professional development, and testing requirements. Few states abandoned th eir established accountability system when the new federal requirements were enacted. Instead, they added another, often disconnected, system of accountability on top of the state approved system currently in place (Sunderman & Kim, 2004).

PAGE 24

24 Because each st ate set its own performance standards there were often wide discrepancies between 2003, Florida was last in the nation with only 18 % of schools meeting AYP There was a significant discrepancy between the state accountability system, which assigned letter grades to schools based on performance, and the mee ting of federal AYP requirements. For example, 56% of schools in Florida m et federal AYP requirements (Linn, 20 0 5). NCLB included provisions for increased accountability the use of scientific research to make educational decisions, and an increase in parental authority when school decisions significantly affect ed their children (Simpson et al., 2004). The goal of opportunity to obtain a high quality education and reach, at a minimum, proficiency on challenging state academic achievement standards and state (U.S. Department of Education, 2009a, p. 1). To provide detailed information to parents, provided information about the quality of education at th Department of Education, 2008). Included in the report card were student achievement data broken down by AYP subgroup (U.S. Department of Education, 2008). Educational researchers h ave questioned whether NCLB expanded the federal r ole through legislating additional requirements and sanctions or whether the legislation was not specific enough since states set their own performance standards. According to Sunderman & Kim (2007) NCLB has expanded federal power to regulate education d ue to mandated AYP proficiency targets and federal selection of subgroups. B y

PAGE 25

25 allowing each state to determine levels of proficiency, some states have lowered their educational standards to ensure that their schools can attain AYP (Fusarelli, 2005). No C hild Left Behind School Choice NCLB was the first means by which parental school choice was enacted. Prior to this, school choice plans had been district or state oriented and had loose ties to accountability as measured by standardized testing. The goal of previous state or local school choice plans was to improve both educational quality and equity by providing incentives for schools to compete for students. Additionally, disadva ntaged students were afforded the opportunity to attend high quality public schools (Hastings et al., 2007). In the decade before NCLB became law, students in a number of large cities in the United States (New York, Chicago, and Philadelphia) already had a ccess to extensive school choice programs (Neild, 2005). Unfortunately, a large percentage of disadvantaged students did not participate in previous school choice programs, e.g., in New York, Chicago, and Philadelphia because school choice information wa s not available or they could not access the information in a form that they could understand. Clearly school choice initiatives must include provisions for reaching out to diverse populations of parents who had not traditionally considered these options. Without this effort, school choice programs often failed the disadvantaged students for which they were created (Howell, 2006). NCLB stipulated that children who attended Title 1 schools that failed to meet AYP requirements in two consecutive years had the opportunity to transfer to another public school (Arce et al., 2005). Giving parents the option to tr ansfer their children to higher performing schools was considered one way to force failing schools to meet federal

PAGE 26

26 AYP standards as well as individual state accountability standards (Betebenner, Howe, & Foster, 2005). NCLB provided a clearly defined relationship between school choice and test based accountability. Prior to NCLB, choice and accountability were relatively distinct from one anothe r ( Betebenner et al., 2005). State and district school choice plans that were initiated prior to NCLB allowed movement of students without measuring the effects of the moves on school and individual student standardized testing performanc e. It was unclear whether there were a high number of students from disadvantaged backgrounds who pursued the option to switch schools and whether these plans were helping these students In response, NCLB required states to develop procedures for ranking sch ools and to provide detailed reports of school performance to parents (Bast & Walburg, 2004). School choice is the responsibility of the Local Education Agency (LEA) who is charged with notifying parents that a school ha s been identified for school improv ement, corrective action, or restructuring (U.S. Department of Education, 2009b). The LEA was required to provide this in formation to parents in an easy to understand format. Once a student elected to transfer to a higher performing school, the LEA was req uired to allow the transferred student to stay in the selected school until he or she had completed the highest grade. Certain circumstances such as when transportation funds are limited or first choice options are closed, require that an LEA g ive priorit y to the lowest achieving students from low income families (U .S. Department of Education, 2009 d ). The guidelines set forth in NCLB were clearly defined for the LEA. Unless a lesser amount is needed, an LEA must spend up to an amount equal to 20% of its Ti tle 1, Part

PAGE 27

27 A allocation to meet NCLB requirements. Specifically 5% must be allocated for choice related transportation, 5% for supplemental educational services, and 10% for a combination of transportation and supplemental services (U.S. Department of E ducation, 2003). A limitation of school choice is that it penalizes schools and their communities when student s or guardians, specifically low income and speakers of other languages, often lack the resources of ti me and knowledge of how the educational and governmental system s oper ate. This often results in economically disadvantaged students attending schools with higher percentages of economically disadvantaged student s than prior to the enactment of NCLB (Arce et al., 2005). Characteristics of Choosers The consistently low number of students who use the school choice option has raised many questions. One such question is: Why do students choose to move, and what is it about these students that make them follow t hrough? The Chicago Public Schools (CPS) have ha d extensive school choice programs for many years. There were several competing explanations for why students in this system who opte d out of their assigned schools outperformed those who stayed. Distance to nearby schools was a strong predictor of the likelihood that a student would opt out of the assigned school (Cullen, Jacob, & Levitt, 2005). Evidence from the CPS study which included a sample of 60,623 students in grades eight through twelve over a period of three years, suggests that academic increases for students who used the choice option were due to motivation level and parent al involvement (Cullen et al., 2005 ). Interestingly, boys

PAGE 28

28 (specifically African Am erican males) were significantly less likely to graduate from high school than females. Economically d isadvantaged students who use school choice most often have parents who received school choice information in an easy to understand format, as required by what schools to send their children to were not uniformly directed by test scores alone. This was true for all parents, not just those who used the school choice option (Hastings et al., 2007). Other factors th at influenced parents were school reputation, safety, location, and curriculum (Betebenner et al., 2005). A study of the Charlotte Mecklenburg Public School District which included a sample of 6,328 non school choice students and 10,134 NCLB school choice students in grades five and eight concl uded that simplified information affects for African American and economically disadvantaged students (Hastings et al., 2007). Parents of children who qualified for sc hool choice provisions purport to care about the same features of schools, foremost among them academics, as do parents whose children already attend higher performing public schools (Howell, 2006). When selecting an alternative public school for their child, parents in underperforming schools consistently choose schools with more advantaged and higher performing students. They seek the perceived advantages of surrounding their children with high achieving peers (Howell, 2006). Parents put academic achievement at the t op of their list of term interest s (Bast & Walberg, 2004). Providing clear information on school test score s

PAGE 29

29 within a choice plan increases the proportion of parents choosin g higher scoring schools for their children (Hastings & Weinstein, 2008). Some parents choose schools based on location (convenience) and extracurricular activities (Bast & Walberg, 2004). The impact of information on choice behavior was greatest for fami lies with higher scoring schools in relative close proximity, implying that providing informatio n i s most effective when parents have a quality alternative within a reasonable distance ( Hastings & Weinstein, 2008). The c hoice provisions of NCLB have introd uced new competitive pressures in public education and have empowered disadvantage d parents to shape the educational lives of their children. School Choice and Student Achievement Prior to NCLB, several school districts already had experimented with school choice. When t he relationship betwee n school choice and student achievement is studied, the results are inconsisten t Results from the school choice program in CPS which allows choice for at least half of the students in the district, indicated som e positive effects for students. In this study, Cullen et al. (2005) established a correlation between opting out of the assigned school and choosing another school and higher educational attainment. In contrast, Betebenner et al. (2005) reported that ther e was no uniform benefit for those students who participat ed in choice in a large western school district with an extensive school choice program. The sample in this study included 402 fifth and sixth grade students Thus, the belief that allowing choice w ould help academic achievement was not supported. The only students who showed a benefit from choice were in the lowest quartile growth and this was observed only in mathematics Also, it was noted that Caucasian and Hispanic parents elected school choice at a comparable rate (Betebenner et al. 2005) Hastings et al. (2007) reported that African American

PAGE 30

30 students were about 15 percentage points less likely to elect school choice than comparison subgroups and that the decision to transfer did not improve p erformance for this subgroup. After conducting field experiments on 115,716 students in 2003 2004 and 125,313 students in 2005 2006, Hastings and Weinstein (2008) noted that transferring a disadvantaged student to a school with high mean t est scores did not result in statistically significant academic gains for that student This was true for African American and economically disadvantaged students. A ttending a higher performing school had a positive impact when coupled with parents who wer e informed and were seeking im proved academic achievement for their child (Hastings & Weinstein, 2008). Schools of choice tended to out p erform public schools on a wide range of outcomes and for children from all socioeconomic backgrounds, meaning that the act of choosing, for whatever reason, led to children attending better schools (Bast & Walburg, 2004). In a study of six districts with a sample size of 4,155 students there was no observable achievement gain across content areas as a result of changin g schools for the total population Achievement results for African American Hispanic and students with disabilities showed mixed results across the districts Some showed statistical significance, but no clear patterns emerged overall for any particular subgroups (Zimmer, Gill, Razquin, Booker, & Lockwood, 2007) In a study of 10,372 third through sixth graders who participated in an intra district school option, Phillips et al. (2009) compared reading achievement for those who switched schools versus those who stayed in neighborhood schools. The r esearchers concluded that for a majority of students, the choice to transfer had no impact on

PAGE 31

31 reading achievement. They also found that students who switched schools were less likely to be Hispanic, economic ally disadvantaged, or English Language Learners. However, they did find gains for s tudents who chose to move away from a low per forming school to attend a high performing school. The choice of school was found to be more important than simply the choice t o switch schools. Just switching to another school of similar achievement reportedly does little to improve student performance (Phillips et al., 2009). In a North Carolina study on intra district choice and student achievement, Okpala, Bell, and Tuprah (2 007) found significant differences in achievement scores of students in schools of choice compared to those in traditional schools, on end of course reading tests at the middle school level. Based on mean scores of 48 students in grades six to eight the schools of choice had higher mean scores than did traditional schools in reading achievement based on end of grade tests (Okpala et al., 2007). The study also found that African American and economically disadvantaged students were less likely to elect cho ice than other subgroups. Hoxby (2002) reported that inter district choice had a positive, statistically significant effect on achievement for eighth grade students across a district. The district scored 3.8 national percentage points higher in reading fo llowing a 1% increase in inter district choice. When considering large groups and testing results, however, reallocation of students had little impact on the overall state or district average. Rather, it affected only the distribution of achievement across schools. When considering achievement test mean scores, s truggling schools sco re lower and thriving schools sco re higher (Hanushek, Kain, Markman, & Rivkin, 200 3 ). Thus, it remains unknown whether the overall choice policy improves district or state testi ng results.

PAGE 32

32 NCLB School Choice, Student Achievement, and Demograph ics NCLB is an important legislative action because it has expanded school choice to students who attended schools that were not meeting state proficiency requirements. This notion of choic e was an important reason why NCLB was initiated, passed, and maintained. Supporters o f school choice policies assume that al lowing parents to choose higher performing public school for their children will improve academic performance (Nicotera et al., 2007). Using a survey and semi structured interviews with 19 parents of eighth graders Neild (2005) found that parents believe that the school choice system offers their children the opportunity to attend schools with advantages not available at their previous schoo l These advantages include higher performing peers, less disciplin ary referrals and additional extracurricular activities. When selecting an alternative public school for their child, parents consistently selected schools with more ad vantaged and higher performing students (Howell, 2006) In a case study of 264 school choice and 1,907 non school choice elementary and middle school students in one district where NCLB school choice was offered, McCombs (2007 ) f ound that transferring did not positively affec t performance or absolute gains in reading, when controlling for prior achievement and student background characteristics. However, students who performed below the mean gained more from transferring schools than t hose who scored above the mean. Also, receiving schools were found to have a significantly higher rate of Caucasian students and lower rate of economically disadvantaged students than sending schools. I n a study of 62,628 second to tenth graders Nicotera et al. (2007) questioned whether students performed better when given the opportunity to transfer from low performing to high performing schools. These researchers found that students who

PAGE 33

33 moved under the NCLB choice provision experienced an initial decline in reading ac hievement during the first school quarter as measured by the Northwest Evaluation Association testing. The small boost in achievement growth that came later in the year did not compensate for the initial losses. Nicotera et al. (2007) also reported that s tudents who did not change schools had faster achievement growth rates compared to those who transferred. Thus, the authors concluded that the goals of NCLB school choice did not appear to work as intended. Researchers studied six urban school districts t o determine the relationship between elementary school choice and reading achievement. The r esearchers concluded that in all six districts there were no positive or negative statistically significant results (Zimmer et al. 2 007). Achievement results for African American and Hispanic students and for students with disabilities who participated in school choice varied ac ross districts. One district out of the six, however, did show statistically positive results in reading gains for school choice students. Overall, h owever, there were no effects on reading achievement across subgroups. The authors cautioned that the results were based on a small sample a nd should lead to further study (Zimmer et al., 2007). In a study of 55 tenth graders in one Florida dis trict, Hall (2010) found that there was no statistically significant relationship between the variables of ELL status ESE status gender, race, and economically disadvantaged status and reading achievement gains, as measured by the FCAT, for students who s elected or declined the school choice option. In addition, there was no statistically significant relationship between school choice and student achievement in reading f or students who transferred from

PAGE 34

34 their Title 1 schools, compared to students matched b y ELL, ESE, gender, race, and socioeconomic status who elected to stay in their neighborhood school. In a study conducted in another Florida school district, Kirkland (2009) matched 103 transfer students with 103 students who had similar demographic charac teristics who remained in their assigned schools. The results indicated that, after one year, there were no significant differences in reading achievement scores based on transfer status as compared to remaining in the assigned school. Also, this study rep orted that Caucasian students transferred at statistically significantly higher rates than Hispanic or African American students. Florida Comprehensive Assessment Test (FCAT) In 1996, educators in the state of Florida identified a core body of knowledge a nd skills that they felt students should have attained at each stage of their school career ( Florida Department of Education, 2004a) This body of knowledge and skills is detailed which include seven content are as: mathematics, language arts, science, social studies, foreign language, health and physical education, and the arts (Florida Department of Education, 2004b). The SSS is divided into four grade level clusters, PreK 2, 3 5, 6 8, and 9 12, which are furthe r divided into benchmarks for student knowledge and performance at each grade level. 2 p 2) During the same period, the Florida Educational Reform and Accountability Commission recommended that a statewide assessment system be developed to measure student progress as related to the SSS This recommendation resulted in the

PAGE 35

35 creation of the FCAT. The FCAT was designed to test the achievement of the SSS and benchmarks in grades 3 through 10. The test has a multiple choice format that includes performance tasks. Test questions require students in the upper grade levels to respon d in writing rather than simply selecting from the four choices offered for each question, to demonstrate concept attainment (Florida Department of Education, 2004b). The primary purpose of the FCAT is to assess st udent achievement of the higher order co gnitive skills outlined in the SSS in mathematics, reading, writing, and science. T he FCAT is a criterion referenced test (CRT) and the 2008 FCAT contained a second part that i s a nationally norm referenced Stanford Achievement Test (SAT 10) that i s used to compare the performance of Florida students in reading and mathematics to the rest of the nation. This norm referenced portion was eliminated after the 2008 FCAT administration Following administration and scoring, score reports are provided for school and student review. For the FCAT scores are reported on a scale from 100 500. In addition, scores are also reported as achievement levels that rang e from 1 to 5 with 5 being the highest (Florida Department of Education, 2004b). In addition to the scale scores and achievement levels, the FCAT also yields developmental scale scores for each test taker. These scores are designed to measure the learning gains of individual students as they progress through the grade levels. Developmental scale scores range from 0 to 3,000, and it is expected that th ese scores increase as student s progress through the grades. A third grader typically scores near the lower end of the scale, and a tenth grader scores near the upper end of the scale. Developmental gain scores a re often used to determine adequate student progress. The state has expected at least a one year gain for each grade level as presented in

PAGE 36

36 Table 2 1. The state has defined AYP for students based on three criteria: (a) an increase of one achieveme nt level from the previous year; (b) maintaining a level 3, 4, or 5 from the previous year; or (c) students at levels 1 and 2 demonstrating a one year gain based on developmental scale scores (Florida Department of Education, 2004b). Table 2 1. One Year Gains in FCAT Developmental Scale Scores in Reading as set by the state of Florida Grade Level Change Developmental Reading Scale Score Gain 3 rd to 4 th 230 4 th to 5 th 166 5 th to 6 th 133 6 th to 7 th 110 7 th to 8 th 92 8 th to 9 th 77 9 th to 10 th 77 The Reading FCAT SSS is divided into subtests. These include Information (25 questions), Literature (26 questions), Words and Phrases in Context (7 questions), Main Idea, Plot and Purpose (28 questions), Comparisons and Cause/Effect (13 questions), and Re ference and Research (3 questions ; Florida Department of Education, 2007). The results of the subtest scores are combined to provide the overall developmental scale score as shown in Table 2 2 (Florida Department of Education, 2007). Table 2 2. Range of F CAT Developmental Scale Scores for each Achievement Level in Grade 4 Reading as set by the state of Florida Achievement Level Developmental Scale Score Range Level 1 295 1314 Level 2 1315 1455 Level 3 1456 1689 Level 4 1690 1964 Level 5 1965 2638

PAGE 37

37 Summary The purpose of this study was to determine whether there is a relationship between the decision to elect school choice (following NCLB provisions) and student achievement. Previous research has demonstrated that in some cases, choice has result ed in increased achievement Such research also has found that parents choose to send their children to other schools when i nformation is shared in an easy to understand format test score information is understood and the receiving school has a positive reputation (Betebenner et al., 2005; Hastings & Weinstein, 2008; Hastings et al., 2007) and that students who chose to move to other schools tend to have proximity to the choice school are motivated to do well, and have involved parents (Cullen et al 2005; Hastings & Weinstein, 2008). Overall, however, the findings in regard to whether or not students who elect choice outperform those who remain in their zoned school are inconclusive For example, Okpela et al. ( 2007) found that students who moved to choice schools outperformed on end of year tests, stud ents who remained behind McCombs (2007) found that students who had scored below the mean showed the greatest improvement. In contrast, Kirkland (2009) and Hall (2010) found that moving to a choice school had no impact on achievement. McCombs (2007) also reported that choice had no impact on students who had initially scored above the mean. Finally, Nicotera et al. (2007) reported that students remaining at their zoned school had faster achievement growth rates than did those who transfer red to choice schools. Researchers have indicated that future studies should continue to focus on choice and its impact on achievement (Betebenner et al., 2005; Hall 2010) McCom bs (2007)

PAGE 38

38 and Kirkland (2009) recommend that future research should compare the achievement gain scores of NCLB transfer students to those who ha ve matched demographics and who remained in Title 1 schools. Typically, studies on school choice have been limi ted because they failed to examine the effects of choice policy on the students who remained in assigned neighborhood sc hools (Betebenner et al., 2005). Overall, there is a gap in the literature on the relationship between NCLB school choice and student achievement for certain subgroups In this regard, the current study, which focuses on fourth grade students, differs from that of Hall (2010) who focused on tenth grade rs The current study also di ffers from that of Kirkland (2009) because it provides a comparison of transfer and non transfer students, specifically in regard to each of the FCAT reading subtests.

PAGE 39

39 C HAPTER 3 METHODOLOGY This chapter presents the research design and methodology used in this study. The chapter includes the theoretical framework as it relates to the methodology of the study research methodology, researcher qualifications and bias, participants, access and e ntry, variables, instrumentation, validity and reliability of the Florida Comprehensive Achievement Test ( FCAT ) and assumptions and limitations. The chapter concludes with a summary. Theoretical Framework for Data Collection and Analysis The theoretical framework for this study is positivism. Positivism assumes that there are social facts with an objective reality apart from the belief of individuals. Quantitative researchers perceive truth as something that describes an objective reality, separate from t he observer and waiting to be discovered (Sale, Lohfeld, & Brazil, 2002). The positivist conception of science is rooted in the belief that objective reality exists and that it can be known only by objective means. This conception also proposes that s ubjectivity in knowledge does not lead to truth (Peca, 2000). According to Johnson (2009), virtually no quantitative research methodologists currently use the term positivist to refer to their work (and they have not for quite a long time). The term postp ositivism, rather than positivism, shows that a researcher is cognizant of changes that have occurred over the past 75 years in the philosophy of social science and research methodology. Most current researchers however, tend to hold more moderate positio ns with respect to varied research methodologies (Johnson, 2009). For this study, the researcher collected and analyzed student performance data to determine whether using school choice had an impact on student academic

PAGE 40

40 performance while remaining cogniza nt that there are factors which effect achievement that cannot be measured Research Methodology The researcher used descriptive statistics to calculate Reading FCAT scores for 4 th grade school choice students at receiving schools and for those who elected to remain at their zoned school. The Reading FCAT mean developmental gain scale scores were categorized by student demographic information including economically disadvantaged status ESE status gender, ELL status and race. This post hoc data set inclu ded the 2008 and 2009 Reading FCAT developmental scores. Additional variables included the FCAT subtests Information ; Literature ; Words and Phrases ; Main Idea, Plot and Purpose ; Comparison and Cause/Effect ; and Reference and Research. A level of significan ce of .05 wa s used, and all tests were two sided. All an alysis was performed using SAS, V ersion 9.2. Research question 1 required comparison group testing. First, chi square tests were used for the binary variables (male/female, ESE/no n ESE, and economically disadvantaged/non economically disadvantaged ). For the binary variables, the null hypothesis can be written as H o : N = S Here, = percentage of male/female, percentage of ESE (non gifted) /non ESE, and percentage of economically disadvantag ed/non economically disadvantaged (N = non School Choice and S = School Choice). Second, categorical groups (race/ELL status) were compared using chi square tests. For the categorical groups, the test of hypothesis can be written as H o : Group and (race/ELL ) are not related, and H a : Group and (race/ELL) are related. Third, numerical variables (overall reading mean/subscale scores) were compared using t tests. For the numerical variables, the test of the n ull hypothesis can be written as H o :

PAGE 41

41 N = S Here, = overall reading mean/subscale means (N = non School Choice and S = School Choice). Research questions 2 and 3 required the use of bivariate testing and regression analysis to identify predictors related to the reading achievement in 2008 and 2009. First predictors that were binary were tested using t tests. For the binary variables, the null hypothesis can be written as H o : 1 = 2 Here, = overall reading mean (1 = male, 2 = female; 1 = non economically disadvantaged 2 = economically disadvantaged ; 1 = non ESE, 2 = ESE (non gifted) ; 1 = non School Choice, 2 = School Choice). Second, categorical predictors were tested using Analysis of Variance ( ANOVA ) with a follow up multiple comparisons Tukey Kramer procedure. For the categorical predictors, the nu ll hypo thesis can be written as H o : 1 = 2 = K Here, = overall reading mean (1 = African American 2 = Haitian Creole, 3 = Hispanic, 4 = other, 5 = Caucasian and 1 = LF, 2 = LY, 3 = LZ, 4 = ZZ). Third, the correlation between 2008 Reading and 2009 Reading was computed using the formula H o : p = 0. Finally, research questions 2 and 3 required the use of multivariate testing (regression analysis). First, the equation for research question 2 can be written as Y = o + 1 (gender) + 2 (race) + 3 (ESE status ) + 4 (ELL status ) + 5 ( economically disadvantaged status ), and the null hypothesis can be written as H o : = 0. Here, Y = overall reading mean in 2008. Second, the equation for research question 3 can be written as Y = o + 1 (reading 2008) + 2 (gender) + 3 (race) + 4 (ESE status ) + 5 (ELL status ) + 6 ( economically disadvantaged status ) + 7 (group), Here, Y = overall reading mean in 2009.

PAGE 42

42 Researcher Qualifications and Bias The researcher has worked in a variety of school settings. Currently, the researcher serves as principal of a Title I elementary school. Students at this school have the opportunity to transfer to a higher performing school based on NCLB school choice requirements. revious work included serving as a principal of a receiving school. Both experiences made this researcher aware of the issue of choice and the question about whether changing schools makes a difference in student work with elementary and middle school students, the importance of peers, school environment, and the ir impac t on student performance was noted. graduate classes at Georgia State University, with a focus on inner city students, increased his i nterest in school choice. The classes that the researcher took in the University of Florida Educational Leadership doctoral program, such as Quantitative Research I and II and Research Design, led the researcher to realize that the relationship between sch ool choice and student achievement could be analyzed from a quantitative perspective. The researcher has worked as a sc hool administrator in both high performing and Title 1 schools and as such, did not enter the study with any bias Creswell (2008) remin d s us that researchers need to take an objective and unbiased approach when conducting quantitative research As a researcher and school administrator, the researcher has the experience to conduct and complete the proposed study. Participants The target p opulation for this study include s fourth grade students who selected the NCLB school choice option during the 2008 09 school year and attended a receiving school as well as those who were offered the option but elected to remain at their zoned

PAGE 43

43 school durin g this school year. The accessible population for t his study was elementary school aged students from a Southwest Florida public school system. This district is comprised of 29 elementary schools (PreK 5), 10 middle schools (6 8), 8 high schools (9 12), an d one smal l (under 1 75 students K 12 ) K 12 school. The student racial make up for the school distric t is approximately 43% Hispanic, 41% Caucasian 6% Haitia n Creole 6% African American 3% Mixed, 1% Asian, and .3% Native American economically disadvantaged and qualifies for the free and reduced cost lunch program. Because the purpose of this study was to examine the relationship between school choice and student reading achievement, the students must have taken the 2007 08 Reading FCAT as third graders and the 2008 09 Reading FCAT as fourth graders. All fourth grade school choice students from the 2008 09 school year were part of the analysis as well as 2008 09 fourth graders who were eligible for choice but stayed at their zoned school. Access and Entry After receiving approval from the University of Florida Institutional Review Board and permission from the school district, the researcher contacted the school system Director of Assessment and Data Management to obtain student scores. The scores were sorted by demographics an d subtest results on two Excel s preadsheets. Variables of the Study Data collection included analysis of student demographic data and subtest data Student demographic variables included economically disadvantaged status ESE status gender, ELL status and race. Subtests included the Information subtest,

PAGE 44

44 Literature subtest, Words and Phrases in Context subtest, Main Idea, Plot, and Purpose subtest, Comparisons and Cause/Effect subtest, and Reference and Research subtest. Instrumentation The FCAT was used to quantitatively measure student performance. In this study, student FCAT developmental scale scores in reading from the 200 7 08 and 2008 09 school years for students who elect ed the NCLB school choice option and those who stayed in their zoned schools w ere analyzed and compared. Validity and Reliability of the FCAT Validity According to the Florida Department of Education (2004a), the FCAT is a technically sound and valid instrument that meets or exceeds professional standards for standardized achievement tests. The achievement of skill s and content described in the SSS Content validity, concurrent validity, and construct related validity have been substantiated. Specifically, t he FCAT has high content validity. I tem specifications were developed items were pilot tested with randomly selected students, items were field tested to determine psychometric properties, and items were reviewed for cultural, language and gender bias (Florida Department of Education, 2007). T he FCAT demonstrates criterion related validity with the Stanford 9 t est. However, the two tests do not provide the same information (Florida De partment of Education, 2007). The FCAT has construct related validity. This was established through convergent and discriminant analyses (Florida Department of Education, 2007). The se state generated reports lack the statistical infor mation necessary to quantify findings related to validity. Shermis and Long (2009) calculated the reliability, convergent, and discriminant validity coefficients

PAGE 45

45 to assess the construct validity for the choice and essay/extended response items. The structural equation model (SEM) analysis provided support for both the convergent and discriminant validity of the test scores. In a study of performance on the reading portion of the FCAT, Scha tschneider et al. (2004) provide d substantial evidence for the construct validity of the FCAT test as a measure of reading comprehension S cores on the FCAT consistently identify children who are both better readers (text level fluency) and who have more v erbal knowledge and reasoning skills. FCAT reading comprehension levels were also consistent with reading comprehension scores from the SAT9 test which reports a comparison of performance with a national sample of students. Reliability Reliability for in a lpha and item response theory (IRT) marginal reliabilities for the FCAT SSS test. Reported reliability coefficients confirm that the FCAT is a reliable test for assessment of educational achievement (Florida Department of Education, 2007). According to the FCAT technical repor t marginal reliabilities indicate that FCAT scores have reliabilities similar to those of other standardized and statewide tests. Test scores can fluctuate; therefore, the FCAT should be viewed as only one indication of student achievement (Florida Department of Education, 2006). These reports fail to share statistical information necessary to quantify the findings related to reliability. Assumptions and Limitations Assumptions For this s tudy, several assumptions were made. First, this study looked at a sp ecific population of elementary aged students identified as eligible for the NCLB

PAGE 46

46 school choice program. It was assumed that the students who used the NCLB school choice option were properly identified and resided within the school boundary of the sending school. It was also assumed that the assigned teaching personnel had appropriate teaching credentials for teaching the assigned grade level and stu dents. Another assumption was that, because the school district had adopted curriculum materials for use in general education classrooms, all students were taught using a similar curriculum and had access to similar resources and instructional materials. Because the FCAT was used as the measure of student achievement, it was assumed that the school personnel who administer ed the FCAT received training, used standardized procedures, and allowed students to complete the examination independently. In addition database had been recorded accurately. Limitations The students selected for this sample include all fourth grade NCLB school choice students and those who remained in their zoned school who took the FCAT in both 2008 and 2009. Due to the lack of randomization in the selection of the sample, the ability to generalize the findings is limited. Summary The underlying premise of the NCLB school choice provision was that disadvantaged students would be nefit from attending high performing schools (Nicotera et al., 2007). The purpose of this retrospective data analysis was to determine whether the decision to elect school choice made a difference in achievement, as evidenced by FCAT Reading Gain Scores in one Florida school district. This study analyzed results for student subgroups including economically disadvantaged status ESE status

PAGE 47

47 gender, ELL status and race. Students who used NCLB school choice were compared to eligible students who elected to r emain in their zoned school.

PAGE 48

48 CHAPTER 4 RESULTS AND ANALYSIS OF DATA This chapter presents the results of the research study. The purpose of th e data analysis was to assess the impact of the NCLB school choice option on student academic performance from 2007 to 2009 in one suburban district in Southwest Florida. The researcher analyzed reading test scores for students who used the NCLB school choice option and those who remained in their zoned schools. The developmental scale scores from the thi rd and fo ur th grade FCAT were used to measure student achievement. The results and analysis for the following research questions were addressed : 1. What is the relationship between School Choice and non School Choice, when compared by gender, race, ELL status, ESE status, economically disadvantaged status Information subtest, Literature subtest, Words and Phrases in Context subtest, Main Idea, Plot, and Purpose subtest, Comparisons and Cause/Effect subtest, and Reference and Research subtest of the Reading FCAT? 2 What were the predictors of reading achievement in 2008? 3. What were the predictors of reading achievement in 2009? Descriptive Analysis The sample group in this study included 953 students who were third graders in 20 08 and fourth graders in 20 09. To be included in the sample group, each student had to generate a score on both the 20 08 FCAT as a third grader and the 2009 FCAT as a fourth grader. In addition, the student had to be eligible for NCLB school choice and either select the option to transfer (School Choice) or re main at their zoned school (non School Choice). In this study, 859 students we re identified as non School Choice and

PAGE 49

49 94 students were identified as School Choice. All eligible students in the district were included in the study resul ts. Table 4 1 provides the demographic i nformation for the group. Table 4 1. Overall Student Demographics Demographic Variable n % Gender Female 438 46 Male 515 54 Race African American 66 7 Haitian Creole 136 14 Hispanic 654 69 Other 18 2 Caucasian 79 8 ELL Status LF (follow up ELL) 159 17 LY (active ELL) 235 25 LZ (former ELL) 116 12 ZZ (non ELL) 443 46 ESE Status No 749 79 Yes (non gifted) 204 21 Economically Disadvantaged Status No 146 15 Yes Group Non School Choice School Choice 807 859 94 85 90 10 Data Analysis Research Question 1 What is the relationship between School Choice and non School Choice, when compared by gender, race, ELL status, ESE status, economically disadvantaged status Information subtest, Literature subtest, Words and Phrases in Context subtest, Main Idea, Plot, and Purpose subtest, Comparisons and Cause/Effect subtest, and Reference and Research subtest of the Reading FCAT? To address this question, comparison

PAGE 50

50 group testing was completed for the binary (male/female, ESE (non gifted) /no n ESE, economically disadvan taged /non economically disadvantaged) categorical (race, ELL status ), and numerical (overall reading mean/subscale scores) variables. The corresponding null hypothesis was developed to test the research question 1: H o 1 : There will be no statistically sign ificant relationship s between School Choice and non School Choice on gender, race, ELL status, ESE status, economically disadvantaged status Information subtest, Literature subtest, Words and Phrases in Context subtest, Main Idea, Plot, and Purpose subtes t, Comparisons and Cause/Effect subtest, and Reference and Research s ubtest as measured by the FCAT D evelopmental Scale Score in 2008 and 2009. First, chi square testing was used to compare 2008 non School Choice and School Choice for the binary variables as presented in Table 4 2. The percentages of students in each group were found to be not significantly different for the variables of Gender ( p = .862) and ESE status ( p = .1749), while differences for economically disadvantaged status were significant ( p = 0005). In this case, 86% of non School Choice students were economically disadvantaged compared with 72% of School Choice students. Due to significance levels above .05 the findings show that it was necessary to fail to reject the null hypothesis for the binary variables G ender and ESE status For economically disadvantaged status however, d ue to the significance level falling below .05, the null hypothesis for this binary variable was rejected

PAGE 51

51 Tab le 4 2. Group Comparison s of non School Choice and School Choice Non School Choice n (%) School Choice n (%) p Demographic Variable Total 859 (90%) 94 (10%) N/A Gender Female 394 (46%) 44 (47%) .862 0 Male 465 (54%) 50 (53%) Race African American 59 (7%) 7 (7%) <.0001 Haitian Creole 123 (14%) 13 (14%) Hispanic 608 (71%) 46 (49%) Other 12 (1%) 6 (6%) Caucasian 57 (7%) 22 (23%) ELL Status LF (follow up ELL) 152 (18%) 7 (7%) < .0001 LY (active ELL) 227 (26%) 8 (9%) LZ (former ELL) 103 (12%) 13 (14%) ZZ (non ELL) 377 (44%) 66 (70%) ESE Status No 670 (78%) 79 (84%) .1749 Yes (non gifted) 189 (22%) 15 (16%) Economically Disadvantaged Status No 120 (14%) 26 (28%) .0005 Yes 739 (86%) 68 (72%) Second, chi square testing was used to compare 2008 non School Choice and School Choice for the categorical variables as shown in Table 4 2. The differences were significant for both variables, race ( p < .0001) and ELL status ( p < .0001). The percentage of Hispanic non School Choice students was 71% while the percentage of Hispanic School Choice students was 49%. Additionally t he percentage of classified ZZ (students not labeled ELL) non School Choice students was 44%, while the percentage of ZZ School Choice students was 70%. Due to the significance levels falling below .05, the null hypothesis for the categorical variables of race and ELL status was rejected

PAGE 52

52 Third, t tests were used to compare 2008 non School Choice and School Choice for the numerical variables as presented in Table 4 3. The percent age correct was significantly different for non School Choice and School Choice on all variables (Overall Reading Mean, Comparison and Cause/Effect subtest, Main Idea subtest, Information subtest, Literature subtest, Reference and Research subtest, and Words and Phrases subtest) as evidenced by all p values of < .0001, with the exception of Words and Phrases, having a p value of .0003. This included an Overall Mean Reading Score of 1,249 with a stan dard deviation of 308.8 for non School Choice. School Choice students had an Overall Mean Reading Score of 1,407.8 with a standard deviation of 296. Due to the significance level falling below .05, the null hypothesis for all numerical variables was rejected Table 4 3. 20 08 Comparing Groups on Reading Achievement Non School Choice M ( SD ) School Choice M ( SD ) p Overall Reading/Subtests Overall Reading 1249.1 (308.8) 1407.8 (296) <.0001 Compa rison and Cause/Effect .59 (.20) .67 (.19) <.0001 Main Idea .63 (.19) .73 (.18) <.0001 Information .63 (.20) .73 (.17) <.0001 Literature .60 (.19) .69 (.18) <.0001 Reference and Research .62 (.26) .73 (.23) <.0001 Words and Phrases .61 (.23) .70 (.22) .0003 Research Question 2 What were the predictors of reading achievement in 2008? This question required the use of bi variate testing and regression to identify predictors related to reading achievement. The r esults are shown in Table 4 4. The corresponding null hypothesis was deve loped for research question 2: H o 2 : There will be no statistically significant

PAGE 53

53 relationships between the control variables studen t gender, race/ethnicity, economically disadvantaged status ESE status and ELL status and overall reading mean scores in 2008. Table 4 4. Predictors of Reading Achievement in 2008 M ( SD ) p multi variate value Demographic Variable Gender Female 1267.2 (302.7) 0.8221 .6162 Male 1262.7 (318.2) Race (1) African American (a) 1262.2 (335.4) <.0001 .0013 Haitian Cr eole (a) 1228.3 (285.4) Hispanic (a) 1245.9 (303.3) Other (b) 1556.6 (423.1) Caucasian (b) 1419.1 (296.3) ELL Status (1) LF (b) (follow up ELL) 1264.2 (286.2) <.0001 <.0001 LY (a) (active ELL) 1092.5 (287.5) LZ (b) (former ELL) 1233.9 (275.1) ZZ (c) (non ELL) 1364.4 (299.4) ESE Status No 1273.6 (278.1) .176 .0089 Yes (non gifted) 1232.3 (409.1) Economically Disadvantaged Status No 1361 (287.8) <.0001 .0852 Yes 1247.3 (312.1) (1) Means with the same letter are not significantly different First, predictors that were binary were tested using t tests. The test of hypothesis was used for gender, economically disadvantaged status and ESE status Gender mean scores were not significantly different ( p = .8221; multi varia te p = .6162). Females had a mean score of 1,267 with a standard deviation of 302.7, while males had a mean score of 1,262 with a standard deviation of 318.2. Economically disadvantaged status mean scores were significantly different ( p = .0001; multivaria te p = .08 52). Students classified as non economically disadvantaged had a mean score of 1,361 with a

PAGE 54

54 standard deviation of 287.8 while economically disadvantaged students had a mean score of 1,247.3 with a standard deviation of 312.1. ESE status scores were not significantly different ( p = .176; multivariate p = .00 89). Students classified as non ESE had a mean score of 1,273.6 with a standard deviation of 278.1, while ESE (non gifted) students had a mean score of 1 ,232.3 with a standard deviation of 409.1. Due to a level of significance above .05 the researcher fail s to reject the null hypothesis for the binary variables gender and ESE status Due to the level of significance falling below .05, the null hypothesis for the binary variable economically disadvantaged status was rejected Second, categorical predictors were tested using ANOVA with a follow up multiple comparisons Tukey Kramer procedure. Scores for race were found to be significantly different ( p = <.0 001; multivariate p = 0.0003 ) Specif ically, scores for African American ( M = 1,262, SD = 335.4), Haitian Creole ( M = 1,228, SD = 285.4), and Hispanic ( M =1245.9, SD = 303.3) subgroups were significantly different from scores for Other ( M = 1,556.6, SD = 423.1) and Caucasian ( M = 1,419, SD = 296.3). ELL status scores were also found to be significantly different ( p < .0 001; multivariate p < .0001). Specifically, scores for LY ( M = 1,092, SD = 287.5) differed significantly from scores for LF ( M = 1,264.2, S D = 286.2) and LZ ( M =1,233.9, SD = 275.1). In addition, scores for ZZ ( M = 1,364.4, SD = 299.4) were significantly different from scores for all other ELL classifications. Due to levels of significance falling below .05, the null hypothesis for the categorical variables of race and ELL status was rejected Research Question 3 What were the predictors of reading achievement in 2009? This quest ion required the use of bivariate testing and regression to identify predictors r elated to reading achievement. The r esults are shown in Table 4 5. The corresponding null hypothesis

PAGE 55

55 was developed for research question 3: H o 3 : There will be no statistically significant relationships between the control variables studen t gender, race/ethnicity, economically disadvantaged status ESE status and ELL status, overall reading mean scores in 2008, and overall reading mean scores in 2009. Table 4 5. Predictors of Reading Achievement in 2009 Correlation p multivariate p 2008 Overall Reading .78 <.0001 .4450 Demographic Variable M ( SD ) Gender Female 1520 (239.7) .0042 <.0001 Male 1468.9 (310.1) Race (1) African American (a) 1430.2 (315.1) <.0001 .0226 Haitian Creole (a) 1469.1 (284.6) Hispanic (a) 1485.6 (271.5) Other (b) 1671.3 (267.9) Caucasian (b) 1600 (289.7) ELL Status LF(a) (follow up ELL) 1467.2 (289.6) <.0001 .1173 LY (b) (active ELL) 1379.6 (245.6) LZ (a) (former ELL) 1462.3 (269.5) ZZ (c) (non ELL) 1569.2 (275.8) ESE Status No 1498.3 (246.9) .3286 .5283 Yes (non gifted) 1470.8 (380.7) Economically Disadvantaged Status No 1566.4 (247.1) .0002 .6103 Yes 1479 (284.8) School Choice Status : School Choice 1580.6 (228.9) .0002 .445 Non School Choice 1482.8 (284.5) (1) Means with the same letter are not significantly different First, predictors that were binary were tested using t tests. The test of the hypothesis was used for gender, economically disadvantaged status and ESE status. Gender mean scores were significantly different ( p = .0042; multivariate p = < .0001). Females had a mean score of 1,520 with a standard deviation of 239.7, while males had

PAGE 56

56 a mean score of 1,468 with a standard deviation of 310.1. Economically disadvantaged status mean scores were significantly different ( p = .0002; multivariate p = 0.6103). Students cla ssified as non economically disadvantaged had a mean score of 1,566.4 with a standard deviation of 247.1, while economically disadvantaged students had a mean score of 1,479.3 with a standard deviation of 284.8. ESE status scores were not significantly different for those classified as ESE (non gifted) or non ESE ( p = .3286; multivariate p = .52 83). Students classified as non ESE had a mean score of 1,498.3 with a standard deviation of 246.9, while ESE (non gifted) students had a mean s core of 1,470.8 with a standard deviation of 380.7. Due to levels of significance falling below .05, the null hypothesis for the binary variables gender and economically disadvantaged status was rejected Due to the level of significance above .05 for the binary variable ESE status, the researcher fail s to reject the null hypothesis Second, categorical predictors were tested using ANOVA with a follow up multiple comparisons Tukey Kramer procedure. Scores for race were found to be significantly different ( p = <.0001; multivariat e p = .0226). Specifically, scores for African American ( M = 1,430.2 with a standard deviation of 315.1), Haitian Creole ( M = 1,469.1 SD = 284.6), and Hispanic ( M = 1,485.6, SD = 271.5) subgroups were significantly different from scores for Other ( M = 1,671.3 SD = 267.9) and Caucasian ( M = 1,600 SD = 289.7). ELL status scores also were found to be significantly different ( p < .0001; multivariate p < .1173). Specifically, scores f or LY ( M = 1,379.6 SD = 245.6) differed significantly from LF ( M = 1,467.2 SD = 289.6) and LZ ( M = 1,462.3 SD = 269.5). In addition, scores for ZZ ( M = 1,569.2 SD = 275.8) were significantly different from

PAGE 57

57 scores for all other ELL classifications. Due to levels of significance falling below .05, the null hypothesis for the categorical variables race and ELL status was rejected Third, the 2008 Overall Reading score was computed and found to have a correlation of .78 with the 2009 Overall Reading score ( p < .00 01; multivariate p = <.4450). Because t he multivariate p was greater than .05, the researcher fail s to reject the null hypothesis for the variable overall reading score Table 4 6 s hows the comparison between non School Choice and School Choice in 2009 on the Overall Reading Mean Score and subtests. Because all of the p values were below .05, it is clear that there was a sta tistical difference between non School Choice and School Choice scores. Table 4 6. 2009 Compar ison Groups for Reading Achievement Overall Reading/Subtest Non School Choice M ( SD ) School Choice M ( SD ) p Overall Reading 1482.8 (284.5) 1580.6 (228.9) .0002 Comparison and Cause/Effect .54 (.52) .60 (.56) .0066 Main Idea .56 (.19) .61 (.19) .004 Information .55 (.21) .61 (.20) .0078 Literature .57 (.19) .65 (.17) .0002 Reference and Research .61 (.30) .73 (.25) <.0001 Words and Phrases .67 (.24) .74 (.21) .0016 S ummary The findings indicated that there are significant differences related to school choice decisions based on demographic variables. In regard to race, Hispanic students accounted for 46% of School Choice students and 71% of non School Choice students. In contrast, Caucasian students accounted for 23% of School Cho ice students and only 7% of non School Choice students. For ELL status, Z Z (non ELL status) students accounted for 70% of School Choice s tudents and 44% of non School Choice students.

PAGE 58

58 In contrast, LY students accounted for only 9% of School Choice students and 26% of non S chool Choice students. S tudents identified as economically disadvantaged accounted for 72% of School Choice stud ents and 86% of non School Choice students. In contrast, students identified as non economically disadvantaged accounted for 28% of School Choi ce students and only 14% of non School Choice students. Additionally, because there was a high correlation (.78) and mult ivariate p value of .4450 between overall Reading scores from 2008 to 2009 it was determined that the school choice decision for the entire group did not significantly improve s tudent reading performance.

PAGE 59

59 CHAPTER 5 CONCLUSIONS An important component of NCLB i s that all students have an opportunity to attend a higher performing school if their current school i s designated as needing improvement (Simpson et al., 2004). Attending a higher performing school may be preferred, but the re has been limited researched focused on two important questions: First, how will school choice affect school quality ? (Hastings et al. 2005). Second, how will school choice affect student achievement? (Hoxby, 2002). This study examined the effects of NCL B school choice on reading achievement for elementary aged students in one Southwest Florida school district. Kirkland (2009) stated that the relationship between NCLB school choice and student achievement should be investigated. Similarly, other researche rs recommended research that focused on the relationship between student achievement and school choice ( Hall, 2007, 2010; McCombs, 2007; Phillips et al. 2009). For this study, the hypothesis was that the School Choice group would outperform the non School Choice group on overall reading achievement gains as measured by the 2008 and 2009 Reading FCAT developmental scale scores. In addition, it was hypothesized that demographic variables would affe ct participation in school choice and reading achievement sc ores and gains. Summary of Results This study was guided by three research questions and their corresponding null hypothesis. In the discussion below, e ach null hypothesis is included with the corresponding research question and results of the data analysis.

PAGE 60

60 Research Question 1. What is the relationship between School Choice and non School Choice, when compared by gender, race, ELL status, ESE status, economically disadvantaged status Information subtest, Literature subtest, Words and Phrases in Context subtest, Main Idea, Plot, and Purpose subtest, Comparisons and Cause/Effect subtest, and Reference and Research subtest of the Reading FCAT? H o 1 : There will be no statistically significant relationships between School Choice and non School Choice on gender, race, ELL status, ESE status, economically disadvantaged status Information subtest, Literature subtest, Words and Phrases in Context subtest, Main Idea, Plot, and Purpose subtest, Comparisons and Cause/Effect subtest, and Reference and Researc h subtest as measured by the FCAT developmental Scale Score in 2008 and 2009. When considering demographic groups, the analysis revealed that the percentages of students that used the s chool c hoice option were significantly different from the percentages f or non School Choice for economically disadvantaged status race, and ELL status. Each of these variables had p values that fell below .05, causing the null hypothesis to be rejected. In contrast, the demographic variables of gender and ESE status had p va lues greater than .05 making it necessary to fail to reject the null hypothesis for gender and ESE status. When considering 2008 subtests, the analysis revealed that there were significant difference s between School Choice and non School Choice students on all subtests. School Choice students scored higher on all measures (Information subtest, Literature subtest, Words and Phrases in Context subtest, Main Idea, Plot, and Purpose subtest, Comparisons and Cause/Effect subtest, and Reference and Research sub test) and this

PAGE 61

61 was found to be significant as evidenced by all p values that fell below .05. Thus, the null hypothesis was rejected for all subtest measures. Research Question 2. What were the predictors of reading achievement in 2008? H o 2 : There will b e no statistically significant relationships between the control variables student gender, race/ethnicity, economically disadvantaged status ESE status and ELL status and overall reading mean scores in 2008. Bivariate testing and regression w ere used to identify predictors related to reading achievement. When considering 2008 demographic groups, the analysis revealed that the mean scores of students who used the s chool c hoice option w ere significantly different from those of non School Choice for economically disadvantaged status race, and ELL status. Each of these variables had p values falling below .05, causing the null hypothesis to be rejected. In contrast, the demographic variab les of gender and ESE status had p values greater than .05 mak ing it necessary to fail to reject the null hypothesis for gender and ESE status when considering reading mean scores in 2008 Research Question 3. What were the predictors of reading achievement in 2009? H o 3 : There will be no statistically significant re lationships between the control variables studen t gender, race/ethnicity, economically disadvantaged status ESE status and ELL status and overall reading mean scores in 2009. Bivariate tes ting and regression were used to identify predictors related to r eading achievement. When considering 2009 demographic groups, the analysis revealed that the mean scores of students us ing the s chool c hoice option w ere significantly different from those for non School Choice for gender, economically disadvantaged status race, and ELL status. Each of these variables had p values falling below .05, causing the null

PAGE 62

62 hypothesis to be rejected. In contrast, the demographic variable of ESE status had p values greater than .05 making it necessary to fail to reject the null hy pothesis for ESE status when considering reading mean scores in 2009 Comparison group testing indicated that there was no significant difference in overall reading gain scores between the two groups of students, School Choice and non School Choice, as mea sured by their 2009 reading developmental scale score in relation to their 2008 reading developmental scale score. Because no significant difference was found ( multivariate p value < .4450; correlation = .78), the null hypothesis was rejected for the Overa ll Reading. While there was a wide gap between the two groups in overall reading mean scores, School Choice students displayed no significant difference in reading gain scores as compared to non School Choice students. Limitations of the Study The student sample in this study consisted of 953 students who received scores on the FCAT Reading test as third graders in 2008 and fourth graders in 2009, in one suburban Southwest Florida school district. Of the students, 94 used the NCLB school choice option an d 859 students elected to remain at their zoned school. This sample represents all eligible students with test scores during the period 2007 to 2008 and 2008 to 20 09. The number of students in the School Choice group, 94, is relatively small when compared to the total eligible sample, 953. There might not have been enough school choice sample participants to detect statistically significant differences, thus a Type II error could have been present. developmental scale scores in reading from the 2008 FCAT test and the 2009 FCAT

PAGE 63

63 test. This measure is limiting because the FCAT test measured performance only during those two testing pe riods. Using additional measures of student p erformance, such as grades, end of course examinations, and/or norm referenced tests, may have provided additional insight. In addition, using only specified grade levels and years decreased the ability to draw conclusion s beyond the selected sample. This study could have been enhanced by increasing the number of grade levels and using information from other districts to increase sample sizes. School differences and issues related to curriculum delivery and suppo rt services were beyond the scope of this investigation. Additionally, the reasons that students selected School Choice or non School Choice were not known. Implications Although research related to the NCLB school choice option and achievement has shown mixed results ( Hall, 2010; Kirkland, 2009; Nicotera et al., 2007; U.S. Department of Education, 2007; Zimmer et al., 2007 ) it is clear that federal po licymakers continue to support s chool c hoice legislation. Prior to NCLB, school choice and accountability for test results were not mea sured or considered (Betebenner et al., 2005). Lawmakers relied on the premise that providing parents with the option to tr ansfer their children to higher performing schools was one way to force failing schools to meet standar ds (Betebenner et al., 2005). The findings in this study add to the mixed results presented in the literature related to NCLB Choice. The overall group performance in reading did not improve due to school choice. This study found that s ubgroups, such as C aucasian and non economically disadvantaged used school choice in direct contrast to the federal goal of providing choice for disadvantaged students. Perhaps policymakers should pay

PAGE 64

64 particular attention to the percentages of students in each subgroup who s elect school choice and fund additional studies to determine growth on achievement measures. Recommendations This study investigated the academic achievement of students who participated in school choice by transferring from their zoned s chool and enrolli ng in a higher performing school, as determined by NCLB guidelines. Additional research is needed to establish the relationship between the use of school choice and improved student academic achievement. This study focused on one suburban district in Sout hwest Florida and the academic achievement of students who chose to transfer under NCLB guidelines. By looking at only the years 2007 20 09 and the cohort that participat ed in testing both years, the number of students available who s elected school choice w as limited to 94. Replicating this study in other district s and in other grade levels would significantly increase the student sample. A larger sample may uncover further differences in academic achievement that were not detected in this study. Increasing the number of participants also could introduce a greater variety of demographic factors, making it possible to better examine the interaction between student subgroups and academic achievement and the findings more generalizable Future research also could use qualitative methods to investigate instructional effectiveness at choice and non choice schools. I nterviews with stakeholders (parents, teachers, students, administrators, and reading coaches) at high and low gain schools could be conducted to d etermine the factors that co ntribut e to the differences between these schools Also, instructional observations at choice and non choice schools could be conducted to describe the quality and nature of reading instruction and level of

PAGE 65

65 support provided to t argeted students. Exploring parent education level as an additional variable could help explain differences in selection of school choice and academic gains for certain groups of students. Additionally, research could explore the reasons for significant di fferences between the performance in 2009, between males and females, who had nearly identical mean scores in 2008. Finally, future research could examine why parents elect school choice for their children and the motivation s that drive their decision s Conclusions The purpose of this study was to determine whether NCLB school choice affects student reading achievement when comparing students who used school choice to those who declined the opportunity. FCAT Developmental Scale Scores were analyzed to com pute reading mean scores for students during the 2007 2008 and 2008 2009 school years. The results of this study indicate that there is no evidence to support the claim that school choice significantly affe cts academic achievement in reading. Students in b oth the School Choice and non School Choice groups made statistically comparable gains on the Reading FCAT from 2008 09 W hile students in both groups made comparable gains it is concerning that students who elected School Choice scored at a statistically significant higher rate than non School Choice students during both testing periods and on all subtests Students classified as Caucasian, non economically disadvantaged and non ELL scored at a statistically significant higher rate than economically disadvantaged, African American, Hispanic, Haitian Creole, and active ELL in both years. While all subgroups are making comparable gains it is clear that the achievement gap is not closing. T his policy does not appear to be helping improve t he academic performance of either group (school choice and non school choice) T he

PAGE 66

66 results question whether the enormous financial commitment made by policy makers is worthwhile. Not surprisingly, the study results indicated that subgroups select s chool choice at varied rates. For example, for the subgroup of race, Caucasian students were more likely to select school choice than were Hispanic students. Students classified as ZZ (non ELL) were more likely to s elect school choice than those identified as LF, LY, or LZ. Students identified as no n economically disadvantaged were more likely to s elect school choice than those identified as economically disadvantaged It is important to remember that the intent of federal NCLB legislation was that all stu dents have an opportunity to a ttend higher performing schools if their current school was designated as needing improvement (Simpson et al., 2004). Specifically, the legislation gave disadvantaged students the opportunity to attend high quality public scho ols (Hastings et al., 2007). Unfortunately, this study shows the opposite to be true. Students identified as economically disadvantaged with ELL classifications, and certain racial groups ( African American and Hispanic) transferred at significantly lower rates than did comparison groups. Arce et al. (2005) reported that t he legislation often caused low income students to attend schools with higher percentages of disadvantaged students than prior to the enactment of NCLB. It appears that the subgroups who t ook advantage of this legislation included those who were the least disadvantaged. Clearly, state and federal policymakers should fund studies to examine both the percentages of students elect ing school choice in each AYP subgroup and the academic performance of these students. Simply implementing a federal policy, such as NCLB School Choice, without evaluating effectiveness does little to help student performance or even support

PAGE 67

67 the continuation of the policy. Determining the impac t on each subgroup should be a mandatory part of federal and state program evaluation.

PAGE 68

68 APPENDIX A U NIVERSITY OF FLORIDA INSTITUTIONAL REVIEW BOARD (UF IRB ) APPROVAL

PAGE 69

69 APPENDIX B SCHOOL DISTRICT APPR OVAL

PAGE 70

70 LIST OF REFERENCES Arce, J., Luna, D., Borjian, A., & Conrad, M. (2005). No Child Left Behind: Who wins? Wh o loses? Social Justice, 32 (3), 56 71. Retrieved September 20, 2009, from http://vnweb.hww ilsonweb.com.lp.hscl.ufl.edu Bast, J. & Walberg, H. (2004). Can parents choose the best schools for their children? Economics of Education Review, 23 (2), 431 440. Retrieved September 24, 2009, f rom http://www.sciencedirect.com.lp.hscl.ufl.edu/science Betebenner, D., Howe, K., & Foster, S. (2005). On school choice and test based accountability. Education Policy Analysis Archives, 13 (4 1), 1 19. Retrieved September 20, 2009, from http://epaa.asu.edu/epaa/v13n41/ Cullen, J., Jacob, B., & Levitt, S. (2005). The impact of school choice on student outcomes: A n analysis of the Chicago Public Schools. Journal of Public Economics, 89 (5 6), 729 760. Retrieved September 19, 2009, from http://faculty.smu.edu/millimet/classes/ec o7321/papers/cullen%20et%20al.pdf Creswell, J. (2008). Educational research: Planning, conducting, and evaluating quantitative and qualitative research (3rd ed.). Upper Saddle River, NJ: Pearson Education. Florida Department of Education. (1994). DOE in formation data base requirements Vol 1: Automated student information systems, automated student data elements: English l anguage l earners, PK 12. Retrieved June 22, 2011, from http://www.fldoe.org/eias/dataweb/database_1011/st112_1.pdf Florida Department of Education. (1996). Sunshine State Standards. Retrieved January 10, 2010, from http://www.fldoe.org/bii/cur riculum/SSS/sss1996.asp Florida Department of Education. (2002). Overview of FCAT Program Retrieved September, 17, 2011, from http://fcat.fldoe.org/pdf/fcss08b1.pdf Florida Department of Education. (2004a). Assessment and accountability briefing book. The Fl orida Comprehensive Assessment Test Retrieved October 23, 2009, from http://www.firn.edu/doe/sas/fcat/pdf/fcataabb.pdf Florida Department of Education. (2004b). FCAT: Florida Comprehe nsive A ssessment Test. Retrieved October 23, 2009, from http://www.fldoe.org Florida Department of Education. (2005 a ). Glossary. Retrieved October 12, 2009, from www.fldoe.org/comptroller/doc/gbglossary.doc Florida Department of Education. (2005 b ). Bureau of K 12 assessment. Retrieved August 10, 2010, from http://f cat.fldoe.org/

PAGE 71

71 Florida Department of Education. (2006). Reading and mathematics technical report for 2006 F CAT t est a dministrations. Retrieved November, 17, 2010, from http://fcat.fldoe.org/pdf/fc06tec h.pdf Florida Department of Education. (2007). Assessment and accountability briefing book: FCAT/s chool accountability/teacher certification t ests. Retrieved November 10, 2010, from http://fcat.fldoe.org/pdf/BriefingBook07web.pdf Florida Department of Education. (2010). Florida Comprehensive Assessment Test. Retrieved September 27, 2009, from http://fcat.fldoe.org/ Fusarelli, L. (2004). The poten tial impact of the No Child Left Behind Act on e quity and d iversity in American e ducation. Educational Policy, 18 (1), 71 94. Retrieved August 28, 2009, from http://epx.sagepub.com Greene, J. (2001). An evaluation of th e Florida A plus accountability and school choice program. New York, NY: Center for Civic Innova tion at the Manhattan Institute Retrieved August 11, 2010, from http://www.hks.harvard.edu Hall, D. (2010). Student achievement and public school choice mandate of No Child Left Behind Act (Ph.D. dissertation, University of Florida, Gainesville ). Retrieved November 11, 2010, from Dissertations & T heses @ University of Florida FCLA (Publication No. AAT 3416676) Hanushek, E., Kain, J., Markman, J., & Rivkin, S. (200 3 ). Does peer ability affect student achievement? Journal of Applied Econometrics, 18 527 544. Retrieved September 18, 2011 from http://edpro.stanford.edu/hanushek/admin/pages/files/uploads/Hanushek+Kain+M arkman+Rivkin%202003%20JAppEct%20185.pdf Hastings, J., Kane, T., & Staiger, D. (2005). Parental preferences and school competition: Evidence from a public school choice program ( National Bureau of Economic Research [ NBER ] Working Paper No. 11805 ) Retrieved November 10, 2010, from www.yal e.edu/~jh529/Hastings_Kane_Staiger_QualityCompetition.pdf Hastings, J., Van Weelden, R. & Weinstein, J. (2007). Preferences, information and parental choice behavior in public school choice ( National Bureau of Economic Research [ NBER ] Working Paper 12995 ) Retrieved November 10, 2010, from http://faculty.maxwell.syr.edu/jmwein04/Weinstein_InformationChoices.pdf Hastings, J., & Weinstein, J. (2008). Information, scho ol choice, and academic achievement: Evidence from two experiments. The Quarterly Journal of Economics, 123 (4), 1373 1414. Retrieved September 29, 2009, from http://faculty.maxwell.syr.educ

PAGE 72

72 Heise, M. (1999). E qual educational opportunity and constitutional theory: Preliminary thoughts on the role of school choice and the autonomy principle. Journal of Law & Politics, 24 (14), 411 458. Retrieved November 12, 2010, from http://library2.lawschool.cornell.edu/hein knowledge about the choice provisions of no child left behind. Peabody Journal of Education, 81 (1), 140 179. Re trieved September 19, 2009, from http://vnweb.hwwilsonweb.com.lp.hscl.ufl.edu/hww/results Hoxby, C. (1998). What do America's "traditional" forms of school choice teach us about schoo l choice reforms? Federal Reserve B ank of New York Economic Policy Review, 4 (1) Retrieved November, 10, 2010, from http://www.ny.frb.org/research/epr/98v04n1/9803hoxb.pdf Hoxby C. (2002) How school choice affects the achievement of public school students. In P. Hill (E d.), Choice with equity Stanford, CA: Hoover Institution Press. Retrieved November 1 0, 2010, from http://www .allianceforschoolchoice.org in Educational Researcher, 38 (6), 449 457. Retrieved June 10, 2011, from http://er.aera.net Kirkland, T. (2009). The influence of the school choice provision, within the No Child Left Behind legislation, on the academic achievement of students and on the demographic composition of Title I schools in Collier County, Florida ( Ed.D. dissertation, University of Central Florida, Orlando ) Retrieved November 11, 2010, from Dissertations & Theses: Full Text. (Publication No. AAT 3401082) Linn, R. (2005). Conflicting demands of No Child Left Behind and state systems: Mixed messages about school performance. Education Policy Analysis Archives, 13 (33), 1 20. R etrieved August 28, 2009, from http://vnweb.hwwilsonweb.com.lp.hscl.ufl.edu McCombs, J. (2007). The school choice provisions of No Child Left Behind: An examination of access, information, and impact ( Ph.D. dissertation, The George Washington University, Washington, DC) Retrieve d November 15, 2010, from Dissertations & T heses: Full Text. (Publication No. AAT 3275146) McDonnell, L. (2005). No Child Left Behind and the federal role in education: Evolution or revolution? Peabody Journal of Education, 80 (2), 19 38. Retrieved August 28, 2009, from http://vnweb.hwwilsonweb.com.lp.hscl.ufl.edu Neild, R. (2005). Parent management of school choice in a large urban district. Urban Education, 40 (3), 270 297. Retrieved September 20, 2009, from http://uex.sagepub.com/cgi/content/abstract/40/3/270

PAGE 73

73 Nicotera, A., Teasley, B., & Berends, M. (2007). An empirical investigation of the No Child Left Behind school choice policy on academic achievement. Washington, DC: U.S. Department of Education, National Center on School Choice. Retrieved November 10, 2010, from www.vanderbilt.edu/schoolchoice/documents/empirical.pdf Okpala, C., Bell, G., & Tuprah, K. (2007). A comparative study of traditional schools and schools of choice in North Carolina. Urban Education, 42 (4), 313 325. Retrieved November 1 5, 2010, from http://uex.sagepub.com.lp.hscl.ufl.edu/content/42/4/313.full.pdf+html Peca, K. (2000). Positivism in education: Philosophical, research, and organizational assumptions. Opini on Papers 1 33. Retrieved June 10, 2011, from http://ericfac.plccard.csc.com Phillips, K., Hausman, C., & Larsen, E. (2009). Intra district transfer and student achievement: A case study of the effects of choice on achievement. Nashville, TN: Vanderbilt University, National Center on School Choice. Retrieved November 10, 2010, from www.vanderbilt.edu/schoolchoice/ Sale, J., Lohfeld, L., & Brazil, K. (2002). Revisiting the quantitative qualitative debate: Implications for mixed methods research. Quality and Quantity, 36 43 53. Retrieved October 1, 2011, f rom http://www.cnr.uidaho.edu/CSS506/506%20Readings/sale%20mixed methods.pdf Schatschneider, C., Buck, J., Torgeson, J. Wagner, R. Hassler, L., Hecht, S. & Powell Smith, K. (2004). A Multivariate Study of Individual Differences in Performance on the Reading Portion of the Florida Comprehensive Assessment Test: A Brief Report Tallahassee, FL: Florida Center for Reading Research. Retrieved October 2, 2011, f rom http://www.fcrr.org/technicalreports/multi_variate_study_december2004.pdf Shermis, M. & Long, S. (2009). Multitrait multimethod analysis of FCAT reading and writing: Or is it writing and reading? Journal of Psychoeducational Assessment, 27 296 310. Retrieved October 2, 2011, from http://jpa.sagepub.com.lp.hscl.ufl.edu/content/27/4/296.full.pdf+html Simpson, R., LaCava, P., & Graner, P. (2004). The No Child Left Behind Act : Challenges and implications for educators. Intervention in School and Clinic, 40 (2), 67 75. Retrieved September 20, 2009, from http://isc.sagepub.com.lp.hscl.ufl.edu/ Smith, E. (2005). Raising standards in American schools: T he case of No Child Left Behind J ournal of Education Policy, 20 (4), 507 524. Retrieved September 20, 2009, from http://pdfserve.informaworld.com.lp.hscl.ufl.edu

PAGE 74

74 Sunderman, G. & Kim, J. (2004). Inspiring vision, disappointing results: Four studi es on implementing the No Child Left Behind Act. Ca mbridge, MA : Harvard University, The Civil Rights Project. Retrieved September 20, 2009, from http://www.eric.ed.gov/ERICDo cs/data/ericdocs2sql/content_storage_0 Thomas, J & Brady, K. (2005). Chapter 3: The Elementary and Secondary Education Act at 40: E quity, accountability, and the evolving federal role in public education Review of Research in Education, 29 (8), 51 67. Retrieved September 26, 2009, from http://rre.sagepub.com.lp.hscl.ufl.edu/cgi/reprint/29/1/51 U.S. Department of Educat i on. (2002). Guidanc e for the Reading First program. Retrieved June 17, 2010, from http://www.edu.gov/readingfirst U.S. Department of Education. (2003). Choice provisions in No Child Left Behind: Nation al Title 1 Directors Conference. Retrieved September 26, 20 09, from http://www.ed.gov/admins/comm/choice/choice03/edlite slide24.html U.S. Department of Education. (2004). A guide to education and No Child Left Behind Retrieved September 19, 2009, from http://www.edu.gov/nclb/overview .pdf U.S. Department of Education. (2008). No Child Left Behind: Helping families by supporting and expanding school choice (Archived information). Retrieved September 24, 2009, from http:www.ed.gov/nclb/choice/schools/choicefacts.pdf U.S. Department of Education (2009 a) State and local implementation of the No Child Left Behind Act: Volume VII Title 1 school choice and supplemental educational services: Final r eport. Retrieved Sept ember 1, 2010, from http://www2.ed.gov/rschstat/eval/choice/nclb choice ses final/choice ses final.pdf U.S. Department of Education. (2009 b ). Elementary & Se condary E ducation: Title I i mproving the academic ac hievement of the disadvantaged. Retrieved September 20, 2009, from http://www.ed.gov/policy/elsec/leg/esea02/pg1.html U.S. Department of Education. (2009 c ). Public school c hoice: Non regulatory guidance. Retrieved September 19, 2009, from www.edu.gov/policy/elsec/guid/schoolchoiceguid.pdf U.S. Department of Education. (2009 d ). Program guidance/policy letters: Office of Innovation and Improvement. Retrieved September 27, 2009, from http://www.ed.gov/about/offices/list/oii/resources/info/g uidance.html Wisconsin Department of Public Instruction. (2009). Economically disadvantaged status data collection and reporting. Retrieved February 8, 2010, from http://dpi.wi.gov/lbstat/dataecon.html

PAGE 75

75 Z immer, R., Gill, B., Razquin, P., Booker, K., & Lockwood, J. (2007). State and local implementation of the No Child Left Behind Act: Volume I Title I school choice, supplemental educational services, and student ac hievement. Washington, DC: RAND. Retrieved November 15, 2010, from http://eric.ed.gov/PDFS/ED497259.pdf

PAGE 76

76 BIOGRAPHICAL SKETCH Brian D ouglas Castellani attended Sachem High School in Ronkonkoma, New York. He graduated from the Uni versity of South Florida in 1991 receiving a Bachelor of Science in elementary e ducation. In the fall of 1991, Brian started teaching at Chapel Hill Middle School in Douglasville, Georgia, where he worked for three years. During this period he earned his Master of Education in A dministration and S upervision from the University of West Georgia In the fall of 1994, Brian transferred to Chapel Hill Elementary School where he worked f or the next three years. During this period he earned his Educational Specialist in Educational Administration from Georgia State University and started his first administrative position as i nstructional l ead t eacher Since then he has served as i nstructional l ead t eacher at Bright Star Elementary Sch ool, a ssistant p rincipal at Chestnut Log Middle School and Corkscrew Elementary School, and principal at Sea Gate Elementary School. He is currently the principal at Lake Trafford Elementary School in Immokalee, Florida. Brian is the oldest child of Roger Robert Castellani and Kathleen Francis Castellani. He is married to Jean Marie Castellani and is the father of two, Kathleen Patricia and Colleen Teresa. Brian enjoys watching his olde r daughter cheer at high school football games, camping with his childr en and friends, a nd caring for his pet turtle and fruit trees.