Citation
Relationship between Selected School and Student Characteristics and Principal Turnover in One Florida School District

Material Information

Title:
Relationship between Selected School and Student Characteristics and Principal Turnover in One Florida School District
Creator:
Burd, Sarah E
Place of Publication:
[Gainesville, Fla.]
Florida
Publisher:
University of Florida
Publication Date:
Language:
english
Physical Description:
1 online resource (107 p.)

Thesis/Dissertation Information

Degree:
Doctorate ( Ed.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Educational Leadership
Human Development and Organizational Studies in Education
Committee Chair:
OLIVER,BERNARD
Committee Co-Chair:
ELDRIDGE,LINDA BURNEY
Committee Members:
MOUSA,BRUCE E
OLIVER,EILEEN
Graduation Date:
8/8/2015

Subjects

Subjects / Keywords:
Academic achievement ( jstor )
Educational administration ( jstor )
High schools ( jstor )
Low socioeconomic status ( jstor )
Mathematics ( jstor )
School principals ( jstor )
Schools ( jstor )
Special needs students ( jstor )
Statistical significance ( jstor )
Students ( jstor )
Human Development and Organizational Studies in Education -- Dissertations, Academic -- UF
principal -- turnover
Genre:
bibliography ( marcgt )
theses ( marcgt )
government publication (state, provincial, terriorial, dependent) ( marcgt )
born-digital ( sobekcm )
Electronic Thesis or Dissertation
Educational Leadership thesis, Ed.D.

Notes

Abstract:
This study examined relationships between principal turnover rate, and selected school characteristics (percentage of minority students, percentage of students with disabilities, percentage of students who are economically disadvantaged, and student achievement in reading/English language arts and math measured by Adequate Yearly Progress (AYP) on the Florida Comprehensive Assessment Test (FCAT)). This study sample comprised public schools (elementary, middle and high). All of these schools were located in one selected Florida school district. Data were collected AYP reports publicly accessed on the Florida Department of Educations website, and the school districts Data Warehouse website. The director of communications at the countys public school district office collected data on the frequency of principal turnover. Data were statistically analyzed using a Quasi-Poisson regression model. This study found that an increase in reading and math scores indicates a lower rate of principal turnover. It also indicates that the higher the minority rates in a school and higher percent of low socio-economic students, the greater the principal turnover rate. ( en )
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.
Thesis:
Thesis (Ed.D.)--University of Florida, 2015.
Local:
Adviser: OLIVER,BERNARD.
Local:
Co-adviser: ELDRIDGE,LINDA BURNEY.
Electronic Access:
RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2016-02-29
Statement of Responsibility:
by Sarah E Burd.

Record Information

Source Institution:
UFRGP
Rights Management:
Applicable rights reserved.
Embargo Date:
2/29/2016
Classification:
LD1780 2015 ( lcc )

Downloads

This item has the following downloads:


Full Text

PAGE 1

RELATIONSHIP BETWEEN SELECTED SCHOOL AND STUDENT CHARACTERISTICS AND PRINCIPAL TURNOVER IN ONE FLORIDA SCHOOL DISTRICT By SARAH E. BURD 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 2015

PAGE 2

© 2015 Sarah E. Burd

PAGE 3

To my Dad, who always believed in me

PAGE 4

4 ACKNOWLEDGMENTS When I began this journey, I never imagined how many people would be involved in making this dream a reality. I thank my family members, friends and the University of Florida faculty members who helped and encouraged me along the way. I could not have completed this without them . Specifically, I thank Dr. Bernard Oliver, my supervisory committee chair, for sharing his insight and wi sdom on my topic. I also thank Dr. Bruce Mousa , one of my University of Florida professors who walked this journey with me over the many year s it has ta ken to reach my goal. His encoura gement and belief in me was steadfast and unwavering even after I left the program for a while after having my triplets. I thank Sungur Gurel for helping with my statistical analysis; Anne Taylor for editing my pa per; and Danielle Chaprnka, Michelle Licata, John Om undsen, Rebecca Mehrer , and Christy Widener . S pecial thanks go to my sweet friend, Sarah Van Gemert, whose insigh ts and friendship are priceless. I thank my mom and dad, who taught me the value of hard work and listening to the direction of the Lord. Finally, I thank my husband, Evan, for pushing me when I needed pushing, encouraging me when I needed encouragement , and loving me always. He was always quick to help with our boys so I coul d reach this goal. I love my boys and pray this will help them know that with God, nothing shall be impossible. I thank my Lord, Jesus Chr ist, who makes the impossible, possible, and who makes the incapable, more than able!

PAGE 5

5 TABLE OF CONTENTS P age ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 8 LIST OF FIGURES ................................ ................................ ................................ .......... 9 LIST OF ABBREVIATIONS ................................ ................................ ........................... 10 ABSTRACT ................................ ................................ ................................ ................... 11 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 12 Background ................................ ................................ ................................ ............. 12 Problem Statement ................................ ................................ ................................ . 14 Purpose of the Study ................................ ................................ .............................. 16 Significance of the Study ................................ ................................ ........................ 16 Research Question ................................ ................................ ........................... 17 Research Hypotheses ................................ ................................ ...................... 18 Null Hypotheses ................................ ................................ ............................... 18 Identification of Variables ................................ ................................ ................. 19 Research Plan ................................ ................................ ................................ .. 20 Summary ................................ ................................ ................................ ................ 20 2 REVIEW OF THE LITERATURE ................................ ................................ ............ 22 Introduction ................................ ................................ ................................ ............. 22 Leadership Turnover and Succession in the Private Sector ................................ ... 23 Leadership in Schools ................................ ................................ ............................. 28 Principal Turnover ................................ ................................ ................................ ... 33 Low Performing Schools ................................ ................................ .................. 34 Turnaround Schools ................................ ................................ ......................... 35 Principal Turnover and Low Socioeconomic Schools ................................ ....... 42 Students with Disabilities (SWD) ................................ ................................ ...... 44 Migrant Students ................................ ................................ .............................. 46 Certification and Training ................................ ................................ .................. 46 Teacher Turnover ................................ ................................ ............................. 47 Effective Principals ................................ ................................ ........................... 48 Principal Assessment Data ................................ ................................ ............... 51 Summary ................................ ................................ ................................ ................ 52

PAGE 6

6 3 METHODS ................................ ................................ ................................ .............. 54 Participants ................................ ................................ ................................ ............. 54 Setting ................................ ................................ ................................ ..................... 54 Instrumentation ................................ ................................ ................................ ....... 54 Florida Comprehensive Assessment Test ................................ ........................ 55 Adequate Yearly Progress ................................ ................................ ................ 56 Procedures ................................ ................................ ................................ ............. 56 Data Sources ................................ ................................ ................................ .......... 58 Access to the Data ................................ ................................ ........................... 59 Demographic Profiles ................................ ................................ ....................... 59 Reading and Math Achievement ................................ ................................ ...... 60 Data Analysis ................................ ................................ ................................ .......... 60 Poisson and Quasi Poisson Regression ................................ .......................... 61 Poisson and Quasi Poisson Regression Assumptions ................................ ..... 63 Summary ................................ ................................ ................................ ................ 64 4 FINDINGS ................................ ................................ ................................ ............... 67 Assumptions ................................ ................................ ................................ ........... 67 Hypothesis Testing Results ................................ ................................ .................... 69 Reading Model ................................ ................................ ................................ . 69 Math Model ................................ ................................ ................................ ....... 69 Minority Model ................................ ................................ ................................ .. 70 Low SES Model ................................ ................................ ................................ 70 5 DISCUSSION ................................ ................................ ................................ ......... 72 Summary of Findings ................................ ................................ .............................. 72 Review of the Research Questions ................................ ................................ .. 72 Review of the Hypothesis and Null Hypothesis ................................ ................ 72 Null Hypotheses ................................ ................................ ............................... 73 Discussion of Finding in Light of Relevant Literature ................................ .............. 73 Reading FCAT and Principal Turnover Findings ................................ .............. 74 Math FCAT and Principal Turnover Findings ................................ .................... 74 Minority Rate and Pri ncipal Turnover Findings ................................ ................. 75 Low SES and Principal Turnover Findings ................................ ....................... 75 Study Limitations and Recommendations for Further Researc h ............................. 77 Implications ................................ ................................ ................................ ............. 77 Limitations ................................ ................................ ................................ ............... 79 Recommendations ................................ ................................ ................................ .. 81 Conclusion ................................ ................................ ................................ .............. 83 APPENDIX A COUNTY PRINCIPAL TURNOVER, TEN YEAR SPAN ................................ ......... 85 B FLORIDA COUNTY MAP ................................ ................................ ....................... 86

PAGE 7

7 C SCHOOL CHARACTERISTICS: STUDENT ACHIVEMENT ................................ ... 87 D DATA FILE ................................ ................................ ................................ .............. 89 E RESEARCH PERMISSIONS ................................ ................................ .................. 91 LIST OF REFERENCES ................................ ................................ ............................... 92 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 107

PAGE 8

8 LIST OF TABLES Table P age 3 1 GVIF: Predictors Flagged for Multicollinearity ................................ ..................... 66 4 1 The CORR Procedure ................................ ................................ ........................ 68 4 2 Reading Model ................................ ................................ ................................ ... 69 4 3 Math Model ................................ ................................ ................................ ......... 70 4 4 Minority Rate Model ................................ ................................ ............................ 70 4 5 Low SES Model ................................ ................................ ................................ .. 71

PAGE 9

9 LIST OF FIGURES Figure P age A 1 Percent of schools' principal turnover rates: TEN YEAR PERIO D ..................... 85 C 1 Elementary school FCAT Reading, Math Scores and number of principals. ...... 87 C 2 Middle school Reading, Math Scores and num ber of principals. ........................ 87 C 3 High schools Reading, Math Scores and number of principals. ......................... 88

PAGE 10

10 LIST OF ABBREVIATIONS AYP Adequate Yearly Progress CCPS Collier Co unty Public Schools CCSS Common Core State Standards CLEM Collier Leadership Evaluative Model ELA English Language Arts ESEA Elementary and Secondary Education Act FAC Florida Administrative Code FCAT Florida Comprehensive Assessment Test FLDOE Florida De partment of Education LEA Local Education Agency NCLB No Child Left Behind NGSSS New Generation Sunshine State Standards RttT Race to the Top SB Senate Bill SEA State Education Agency SES Socioeconomic S tatus SINI Schools in Need of Improvement SSA Studen t Success Act SSS Sunshine State Standards SWD Students with D isabilities

PAGE 11

11 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 RELATIONSHIP BETWEEN SELECTED SCHOOL AND STUDENT CHARACTERISTICS AND PRINCIPAL TURNOVER IN ONE FLORIDA SCHOOL DISTRICT By Sarah E. Burd August 201 5 Chair: Bernard Oliver Major: Educational Leadership This study examined relationship s b et ween principal turnover rate , and selected school characteristics ( percentage of minority students, percentage of students with disabilities, percentage of students who are economically disadvantaged , and student achievement in reading/English language art s and math measured by Adequate Yearly Progress (AYP) on the Florida Comprehensive Assessment Test (FCAT) ) . This study sample comprised public schools (elementary, middle and high). All of these schools were located in one selected Florida school district . Data were collected AYP reports publicly accessed on the Florida Department of Education w ebsite, and the school D ata Warehouse website. public school district office collected data on the freq uency of principal turnover . Data wer e statistically analyzed using a Quasi Poisson regression model. This study found that an increase in reading and math scores indicates a lower rate of principal turnover. It also indicates that the higher the minority rates in a school and higher percent of low socio economic students, the greater the principal turnover rate.

PAGE 12

12 CHAPTER 1 INTRODUCTION Background M ajor school reform agendas, as recent as ten years ago , were lacking in the area of school leadership ( Porte r, Murphy, Goldring, Cravins, & Elliot, 2008 ; Hull, 2012; The Wallace Foundation, 2013 ). Today, with the role of the principal evolving daily, improving school leadership is a greater priority in regard to school reform (Anthes, 2002). The Wallace Foundat ion known for their work with improving and promoting positive school leadership noted the growing prominence of school leadership as a topic in the national policy conversation about education reform , and as a priority in programs suc h as Race to the Top [ RttT ] (P orter et al. , 2008; Seashore Louis, Leithwood, Wahlstrom, & Anderson, 2010 ; The Wallace Foundation, 2012) . The multi faceted responsibilities involved with school leadership continue to Hayden, Riggens Newby , & Zarlengo, 2002 , Acknow ledgements, para.1 ). The importance of a principal behavior in building the school and classroom conditions that promote student learning has been established in large scale studies (B ryk, Sebring, Allensworth, Luppescu, & Easton, 2010; Seashore Louis et al., 2010). Without effective educational leadership in place , efforts to raise student achievement become nearly impossible (Bouchard et al., 2002). influence upon students, just behind teacher influence, and are often an underestimated role ( Hull, 2012; Leithwood, Seashore Louis, Anderson , & Wahlstrom, 2004) . In years pas t , effective principals and good managers (Bouchard et al., 2002, p.1) . instruction formed th e

PAGE 13

13 et al., 2002 , p.1 ). Some research argues that the school principal is , and has historically been, the primary indicator of stud ent achievement and school success (Bouchard et al., 2002). As the role of principal expands and evolves, an insurmountable challenge arises, if the success or failure of a school rests on this (Bouchard et al. , 2002). Historic ally, a school principal performed like a manager and overseer. T oday , a new concept of school leader h as evolved closer to a model suggested by Jim Collins ( 2001 ): disciplined leaders who rigorou sly focus on people and culture . P rincipals now must commi t to creating a school wide culture , which embraces not only high standards, but also the success of all students ( Murphy, 2008 b ; Wahlstrom & Louis , 2008 ; The Wallace Foundation, 2013 ). With the No Child Left Behind [ NCLB ] (2001) law in place since 2001 , a ll public school principals are held accountable for meeting the required Adeq uate Yearly Progress (AYP). A s defined by NCLB, Adequate Yearly Progress uses standardized tests to determine academic success of all students and schools throughout the nation (NCLB, 2001). Student achievement was to increase each year until 2014 when every school was to reach 100% proficiency on the reading and math state standardized tests (NCLB, 2001). In 2003, in accordance with NCLB , various repercussions were established for those schools that did not meet the AYP standard (Anth es, 2002). The consequences varied from replacing the school administration to the school being taken over by the state. For these reasons, the primary focus and duties of principals shifted to some thing new . No longer was the principal merely the manager and overseer of the school. The principal has become much more than that. He or she is regarded as one who drives student achievement , and is responsible for school improvement. This has

PAGE 14

14 caused a dded pressure to an already stressful job . According to NCLB (2001), if the school is not performing well , the principal is often fired or reassigned. The percentage of low socio economic status (SES) students, students with disabilities, and minority st udents are variables proven to have negatively affect ed overall student achievement. Socioeconomic status continues to be one of the most reliable predictors of student performance (Rainwater & Smeeding, 1995; Rodgers & Payne, 2007). This study addresses variables such as low SES rates and minority rates . It was determined that some of the variables were significant predictors of principal turnover . T he need for school systems to provide adequate succession planning for schools was identified. Principal turnover , ) a term borrowed from the private sector (Murphy, 2008 a ) , is a relevant concern in education today. Problem Statement School distric ts commonly replace their principals in an attemp t to increase student achievement , now that NCLB (2001) specified this option in the law. This may have a negative impact on student achievement as measured by the Florida Comprehensive Achievement Test (FCAT). Principal turnover further affects school m or ale and teacher turnover (Meyer & Macmillan, 2011) . School cult ure takes many years to develop; school impro vement does not occur overnight. Thus, f requent principal turnove r negatively affects Seashore Louis et al., 2010; Leithwood , Louis, Wahlstrom, Anderson , & Mascall , 2009 ). According to T he Wallace Foundation (2013 ), principals should be in a school at least five to seven years for optimum impact. A high rate of principal turnover has negative effects on school

PAGE 15

15 performance, t herefore schools with higher principal turnover rates tend to perform lower overall on standardized math and reading tests (Seashore Louis et al., 2010 ). Employee turnover and employee attrition differ in definition. B oth occur when an employee leaves a n organization . Turnover results from a variety of employment actions, such as discharge, termination, resignation or job abandonment 2014 , para. 1 ) . However, when an employee retires or when the company eliminates a position, it is referred to (Mayhew, 2014) . Mayhew (2014) further states, when turnover occurs, the company seeks s omeone to replace the employee, but i n cases of attrition, the employer leaves the vacancy unfil led or eliminates that job role ( see Employee Tur nover vs. Attrition, para. 2) . It would benefit school districts to pay close attention to school characteristics that lead to a high rate of principal turnover in order to minimize unnecessary replacements . The No Child Left Behind Act (NCLB) and RttT offer t he option of replacing the principal as part of a turnaround policy for struggling schools. However, consistent principal turnover can actually impair a school and stunt school growth, since principal turnover impacts school performance (Beteille, Kalogri des & Loeb , 201 2 ). On the other hand, replacing the principal may significant ly improve a . This particular Florida county had 46 % of their schools with 4, 5, 6, or 7 principals with in a ten year span (see Appendix A ) , while the natio nal average of principal turnover is between 15 30% ( The Wallace Foundation , 2012) . T he new principal must be highly effective , and must have at least five years to improve school performance, including implementing innovative strategies ( Seashore Louis et al., 2010 ; Port in et al. , 2009 ; The Wallace Foundation, 2013 ). If schools replace an ineffective principal with a highly

PAGE 16

16 effective leader , they should offer the new principal incentives to remain at the school for more than five years . This could prove to benefit the school greatly in the area of student achievement ( Porter et al., 2008 ; Leithwood et al., 2009 ; The Wallace Foundation, 2013 ). Part of the retain new, highly effective principals at schoo ls for five to seven years , as per the aforementioned research. Purpose of the Study The purpose of this study was to determine the relationship between school and student characteristics , and principal turnov er. This research examined relationship s amo ng selected school and student characteristics , and principal turnover in one Florida county school distr ict. S chool and s tudent characteristics used as predictors include performance variables (FCAT math/FCAT reading), percentage of minority students, per centage of students with disabilities (SWD), percentage of low socio economic status (SES) students (as determined by free and reduced lunch program), and school type. The strength of the relationships was measured using Quasi Poisson regression. This st udy will add to the body of knowledge on school and student characteristics , and the relationship of these variables to principal turnover rates . This will provide public school systems with valuable information to aid in future principal succession planning, as well as principal transfers within systems. The educational implications of principal turnover are far reaching , and well worth investigating . Significance of the Study Unt il recently, little research was performed on the principa student achievement. I ncrease d accountabilit y and data that are more accessible are beginning to change this . Research studies indicate a significant relationship between

PAGE 17

17 principals and student achievement ( Duke, Carr, & Sterrett, 2 012; Lo eb & Reininger, 2004; Murphy, 2008 a ; Waters, Marzano , & McNulty, 2003 ). One study found that a school led by a highly effective principal performed 10 perce ntage points higher than a school led by an average principal (Wate rs et al., 2003). B as ed on value added scores in one study , percentile to between the 54th and the 58th percentiles in just one year, depending o n the type of analysis (Branch , Hanushek , & Rivkin, 2 012 , p. 11) . The school level, the demographi cs of the students , and the initi al performance of the students all influence principal impact ( The Wallace Foundation, 2013). As significant as principal presence, a principal absence is just as impactful . In a stu dy of Miami Dade County Public Schools (MDCPS) , researchers found that students experiencing a new (to their school) principal performed lower than they had with their previous principal (Beteille et al., 2012 ). If the new principal had prior experience ev en at another school the negative effect was minimized (Beteille et al., 2012 ). Similar results were found when studying principals in North Carolin a (Miller , 2009). Because of the importan role, this study aimed to determine whether there is a relationship between principal turnover , and various school and student characteris tics. I n an effort to assist public schools in succession planning and reten tion of quality principals, this study adds to the bod y of research on Research Question RQ1 : Is there a statistically significant relationship between selected school and student characteristics , and principal turnover? This research question led to the following sub research questions: RQ1a : Is there a statistically significant relationship between principal turnover rate , and percentage of students with disabilities ?

PAGE 18

18 RQ1b : Is there a statistically significant relationship between principal turnover rate , and percentage of minority st udents? RQ1c : Is there a statistically significant relationship between principal turnover rate , and percentage of low socio economic status students ? RQ1d : Is there a statistically significant relationship between principal turnover rate , and reading FCAT scores ? RQ1e : Is there a statistically significant relationship between principal turnover rate , and school type ? RQ1f : Is there a statistically significant relationship between principal turnover rate , and math FCAT scores? Research Hypotheses In determining whether there is a relationship between principal turnover rate , and various school and student characteris tics , t he following hypotheses were developed. H1 : Selected s chool and student characteristics are significant predictors of principal tu rnover rate. H1a : SWD is a statistically significant predictor of principal turnover rate. H1b : Percentage of minority students is a significant predictor of principal turnover rate. H1c : P ercentage of economically disadvantaged students is a statistic ally significant predictor of principal turnover rate. H1d : Reading FCAT scores are statistically significant predictor s of principal turnover rate. H1e : School type is a statis tically significant predictor of principal turnover rate. H1f : Math FCAT sco res are statistically significant predictors of principal turnover rate. Null Hypotheses H 0 1 : S elected s chool characteristics are not significant predictors of principal turnover rate.

PAGE 19

19 H 0 1a : SWD is not a statistically significant predictor of principal tur nover rate. H 0 1b : The percentage of minority students is not a statistically significant predictor of principal turnover rate. H 0 1c : The percentage of economically disadvantaged students is not a statistically significant predictor of principal turnover r ate. H 0 1d : Reading FCAT scores are not statistically significant predictors of principal turnover rate. H 0 1e : School type is not a statistically significant predictor of principal turnover rate . H 0 1f: Math FCAT scores are not statistically significant pre dictors of principal turnover rate. Identification of Variables For the purpose of this study, the following are variables of interest: A DEQUATE Y EARLY P ROGRESS (A YP ) : The federal No Child Left Behind A ct requires that states establish performance g oals fo r all schools, districts and the state to ensure that all students reach 100% proficiency on state assessments by 2014. AYP refers to the intermediate yearly goals that each state must establish. Test scores are analyzed yearly to determine if schools, dis tricts and states have reach ed the intermediate goals, (reached AYP ) . F LORIDA C OMPREHENSIVE A CHIEVEMENT T EST : FCAT L OW SOCIO ECONOMIC STATUS (SES): percentage of students who qualify for free or reduced lunch in the state of Florida (Florida Dep artment of Education). M INORITY : percent age of students categorized as Black, Hispanic, American Indian/Alaskan Native, Asian/Pacific Islander, or multiracial under Florida guidelines (Florida Department of Education). P RINCIPAL T URNOVER : number of times a school c hanged principals during the 2002 03 through 2012 13 school years (Bruggink, 2001). Turnover may result from a number of employment actions, such as discharge, termination, resignation or job abandonment. W hen turnover occurs, the company seeks s omeone to replace the employee, unlike attrition, where the employer leaves the vacancy unfilled or eliminates that job role (Mayhew, 2014) . S TUDENTS WITH D ISABILITIES (SW D): percentage of students receiving special education services in the school (Florida Depart ment of Education, 2012).

PAGE 20

20 S CHOOL TYPE ( ELEMENTARY , MIDDLE , OR HIGH SCHOOL ): For differentiation purposes, 1 was elementary, 2 was middle, and 3 was high school (in the data, for school type). Research Plan This study used a Quasi Poisson regression re search design to determine if there was a significant relationship between principal turnover rate , and select school and student characteristics . T he following variables were used : low socio economic student rate (low SES) , students with disabilities rate (SWD) , minor ity rate (see definition, page 18) , student achievement based on FCAT reading and FCAT math, and school type. O ther research indicates that principal characteristics play into principal turnover as well, however due to privacy policies of the district, this information is not included in this study. A regression researc h design was most suitable for this study because the variables already existed , and no treatment was applied by the researcher (Ary, Jacobs, Razavieh, & Sorensen, 2006). E xperi mental manipulation was not used , nor pre or post testing , or random assignment of subjects to conditions because events had already occurred , and manipulation of variables would have been unethical . Determining whether principal turnover rates affect st udent achievement would be impossible given all the extraneous variables. Th e primary concern was determining relationships among principal turnover, students with disability rate, minority rate, economically disadvantaged rate, student achievement, and sc hool type. Summary The No Child Left Behind Act of 2001 (NCLB) placed increased pressure and accountability on schools to make significant gains from year to year . The NCLB guidelines specified removal of the principal if student achievement does not meet state requirements for AYP ( Anthes, 2002). This change in administration may have more

PAGE 21

21 negative than positive consequences because of possible harmful effects on the school culture and its student achievement. R etain ing quality principals for a suff icient period of time and assist ing sc hools in succession planning, as opposed to disrupt ing school culture and student progress , would be beneficial for optimum student achievement ( Loeb & Reininger, 2004 ; Marzano, 2003; Meyer, Macmillan & Northfield, 200 9; Samuels, 2012).

PAGE 22

22 CHAPTER 2 REVIEW OF THE LITERATURE Introduction During the 1990s to early 2000s , the modern world faced a cavernous gap in the area of leadership in both the public and private sector s . governance that is, the pervasive incapacity of organizations to cope with the expectations of their constituents is [ now ] an overwhelming facto (Rothwell, 2010 , Preface, para. 3 ). The days of apprenticing oneself under a master tradesman are , by and la rge, gone . According to Rothwell (2010 , Preface, para. 4 ), employment contract has changed the relationship between workers and their organizations. Employee longevity was once somet hing that was cherished and respected, but recently longevity is not easily attainable. Likewise, e mployee loyalty is a , less on the people who work for them. In the 199 0s, many businesses began drastically downsizing, and this continues today (Rothwell, 2010). This has led to a problem with em ployee engagement: 19% of emp loyees in the United States actively work against hanging demographics makes the identification of successors key to the future of many organizations when the legacy of the cutbacks in the middle management ranks, the traditional training ground for senior executive positions, has begun to be felt. Demog raphics tell the story: The U.S. , Preface, para. 4 ).

PAGE 23

23 Leadership Turnover and Succession in the Private Sector The crisis of leadership has been most evident in the business aren a . However, this may also extend to areas of education. Joseph Murphy of Vanderbilt University suggests that the information provided to turning around failing businesses could be applied to failing schools as well (2008a). Specifically, the literature i ndicates a strong correlation between turnaround leadership in both business and education. That being said, Murphy looks at the private sector, which also struggles with leadership turnover, but approaches it in a much different way than public education ( Charan, 2005; Murphy, 2008a; Murphy, 2008b). Charan (2005) stated, EO succession process is ( p.105 ). Despite the money budgeted for talent recruitment, the talent pool is still lacking. As a result, the growth of these companies is limit ed (Ready & Conger, 2007). Nearly all human res ( 2007 ) study of 40 worldwide business organizations stated that their pipeline of high potential employees was insufficient to fill strategic management roles (p.1) . Th e survey indicated two primary reasons . First, the professional development or leadership programs no longer align with company needs to stay competitive in modern markets . In an effort to save money, some organizations cut jobs that would offer potential leaders the hands on, valuable training that could prepare someone to easily transition into leadership positions . Second, Ready and Conger (2007) point out that human resource succession residing leaders . Succession planning must be a top down priority in order for organization success and longevity. There are companies out there that manage succession well. Ready and Conger (2007) note two corporations who train up leadership from with in , and rapidly fill organization needs: Proctor and Gamble, and HSBC Group . Although

PAGE 24

24 these corporations are managed differently, Ready and Conger state that they both focus functionality (rigorous talent processes that support strategic and cultural objectives) and vitality (management's emotional commitment, which is reflected in daily actions) 2007, p. 2). Functionality and vitality are two key components to a thriving succession management system. Hill (2008) takes a different approach to chan ging leadership: a global view. Specifically, she feels that we can learn from developing countries and people groups that have only recently been given global business opport unities. In South Africa, Hill notes that vided rigorous leadership preparation ( 2008, p. 123) , to whom previously such ventures were denied. Hill (2008) points out two key elements in both developed and developing countries that she believes will be vital to the future growth of businesses: leading from ( p.123 ). According to Hill (2008) , l eading from behind a time, for the good of the company. C ollective genius refers to looking at the strengths of the group , and fostering these strengths to channel innovation. Hill (2008) further cites HCL T echnologies and IBM as being noteworthy in these areas, though finding leaders who cultivate thes e attrib utes can be challenging. According to Charan (2005), positions such as chief executive officer (CEO) are short lived , and the turnover rate is higher than normal. What is prompting the high turnover rate for top executive s ? Though hired because of their experience, many CEOs lack the development and executive guidance necessary to cultivate them into a long term solution for leadership. Much of the professional development that most

PAGE 25

25 companies invest in rest with the middle management positions , while the time and commitment necessary to discover the top quality CEO candidates is neglected. Moreover , when companies seem to h ave the right person for the position, the effort to develop a succession plan is lacking and, as a result, a fragile foundation is p ut into place, ready to crumble at the first short fall the company experiences (S c haeffer, 2002). Succession planning is merely creating a list of highly potential employees who are capable of filling upcoming vacancies (Schaeffer, 2002). The willingness to be a great leader may be evident in the candidates, but the experience necessary to follow through as a great leader is discounted , and the company suffers long term . According to Charan (2005 ) nce outside CEOs often bring in their own teams, can cause the company to lose focus, and are especially costly to be rid of Employing current employees, equipping them with tools for leadership success, and being cognizant of employee and orga nizational strengths are all important when succession planning (Charan, 2005). Well known companies such as Eli Lilly, Dow Chemical, Bank of America, and Sonoco Products have succeeded in creating a pipeline or pool from which they develop and recruit a spiring leaders (Conger & Fulmer, 2003) . Conger and Fulmer (2003) cite five best practices drawn from these successful organizations. These include an intense focus on growth , identify critical leadership positions , emphasize transparency , administer sum mative and formative frequent assessments , and remain flexible. In Eli Lilly s action learning program, aspiring leaders are assigned a strategic issue to help resolve so they can obtain some hands on training in leadership development . Many best practic e companies rely on w eb based succession -

PAGE 26

26 management tools to map out succession plans. Employees are required to update their profiles and upload needed documents . Interestingly, it was found that b est practice organizations also assess the strengths and weaknesses not only of individuals applying for positions, but of the entire group , to ensure that the applicant would be a good fit overall with those who currently work there (Conger & Fulmer, 2003). These companies commit to continual change , and meet the evolving needs of the organization. Just as in education, a change in leadership is not always easy . Reverting back to past managerial practices and coping mechanisms is the instinctual reaction of most executives when faced with a crisis (Schaeffer , 2002). Similar to schools, w hen a company does well, its CEO (principal) is recognized and praised . When it does poorly, the CEO (principal) is blamed and often let go, or transferred . Unfortunately , investors now identify specific CEOs as the primary reason a company succeeds or fails ( Collins, 2007 ) . Likewise, stakeholders in the education arena tend to associate certain high level individuals as the indicator of school performance. In reality , most companies perform ances remain status quo after fir ing a CEO . Even w company wit Wiersema (2002, para. 1 ). The blame for such poor results, Wiersema said , Lies squarely with boards of directors ( 2002 , para. 1) . The board of directors often lacks the depth of underst anding of the business needed to successfully replace a CEO . The focus tends to be on restoring investor confidence as quickly as possible, and not necessarily on what is in the best intere st of the company . The same issues may continue, even with a new leader in place . According to Wiersema (2002), s uccess lies in having a good board ; t his includes

PAGE 27

2 7 members who take an active role in replacing and equipping a new leader, and less about plea sing the investment community . Instituting these best practices is likely to contribute to organizational stability ( Wiersema, 2002 ). Perhaps , this would decrease the amount of CEO upheaval. Likewise, a healthy school board and superintendent one party n ot having too much power -, looking out for the good of the student, and not just a quick recovery to please lawmakers, could be most beneficial for schools. Harvard Business School professor Bower (2007) found , through studying 1,800 successions, t hat businesses had more long within their organization . He points out that l ack of succession planning results in firms turning to outsiders to fill executive positions : w hen the time comes to name a new chief exec utive, more firms turn to outsiders. According to Bower (2007) , there are strengths and weaknesses for both insiders and outsiders. Insiders are sometimes blind to the need for great innovation, while o utsiders may see the need for the changes, but do no t know the organizational culture well enough to make the changes. Bower (2007) further states that outsiders . leadership ca ndidates that are home grown ; t hey are i nsiders who have an outside point of view. This includes insider candidates being given opportunities to manage and know the culture, yet still be mentored to preserve their outside perspectives (Bower, 2007). This, in theory, preserves the business culture and guiding principles, while ena bling fresh ideas and changes. Collins and Porras (1994) , in a six year research project at Stanford University Graduate Sc hool of Business , studied what makes exceptional business truly exceptional. They studied eighteen long lasting companies that had an average age of

PAGE 28

28 nearly 100 years . Each company was examined in comparison to competitors at multiple time increments throughout the age and growth of the company. They identified companies like General Electric, 3M, Motorola, Johnson & Johnson, Boeing, Wal Mart, Hewlett Packard, Procter & Gamble, Walt Disney, and Philip Morris Collins and Porras (1994) also found that several commonly held myths were just simply not true. This includes the myth that the only way for companies to implement significant change is to hire a CEO from outside their current company . One of the most grown man agement plan six times more often than their competition, who were less successful . According to their study, there is significant evidence to support the idea of promoting from inside the corporation , and the ability to stimulate significant change . These companies have shown repeatedly that they do not need to hire top management from the outside in order to get change and fresh id eas. Likewise , their study demonstrates the challenge of remaining a highly visionary company when hiring executives from outs ide the organization . These findings are also consistent with the education arena . If schools want to consistently perform well, the literature reveals the importance of providing a strong pipeline or pool program as a successful way to train up new lead ers, while still embracing change and staying competitive. Leadership in Schools Leaders hip accountability is not only being emphasized in the private sector, but within public schools as well . Recent changes in American society are contributing to a more intense look at the prin cipalship (Levine, 2005). High stakes learning and increase d accountability place greater demands on the principal. In 2001, the No Child

PAGE 29

29 Left Behind L aw (NCLB , 2001) reinforced the idea that principals are accountable for Adequ ate Yearly Progress (AYP). In 2009, President Obama administration designated $4 billion toward turning around low performing schools (U . S . Department of Education [DOE] , 2010). Consistently low performing schools are eligible under this initiative to receive federal fu n ding. However, i n order to receive the funding , they must commit to drastic changes as outlined in the law. One of the changes include changing staff restructuring , and often a change in principal (Dillon, 2011; Tucker, 2010; U . S . Dep artment of Education, 2010). Since p rincipals affect the school and its student performance , this can be a beneficial change at times (Hallinger & Heck, 1996, 1998; Leithwood et al., 2004) ; h owever, Dillon (2011) believes that improvement effo rts can be compromised when there is a change in leadership. Thus, t he decision to restructure a school should be carefully considered. B etween 2011 and 2014, under the Obama administration, the National Governors Ass ociation developed Common Core state s tandards with funding from Was hington, D.C. (Burke, 2013 ; Eilers, 2012 ). Common Core state standards are a highly controversial subject, due to strong feelings on both sides about fede ral intervention into education and local control being taken away (He rzog, 2014) . The Some claim that the CCSS represent s a daunting challenge for school leaders (Riddile, 2012). Th e principal job structure is changing . Principals now work in direct partnership with other school based leaders. Responsibility, accountability, and stress level have all

PAGE 30

30 leadership role . As these ch anges in education take place, b ased on research, it is likely that this type of change in job focus could create a higher rate of principal turnover. Lawmakers tied $4.35 billion in Race to the Top (RttT) grants to individual state adoption of these stan dards (Burke, 2013). The adoption of Common Core plays a major role in securing a No Child Left Behind waiver. The waivers were created by lawmakers when they realized that was unattainable. In 2012, Fl orida submitted an E lementary and S econdary E ducation A ct (also known as NCLB) Flexibility Request (a re quest for a waiver) (DOE , 2014 ). According to the website, the flexibility waiver enabled Florida to streamline the accountability system . Bet ween 2011 and 2014, the State Education Agency (SEA) and local education agencies (LEAs ) specifically, the RttT and Leader Preparation Committee worked together to revise Florida Principal Leadership Standards (SBE Rule 6A 5.065, Florida Administrative Cod e). The Florida Administrative Code define s expectations for effective school administrators maintaining standards based learning for student achievement (DOE, 2014). These changes influenced evaluations of admin istrators and teachers by LEAs. Two primary documents represent guidelines for local teachers and principal evaluation systems: Section 1012.34 of the Florida Statutes (Personnel evaluation procedures and criteria) , ing LEA Memorandum of Understand ing (MOU, Attachment 10b). F or implemen tation of evaluations, two governing rules are in effect to help LEAs: Rule 6A 5.065, FAC , The Educator Accompli shed Practices (Attachment 10c); and Rule 6A 5.080, FAC, Florida Princip al Leadership Standards (Attachment

PAGE 31

31 10d). These rules guided the development of this district newest evaluative model. This model has been in effect for three years , (2003) School Leadership Evaluation Florida Model. Marzano (2003) indicates that the leadership of the school principal is the most crucial aspect of effective school reform . 172) . Waters, Marzano, and McNulty (2003) identified twenty one leadership responsibilities associated with student achievement. Amo ng those twenty one responsibilities were school culture, order, discipline, situational awareness , and in et al. , 2003 , p.3 ), to name a few. All of the responsibilities are critical, but naturally take time, energy, and res ources in order to be effective . The Florida is based on (2003) School Leadership Evaluative Model , and describes five domains (1: A data driven focus on student achievement, 2: Continuous improvement of instruction, 3: A guaranteed and viable curriculum, 4: Communication, cooperation, and collaboration, 5: School climate) , with 26 elements of school administrator behavior that research shows has an impact on raising stude nt achievement ( Marzano, 2003 ). A principal wh o performs well in these areas is considered to be applying successful leader ship characteristics. Due to the Marzano philosophy (2003) behind what a highly effective principal looks like in the state of Florida, if principals are not observed performing well in the designated areas, they could become discouraged and decide the job is not for them. This increase in accountability and rigor could be a cause of principal turnover: principals could leave the job altogether if doing poorly or cannot keep up, or be transferred within district ( if doing well ) to help with turnaround schools.

PAGE 32

32 In 2011, Florida Governor Rick Scott signed Senate Bill (SB) 736 The Student evaluation is based on the performance of the students assigned to the school over a 3 teachers, impr ovement in the percentage of classroom teachers evaluated at the effective or highly effective level, other leadership practices that result in improved (SB 736) . According to recent research, many are opting out of the principalship, feeling it is just , McCloud , & Podmostko, 2000). and more quality le aders are opting out. M ore recently, there has be en a change in how district level leadership deals with the accountability of principals (Samuels, 2012; Fuller, 2012 ). The idea of ( hi gh turnover rate of principals) is a seemingly pop ular strategy . The strategy suggests that frequent change of school based leadership ( in an effort to make the grade ) will render positive results (Fuller, 2012 ). P has long lasting effects , and seems to be most detrimental in challenging schools (Samuels, 2012). Districts eager to make changes may actually be causing more harm than good. Principals who are new to a school encounter many issues that principals promoted from within a school do not (Lashway, 2003; Weinstein, Jacobowitz, Ely, Landon , & Schwartz, 2009). Not only do they feel the need to quickly adapt and assimilate to a new environment, but they also have the day to day procedures of being

PAGE 33

33 a principal to learn and implement (Lashway, 2003; Weinstein et al., 2009). First year p rincipals have added stressors, which could be an implication of the high turnover rate for principals in the first five years. In a study of New York City (NYC) public schools , there was very little significance between principal experience as a teacher , and math and reading scores ( Cohen, McCabe, Michelli , & Pickeral, 2009 ). However, of greater significance was principal tenure at the same school , and math and reading scores ( Clark, Marorell, & Rockoff, 2009 ). Dhuey and Smith (2011) studied principal i mpact in British Columbia , and found that it took a principal being at the same school five years to measure significant impact. Principal Turnover On average , across the United States, annual principal turnover rates range between 15 to 30% ( Béteille et al. , 2011; Dhuey & Smith, 2011; Gates et al. , 200 5 ). Fuller, Young, and Orr (2007) found that 50% of three different cohorts, principals left their Texas schools within the first three years. Small, rural districts in Washington State indicated that 56 % of principals left their schools within five years after their initial placement (Elfers, Plecki, & Knapp, 2006). Illinois and North Carolina indicated a principal turnover rate of 63% and 79%, respectively, at the six year mark (Gates et al., 200 5 ). Wei nstein and colleagues ( 2009 ) found that between 1993 and 2002 (approximately ten years), new high schools in New York City experienced high principal turnover rates: 48% of schools experienced one princ ipal change , and 36% experienced two or more (Burkhaus er, Gates, Hamilton , & Ikemoto, 2012).

PAGE 34

34 The RAND Corporation study ( Burkhauser et al., 2012 , p.17 ) states and case study analysis lend support to the notion that rapid turnover ste ms from school, district, or [Chief Marketing Officer] CMO cho ices rather than from individual principal choices The survey responses of the RAN D Corporation study indicate that frequent turnover is not principal driven, but district driven. This is the i dea prev iously mentioned principal churn. The principals mentioned not voluntarily moving positions. Many planned to seek principal jobs outside of their district, or pursue a job other than principal within their same district (Burkhauser et al., 2012). Low Performing Schools Low performing schools appear to endure the most negative effects of principal churn size fits all formula ; t here is no quick fix. At times , it is logical to allow principals extended time to implement change . Ultimately, districts must consider each situation on a case by case basis. Districts must ensure that they are hiring the best school leaders for each job (Samuels, 2012). Under NCLB/ESEA Academic Assessm ent and Local Educational Agency and School Improvement 1116 (b) , school improvement is mapped out, along with consequences for lack of adequate yearly progress. Granted, the schools that require restructuring are Title 1 schools , which have had many oppor tunities for improvement. According to the Act, restructuring occurs after six years of no adequate yearly progress, known as the Fourth Tier (Rivers & Sattler , n.d. ). When the state steps in, it is not always to reconstitute schools in need of improvemen t (SINI). This action is

PAGE 35

35 Turnaround Schools Turnaround schools are in crisis. Until recently , there was little empirical start in the private sector, where turnaround refers to corporations, churches, and political parties , and data has been collected through the years ( Brobeck , 2011). The private sector typically has a very cut and dry response to turnaround: immediately f ire the bosses. In education, however, there is a much longer standing plan in place ( three to five years), and often the employees are the first to be terminated. Another difference between turnaround in the private sector and turnaround in education is that the former revolves around budgets and money. Whereas other areas begin with efficiency moves, education is slow to make changes ( Brobeck , 2011) to say, ( as cited in Brobeck, 2011 , para.12 ). The private sector handles turnaround much more aggressively. Unlike the No Child Left Behind initiative , the U.S. School Improvement Grants (SIG) take s a targeted approach to identifying low performing schools , and providing funding to turn them around. The grants permit four different interventions , which are intended to improve the academi c achievement of the lowest performing five percent of schools. Of the various interventions possible, actually closing a failing school and sending its students elsewhere would appear to be the most extreme solution, and, in fact, it is the intervention l east chosen by districts receiving SIG funding. The independent think tank, Education Sector (2011) % of SIG schools are pursuing a

PAGE 36

36 transformation intervention, 21 % are implementing a turna round model, and 4 % have chosen to restart as charter or privately managed schools. Only 2 % of SIG schools are slated to be closed. It is not known yet w hether such a large number of consistently low performing schools can be turned around . Many p ast turn around efforts have not been entirely successful examined the effects of turnaround efforts in a group of 2,025 low performing schools in 10 states. Schools both district and charter schools -were tracked from 2003 04 to 2008 09. After five years, 80 % of the district schools , and 72 % of the charter schools remained low performing. Ten percent made moderat e improvement, while only 1 % improved dramatically. Only 19 % of the charters , and 11 % of district schools were closed ( St uit, 2010 ) . , Stuit (2010) concluded that turning around a school can be extremely challenging , and charter authoriz ers and education policy makers should give more consideration to closing failing schools, especially when higher performing schools are close by. However, Collins (1995 a ) reported , that urge dramatic change and fundamental transformation on all fronts are not only wrong, they are dangerous. Any great and enduring human institution must have an underpinning of core values and a sense of timeless purpose t hat should never change a. 4). Whether restructuring , and must be handled carefully. Private sector organizations can move quickly and

PAGE 37

37 efficiently, and maybe it is because they do not necessa rily involve critical stakes , such as the education and future of children ( Brobeck , 2011) . If corporations took five years to implement a plan of action, they would be bankrupt ( Brobeck , 2011) and long gone. Mass Insight Education, a 501(c)(3) non profit organization based in Boston, Massachusetts transform s public schools into high performance organizations , and aims to minimize the achievement gap between high poverty and low poverty schools (Calkins, Guenther, Belfiore, & Lash, 2007) . They argue tha t comprehensive and dramatic change in low performing schools will raise student achievement within two years of restructuring (Murphy as cited in Brobeck , 2011) . Mass Insight Education, a nd other similar organizations, provide s a service to attempt to sa ve a school from failing, before the State intervenes and restructures it. Restructuring includes many options, one of which is reconstitution (Rivers & Sattler , n.d. ). One of the steps, according to the law, is to change the governance structure of the s chool . T his can involve removal of the principal. With this added weight and accountability level, principals are feeling the pressure more and more. In some cases, before the school can get to the point of restructuring, the s uperintendent that a principal move t o another school, as a preventive measure. This can be a positiv e or negative move if a principal is doing a great job in one location , a superintendent may make a strategic move to help a struggling school with a new, proven princi pal. Generally, when a principal leaves a position, it is not due to being fired (Gates et al., 2005). Although there are no national statistics , as a whole , on principal turnover, small scale studies show that most principal turnover occurs within distri cts not

PAGE 38

38 dismissals (Gates et al., 2005; Loeb et al., 2010 ; Ringel, Gates, Chung, Brown, & Ghosh Dastidar, 2004 ). Murphy ( as cited in Brobeck, 2011) believes that the first move in turning around a failing school should be to replace the principal . Declini ng schools share a common denominator: poor leadership (Duke, 2008). According to Duke (2008) , capable leaders are extremely important when turning around schools . Leadership reassignment is , at times, believed to be in the best interest of the schools. Frequent turnover is believed by some to benefit overall school improvement. However, the majority of literature does not provide sufficient evidence that this is beneficial. According to Loeb et al. (2010), the majority of principal turnover is voluntar y. According to Fuller (2012), there are three key reasons principal turnover is worth looking into . First, high principal turnover often causes an increase in teacher turnover, which is related to poor student achievement (Béteille et al., 2011; Fuller, 2012; Fuller et al. , 2007) Not only can teacher turnover negatively impact school improvement efforts, it can negatively affect other school outcomes as well ( Fuller, 2012; Fuller et al. , 2007 , Ronfeldt, Loeb , & Wyckoff, 2011 ) . Moreover, it can create fi nancial burden s on the district (Levy, Fields , & Jablonski, 2006). Ronfeldt et al. (2011) turnover has a significant and negative effect on student achievement in both math and ELA . . . [and] is particularly harmful to students in scho ols with large populations of low perfor (p. 6) . Principal turnover affects teacher turnover, which strongly impacts student achievement. Second , recent research confirms that there is a direct relationship between principal turno ver , and school and student outcomes ( Béteille et al., 2011 ; Burk hauser

PAGE 39

39 et al. 2012; Miller, 2009 ). The strongest impact seems to manifest itself immediately after turnover occurs ( Béteille et al., 2011 ; Burkhau s er et al. , 2012; Miller, 2009). Sch ool distr icts should keep this information in mind when creating principal succession plans. Finally , the literature indicates that frequent principal turnover lends itself to teachers not investing in vital changes, but rather choosing to ( Hargreaves et al. , 2003, p. 8). If teachers are not investing in the school improvement initiatives, the necessary growth of the organization will be stunted (Fullan , 199 9 ). Having principals committed to a minimum of five years could increase the likelih ood of large scale school improvement (McAdams , 1997; Seashore Louis et al. , 2010). The national MetLife Survey of the American Teacher: Challenges for School Leadership (2012) indicates that three of four K 12 public scho ol principals feel that the job co mplexity has become more than they can handle. Approximately one third says they will most likely change careers (Heitin, 2013). Most principals claim that the principal job has changed over the last five years, and they are often under great stress (H eitin, 2013). There has been a significant decline in job satisfaction : 68% said they , and 59 % percent said so in 2013 (Heitin, 2013). Berrong (2012) found that the increases in responsibilit y and stress level are causing pr incipals to reconsider their career choice, and leave the profession at a higher rate than ever before. For example, principals in Illinois and No rth Carolina have a yearly turnover rate between 14 and 18% ( Rand, 2007). Two thirds of principals leave the ir position within the first six years on the job in New York (Papa, Lankford, Wyckoff, 2002). The majority of these were intra district moves , or principal positions in another

PAGE 40

40 district . Also, the larger the student population, the higher the principal t urnover rate (Papa et al., 2004 ). According to Fuller, et al. (2007), p rincipal turno ver occurs less in suburban areas, schools with small student populations, and schools with higher principal salaries. In Texas, 50% of administrators left their positio ns within the first five years of their career. W ithin 10 years, 75% of principals left school level leadership pos itions. Principal characteristics such as sex and age also factor into principal turnover rate. F emales leave at a greater rate than males . Additionally , age played a significant role in principal turno ver. Principals older than age 46 were more likely to leave their principal position than principals 46 and younger (Fuller et al . , 2007). A nnual turnover rates for prin cipals range between 15 % and 30% (Beteille et al., 2012 ) . Like leadership turnover in other professions , turnover rates at more challenging schools are on the higher end of that spectrum (Beteille et al., 2012; Boyd, Grossman, Lanksford, Loeb , & Wyckoff , 2008; Clark et al., 2009 ). For instance, large, urban districts have similar turnover rates: Miami Dade County Public Schools (MDCPS) is 22%, Milwaukee is 20%, San Francisco is 26%, and New York City is 24% (Beteille et al., 2012; Boyd et al. , 2008; Clark et al. , 2009 ) . However, within MDCPS , the turnover rate is higher poverty schools (28%) , versus lowest poverty schools (18%) (Bete ille et al., 201 2 ). New York City school findings were similar (Clark et al., 2009). Low performing, low SES schools that lose an effective principal , often struggle due to having a replacement that is less experienced and less effective (Beteille et al., 2012; Branch et al., 2012).

PAGE 41

41 The Wal lace Foundation is conducting a five year study on the following large school districts : Charlotte Mecklenburg Schools, North Carolina; Denver Public Schools, Colorado; Gwinnett County Public Schools, Georgia; Hillsborough County Public Public Schools, Maryland. The study is called the Wallace Principal Pipeline Initiative. According to The Wallace Foundation (2013) that when an urban district and its principal training programs provide many talented aspiring principals w ith training, evaluation, and support following these specifications, the result will be a pipeline of principals able to improve teaching quality and st udent Summary, para.2 ). This theory is similarly aligned with the aforementioned researc h and trends on corporations. Not all principal turnover is negative (Brown, 1982; Denis & Denis, 1995). Some turnover side effects include purging ineffective leaders and cultivating new innovative ideas (Brown, 1982; Denis & Denis, 1995 ) . If ineffect ive leaders are often the principals who leave, it could end up being beneficial to the school. Too much turnover, however, can harm school culture, induce teacher turnover, add fiscal burdens to the district, and create overall organizational instability (Abels on & Baysinger, 1984; Hull, 2012; Grusky, 1960; Mobley, 1982). Leadership turnover often causes disruption and upheaval in the workplace, whether it is within corporations, higher institutions, or K 12 schools. The question remains, how do school ch aracteristics relate to principal turnover? Specifically, how do SES, SWD, minority rate, FCAT reading scores , FCAT math scores , and school type relate to principal turnover in K 12 public schools?

PAGE 42

42 Principal Turnover and Low Socioeconomic Schools There i s plenty of literature supporting a link between low SES and student achievement ( Hull, 2012; Kannapel, Clements, Taylor , & Hibpshman, 2005) . The link is so well established that people often make the false assumption that low SES students are destined for poor academic performance. Despite the fact that many educators do not buy into this way of thinking , it is true that high poverty students and schools face a unique set of obstacles to overcome. Disheartening trends appear when principal changes and ca reer paths are studies, in regard to urban schools ( Fuller et al., 2007 ; Mitang, 2003; Papa et al., 2002 ). It is very difficult to attract and retain principals at schools with high percentages of low income students (Mitang, 2003). New York urban school p rincipals tend to be less experienced , and have graduated from less prestigious colleges (Papa et al., 2002). Papa and colleagues (2002) also found that New York urban principals moved to new positions out side their school distric t more often than their su burban districts peers did. S uburban principals also remained in their positions more often , while urban principals left the principalship at a higher rate. A study in Texas using low socioeconomic status as variables to determine principal turnover rates found that p rincipals in low SES sc hools were promoted to district leadership positions at a higher rate than princi pals at high SES schools (Fuller, Young, & Orr, 2007 ) . This created a pipeline to district level positions through low SES schools. By d SES schools to leave their schools to fill the low SES principal jobs. In North Carolina, the schools with the highest principal turnover rate were the highest poverty schools (Clotfelter, La dd, Vigdor, & Wheeler, 200 6). Principals in low performing, high poverty North Carolina schools tended to have

PAGE 43

43 previously been teachers or assistant principals at the school they were leading; also , they had lower principal certification test scores than high achieving, low poverty schools (Clotfelter et al. , 2006). (Loeb et al., 2010; Papa et al., 2002a). It is not unusual for school leaders to feel that as their years of tenure increases, a benefit of seniority would be t ransferring to a school with a higher SES population (Loeb et al., 2010; Papa et al., 2002a) . It is very rare for principals to receive a salary increase when they change schools within a district; therefore , int ra district transfers can provide non finan cial benefits ( Horng, 2009 ) . Safety, school culture, facilities, and student characteristics are some of the strongest indicators of teacher and principal preference when choosing a school (Horng, 2009; Loeb et al., 2010; Loeb & Rein inger, 2004). Schools w ith less appealing characteristics typically do not receive as many applicants for school leadership positions , and therefore cannot be as selective about who they place in s aid positions (Roza, 2003). S tudent absences are lower in schools led b y effective principals than they were in schools led by less effe ctive principals (Branch et al., 2012 ) . T he impact was most obvious in low performing and high poverty schools than in high performing and low poverty schools (Branch et al., 2012; Clark et al., 2009) . Likewise, principals can affect can have a graduation rate nearly 3 percentage points higher than a high school led by an average principal (Coelli & Green, 2012 ). However , it t akes time for even highly effective principals to hav e such an impact. On average, principals graduation rate begins in their second year at the school. The full effect is felt at the

PAGE 44

44 four year mark (Coelli & Green , 2012 ). This i ndicates yet another incentive for districts to encourage principals to remain in the same position for a minimum of five years. Principals were found to be most effective in high poverty schools than principals in less challenging settings. A study by the CALDER Center ( Branch et al., 2012 ) used value added model (VAM) scores to determine the impact of principals. The impact, according to VAM scores , was poverty schoo ls as in low poverty ( Branch et al., 2012 ) . In 2009, the CALDER Center conducted a similar study which found that the greatest impact of principals was in large , high poverty schools than in large , low poverty schools (Branch et al., 2009, 2012; Clark et al. , 2009; Loeb et al., 2010). P rincipals appear to have most impact in high minority, low performing schools (Branch et al., 2012) . High performing, high poverty schools share commonalities t hat differ significantly from practices in lower performing, high poverty schools . T hese traits includ e a school wide ethic of high expectations; caring, respectful relations between stakeholders; a strong academic and instructional focus; regular assessment of individual students; collaborative decision making structures and a non authoritarian principal; strong fa culty morale and work ethic; and coordinated staffing strategies (Kannapel et al., 2005) . A strong principal who has built a positive rapport with students, faculty, and the community would be desirable in a high poverty school in order to implement thes e traits effectively . Students with Disabilities (SWD) The only subgroup that did not make annual yearly progress on the 2013 FCAT reading and math scores was students with disabilities (SWD) ( F lorida Department of Education, 2013) . This particular popul ation faces unique challenges , and not just in the

PAGE 45

45 area of academics. According to Dyson (2010 ), students with disabilities may have aking, reading, writing, reasoning, and mathematical 44) . These challenges t ypically obtain knowledge at the same rate as their non disabled peers (Cortiella, 2007 a , 2007b ). It has been shown that students with disabilities with average intelligence are not as successful as students without disabilities of equal intelli gence, because of their cognitive processing deficits ( Johnson, Humphrey, Mellard, Woods, & Swanson, 2010 ). According to Johnson et al. (2010), a primary characteristic of SWD s is poor academic performance. Many students cla ssified as SWD require an individual education plan to meet their specific needs (Mattison, 2008). Many SWD s find reading in particular a challenge . Reading fluency is a skill that students must develop at a young age , or long term reading comprehension issues may arise (Chard, Vaughn, & Tyler, 2002). According to Torgesen (1989), many students with poor reading s kills early in life continue with comprehension difficulties , and identifying or decoding printed words . These students rarely catch up , becaus e they lack the reading practice needed to restore missing skills ( Rashotte, Torgesen, & Wagner, 1997 ). Students with disabilities are more likely to drop out of school (Thurlow, Sinclair & Johnson, 2002); t he primary reason for the drop outs was poor ac ademic performance. Although the F ederal G overnment has spent billions of dollars to help close the achievement gap between students with and without disabilities, the gap still remains ( Meyer, 2004 ). On e of the performa nce goal s was to the achievement gap between SWD and students without disabilities , however, the SWD s are still found lacking on the state achievement tests (McLaughlin,

PAGE 46

46 2010 ) . The No Child Left Behind Act requires that all students , regardless of disability , take state annual assessments to determine adequate yearly progress . Migrant Students families. There are many challenges faced with educating migrant stude nts. Poverty, lack of educational support, and discontinuity of education are just a few of the issues that migrant families face (Education World, 2013). Some counties offer migrant students summer school to offset these difficulties , and help with sta bility and continuity of learning. Migrant students may spend part of the year in one state, and then move to another state for the remainder of the year. Educators have difficulty tracking students who move state to state (Crary, 2005) , and many times, the students fall behind and end up dropping out of school (Education World, 2013) . The cycle of un education and poverty that their parents came to America to leave behind, continues (Crary, 2005). With the country, and specifically the state of Florida , in one of the largest demographic shifts, and immigration influencing social, political and education reform, this subject cannot be ignored (Green, 2003). The Florida county that this study was conducted in has an area that serves migrant families, and has consistently low performing schools. Likewise, the principal turnover rates at these schools are high. It could be that the high migrant population influences the principal turnover rate, but due to lack of documentation, this was not a variable in this study. Certification and Training A relationship has been found between schools with poor student achievement on standardized tests , and the principal (Baker & Cooper, 2005). Principals, who failed the principal certification test and had to take it m ore than once, generally oversee

PAGE 47

47 schools with a larger amount of uncertified teachers, than do principals who took the test once and passed (Baker & Cooper, 2005). Baker and Cooper (2005) also discovered that schools with higher percentages of minority stu dents were led by principals who initially had failed their certification exam , whereas high performing schools were found to have the opposite circumstances: higher numbers of certified teachers and principals who did not fail their certification test . Pr incipals who attend notable universities are more likely to hire and retain highly qualified teachers than principals wh o attend less selective universities ( Baker, Orr , & Young, 2005; Clotfelter et al., 2006). Additionally , the principals who ac quired the ir trainin g at highly selective universities were placed into school leadership positions more rapidly than their peers who re ceived their training from less selective univ ersities (Fuller et al., 2007). Policymakers and district leaders must put better p rincipal selection processes in place in order to find the best qualified candidates to fill positions in struggling schools (Samuels, 2012). Teacher Turnover Principal turnover can lead to increase d teacher turnover, which causes negative ramifications t o the overall school performance (Béteille et al. , 2011; Fuller et al., 2007; Miller, 2009; Papa, 2007). It should be noted, that this type of research is co rrelational rather than causal; it is difficult to determine whether principal turnover is leading to , or being caused by these outcomes. I n Texas schools, there was a significant relationship found between principal tenure and percentage of veteran teachers working at a school (Fuller et al., 2007) . The more experienced principals retained teachers , and ge nerally hired veteran teacher s (Fuller et al. , 2007). These veteran principals also had lower numbers of uncertified teachers (F uller et al. , 2007).

PAGE 48

48 Results from t eacher surveys school, base d on principal leadership critique (Sparks, 2013). This points out a critical element for districts to consider , and emphasizes the need for a strong selection and placement process for principals within their district. Effective Principals a subjective term. Stogdill (1974) , as well as Bass and Stogdill (1990) , , divergent interpretations . Burns (1978) states, is one of the most observed and leas t understoo d phenomena on earth ( p. 2). According to Dave Ulrich of University of Michigan, in April of 2014 , there were 480,881 books whose topics hav e to do with effective leaders. Most common definitions include both guidance (of the school) and exercising infl uence (Leithwood & Riehl, 2003) . E ffective leaders influence student learning by casting a vision, setting school wide goals, and making sure resources and procedures are in place for effective instruction (Leithwood & Riehl , ctive leadership definition focuses more on results and student achievement. Murphy , Elliot, Goldring, and Porter ( 2006) emphasize learning centered leadership , which includes a process of influencing others, rather than a set of traits or skills. Process , influence and purpose are all what they claim mak es effective principals (Murphy et al. , 2006). As stated previously, the term s not easily defined. However, for the purposes of this study , an principal is one whose students make greater than average gains than similar students in other schools , as measured by the Interstate School Leaders Licensure Consortium (ISLLC) . In the field of education, we have The Interstate School Leaders

PAGE 49

49 Licensure Co nsortium Educational Leadership Policy Standards (ISLLC) of 2008 , which Standard 1: A school administrator is an educational leader who promotes the success of all students by facilitating the dev elopment, articulation, implementation, and stewardship of a vision of learning that is shared and supported by the school community. Standard 2: A school administrator is an educational leader who promotes the success of all students by advocating, nurtur ing, and sustaining a school culture and instructional program conducive to student learning and staff professional growth. Standard 3: A school administrator is an educational leader who promotes the success of all students by ensuring management of the o rganization, operations, and resources for a safe, efficient, and effective learning environment. Standard 4: A school administrator is an educational leader who promotes the success of all students by collaborating with families and community members, res ponding to diverse community interests and needs, and mobilizing community resources. Standard 5: A school administrator is an educational leader who promotes the success of all students by acting with integrity, fairness, and in an ethical manner. Standar d 6: A school administrator is an educational leader who promotes the success of all students by understanding, responding to, and influencing the larger political, social, economic, legal, and cultural context. According to ISLLC (2008), vision and teache r buy in are important, as is building a positive school culture. An effective leader also collaborates with the community , and encourages community involvement from the school members . The effective leader is to lead a school ethically and cognizant of h is or her great, multi faceted influence. One important key to principal effectiveness is strong instructional leadership ( Robinson, Lloyd, & Rowe, 2008). Instr uctional leadership includes everything from designing instructional strategies to implementing graduation requirements and the many steps in between (Murphy, 2008b ) . Principals also affect students by selecting

PAGE 50

50 and hiring effective teachers, keeping teachers motivated, and ensuring the selected curriculum is being taught (Branch et al. , 2011; Brew er , 1993; Eberts & Stone, 1988; Fuller et al. , 2010). Samuels (2012) examined principal success and how various approaches and activities influenced success . He found that rincipals focused most or all of their time on the following activities: promoting data use, observing classrooms, creating a healthy school culture, forming leadership teams, and promoting teacher professional development (para. 10). However, he found no link between student success , and principal time spent on each of these activitie s . The principals who were highest in success rate were those who effectively implemented change, and those whose staff bought into the intended changes . S taff buy to stay in the position a second year. Furth ermore, according to Samuels (2012), p rincipals who respect existing school philosophies, rituals and practices of the school culture , are more likely to be successful. He also found that t his practice of including existing traditions into school improvem ent plans encourages staff cohesion , which influences school culture ; thus, employees are more likely to embrace the necessary changes needed for school improvement. Peter Drucker, author of The Practice of Management (1954) and The Effective Executive ( 1967) , says that an unclear vision from a leader is one of the easiest ways for organizational failure . A clear, concise vision from the principal of a school is vital . It allows teachers, students, and parents to know where they are headed and what is e xpected. Vision and mission are so crucial to leadership, that without them, failure is

PAGE 51

51 imminent (Drucker, 1967). Setting clear goals and a vision on how to reach those goals help overall school performance ( Druker, 1967; Marzano, 2003; McEwan, 2003 ) . A s discussed earlier, Florida has adopted a principal evaluation tool based on , , . the Florida Principal Leadership standards, SBE Rule 6A 5.080. Principal Assessment Data The Dallas examination principal policie comprehensive longitudinal data systems collect limited information about (Sparks, 2013 , para. 5 on their rada other states , like Rhode Island, are developing over arching systems to track principals from their training programs through licensing, placement , and school leadership , para. 5 ) . Former Flo rida Commissioner of Education , Dr. Tony Bennett , said states need to establish criteria to evaluate princ state and district tify whether a principal is effe ctive by using data chievement status and teacher quality , para. 17 ). Interestingly, according to recent reports, despite tight budgets, Denver Public Schools has hired more people to coach and evaluate leaders (Gill, 2013). This indicates a strong need for principal support and accountability. Denver Public Schools Superintendent Tom Boasberg says that principal supervisors are the most important li nk between district leadership and our schools (Gill, 2013).

PAGE 52

52 This study found that principal turnover dat a is not collected at the state DOE level, at least not in a comprehensive manner. However, a similar study on principal turnover from the fourth largest county in the United States, Miami Dade County Public Schools (MDCPS) found that their rates of turno ver were similar to o ther studies : Milwaukee and North Carolina schools have annual turnover rates of around 20%, San Francisco has a turnover rate of 26%, and New York City has a turnov er rate of 30% (Bet eille et al. , 2012 ). Although the demographics are different between MDCPS and the selected county for this study , there are various similarities ; the MDCPS study was helpful in gathering comparable research and a review of the literature. Summary According to the research presented, it is clear that ch ange in leadership is a pressing issue not only in the business world, but in education as well. Quality principals are a necessary part of the school success. Cash (2008) points out that may be no clear definition of the word leadership, th e research is very clear in always identifying effective leadership as one of the most critical components in effective principal leadership research indicates that the principal job responsibilities are becoming too numerou s and overwhelming (Gilson, 2008). The abundant responsibilities make it a challenge to determine whether a principal has direct impact on student achievement. Furt her research should be conducted on the relationship between effective principals and stu dent achievement . It takes time for a principal to implement change and affect school culture. It could be implied that principals who adequately communicate and implement their vision, can affect school culture and ultimately student success . Likewise, principals who are forced to change schools frequently may

PAGE 53

53 not be given the opportunity for success ful outcomes . In education , we must strive to see past the common use of buzzwords and fads , and implement timeless qualities in ways that are applicable to 1994) . School leaders should , leaders . This study will help schools determine what retains quality principals , and what turns them away. It will determine the relationship between principal turnover , and select school and student characteristics in one Florida school district.

PAGE 54

54 CHAPTER 3 METHODS Participants T his study was based on public schools in one selected school district in Florida ( A ppendix B ). The school d istrict has 47 schools , and serves a total student population of 44,573. There are 29 elementary schools, 10 middle schools, and 8 high schools. One of these schools physically meets in one location an d goes by the same name, but reports data as three independent schools: elementary, middle and high school (a combined school). There are also 12 Alternative School Programs. Participating schools were a combination of elementary schools, middle schools and high schools. Charter schoo ls and alternative programs were eliminated. Setting The setting for this study was in one selected school district in Florida . The schools populations vary fr om 200 students to nearly 3,000. The schools have other differences such as a varying percenta ge of SWD, ELL, and economic status. I chose not to restrict participation because of these differences, but rather compare these variables with the frequency of principal turnover. Therefore, only charter schools and alternative schools and programs were excluded. Instrumentation Various instruments wer e used in data collection. For the purpose of this study, the 2013 student scaled scores from the FCAT were used, as calculated for AYP by the state of Florida , in reading and math (Appendix C ) . Principal turnover was measured by th e number of principals leading a given school. Schools with the same principal for the past 10 years (or longer) were given a

PAGE 55

55 one as the principal turnover r ate, after offsetting the number of years. Schools with ten separ ate pri ncipals over the past 10 years were given a ten as the principal turnove r rate. P rincipal turnover was bas A ugust through J une. It was more logical for this study to use an academic year than a traditional ca lendar year . However, w ithin the private sector, turnover may be measured in fiscal year or calendar year. When collecting data, the numeric data were scaled on a 1 to 10 scale ( discussed further in the Procedures section of this chapter ) . The percentage of SWD was measured by the state of Florida as the percentage of students at a given school who were qualified and participated in the given school s special education program for the year . The percentage of minority students was reported to the state of Florida as the percentage of students at a given school who are not Caucasian (FLDOE, 2014) . The percentage of economically disadvantaged students was calculated by the state of Florida as the percentage of students in a given school who qualify for the free or reduced lunch program , a nd was collected Data Warehouse program. Principal turnover, the percentage of SWD, the percentage of minority students, and the percentage of economically disadvantaged students for ea ch school we re the predictors for this study. School type was defined in the dat (Appendix D) . F lorida C omprehensive A ssessment T est According to the Florida Department of Educatio website, the FCAT began in 1998 as part of Florida's overall plan to increase student achievement by implementing higher standards. FCAT is a criterion referenced assessment measuring progress in students (grades 3 11), in the areas of reading, math, science, and writing (FLDOE,

PAGE 56

56 2013). The FCAT also has a norm referenced component in math and reading, which The FCAT measures student progress in meeting the Florida Sunshine Sta te Standards (SSS) . According to the FLDOE During the 2010 2011 school year, Florida began the transition from the FCAT to the FCAT 2.0 and Florida End of Course (EOC) Assessments. (highest). For the purpose of this study, percentage of proficiency was used that is, FCAT scores 3 and above (FLDOE, 2013) . A dequate Y early P rogress According to the The No Child Left Behind legislation requires all states to report student achievement based on results of reading, mathematics, and writing statewide assessments and, also, high school graduation rates for all schools, districts, and the State. The AYP Report informs the public of student s groups, economically disadvantaged groups, students with disabilities, and students . Each subgroup must reach proficiency for their school to ma ke AYP . Procedures The Florida Department of Education publishes public access data, which was gathered for the purpose of this study. Therefore, c onsent for participation was not needed since all data collected is public information , and archived by the Fl orida publ ic school system . This study was performed af ter gaining IRB approval to collect data . After collection , all per tinent data were organized in a Microsoft Exc el spreadsheet. Column 1 was school number (numbers 1 through 47). Column 2 was

PAGE 57

57 school name ( lat er concealed to protect id entity). C olumn 3 was the 2013 reading score. Column 4 was the 2013 math score. C olumn 5 was the principal turnover rate. C olumn 6 was the perce ntage of SWD. Column 7 was the perce ntage of minority students. Column 8 was the percentage of econ omically disadvantaged students . Column 9 was the sc r middle, The schools were arranged in Microsoft Excel in no particular order, and the names of the schools were chan ged to numbers rather than actual school names , to protect school identity (Appendix C ). Data were imported into R 3.0.2 (R Core Team, 2013) for analysis, where the mean and standard deviations were calculated for each variable. A Poisson model regression was run, at which time various predictors were flag ged for multicollinearity because of the high correlations between predictors and big numbers for generalized variance inflation factors. A variance inflation factor (vif) function from ( Fox & Weisberg, 2011 ) package was used to estimate the generalized variance inflation factor, testing the sev erity of the multicollinearity . According to Martz Multicollinearity refers to predictors that are correlated with other predictors (para. 3) . It occurs when the model includes multiples factors that are correlated not just to the response variable, but also to each other. In other words, it results when factors are a bit redundant. For example, one could think of multicollinearity in a foot ball game (Martz, 2013) . It is easy to give credit for a sack when one opponent has tackled the quarter back. However, if three players tackle the opponent at the same time, it is much more difficult to determine who should get credit, and how much each one contributed to the sack ( para. 4 ) . Martz further states, Multicollinearity increases the standard errors of the coefficients para. 3 ) . Overinflating standard errors make some

PAGE 58

58 variables statistically insignificant when they should be significant. Without multicollinearity (and with lower standard errors) , those coefficients might be significant. A dispersion test function from the ( Kleiber & Zeilus, 2008 ) package was used to estimate statistic for over dis persion. T here was a problem with under dispersion. Four Quasi Poisson regression models were then run to compensate for the under dispersion concern (R Core Team, 2013). The Quasi Poisson regression model assumes variance is equal to the mean , and this was found to correct for the under dispersion. The Poisson model was determined to be an unrealistically conservative model , and the Quasi Poisson model w as deemed a better fit for this study. Data Sources T he Florida Public Education Report Card and the annual Adequate Yearly Progre ss R eport are compiled through data from each school district . Electronic surveys collect data on student achievement and various other areas including system staffs, financial records, student information, and Full Academic Year (FAY) students (FLDOE, 2 013) . Once data were collected, reports are sent to each school to ensure data qual ity and validity . School system personnel were responsible for making any needed changes to their data. After corrections were made, final reports were submi tted again. The system report cards are then released to the public via the Florida Office of Student Achievement (FOSA) , and as AYP reports on the Florida Depar tment of Education website. These two public websites, as well as the local district Data Warehouse, provide d the data sources for this study in the areas of math a nd reading student

PAGE 59

59 achievement, percentage of minority students, percentage of economically disadvantaged students, and percentage of SWD. Access to the Data T Data Warehouse program and the Florida Department of Education website were used to c ollect data for each school in this study . A Microsoft Excel file was created listing eac h school in the first column . Then , eight more columns were created: reading score, math scor e, p rincipal turnover rate, percentage of SWD, percentage o f minority students, and percentage of low SES students (lunch status) , years school has been in existence, and coded school type . T hese data were entered in the row corresponding to the particip ating schools . The end result was one ata listed in one location ( Ap pendix C ). All data were publicly accessible for viewing on each website. Email correspondence and ind ividual principal meetings were used to co llect the principal turnover data ; however , a r equest to send a mass email to all participating school principals was denied. After this, the princ ipal turnover data was requested through the district Supervisor of Communications and Communit y Relations. A request was granted for access to public records for this data . All data collected and entered into Microsoft Excel were checked twice with AYP reports for accuracy. Principal turnover data was entered as soon as the data were received , to ensure that data were entered correctly. Demographic Profiles Demographic profiles were created , and stored in a Micro soft Excel database. Principal turnover rate was then calculated for each school. This data was collected from the previous 10 years and given a ratin g : one represented a school with only one

PAGE 60

60 principal over the past 10 years , and ten represented a school with ten differ ent principals over the past 10 years. The number of princip als per school over the past 10 years was the principal turnover rate for th is study. The math and reading proficiency percentages were also stored for the 2013 school year. Lastly, AYP subgroup data were stored for 2013 in the areas of percentage of SWD, percentage of economically disadvantaged students, and percentage of minorit y students. Reading and Math Achievement The Adequate Yearly Progress scores calculated from the FCAT in math and FCAT reading were published , and provided to each school district by the Florida Department of Education in the annual AYP report; therefore, no extra permission s w ere needed to use this information (Appendix E ). Student names were not used , since the data collected represented the average scores of the entire school, with only prin cipal succession frequency was compiled into a Microsoft Excel spreadsheet. Principal turnover data were collected over a 10 year period, while FCAT data were only collected for the 2012 2013 school year. Expan ding the study to include a 3 year trend in FCAT scores may provide dif ferent outcomes. It is possible that student achievement drops during the first year of principal turnover , and increases as the new prin cipal establishes leadership. This study only examined principal turnover rates over a 10 year period. Data Analysis Q uantitative data were sorted , and stored in a Microsoft Excel spreadsheet . The Poisson regression model was used to test significance, which is a general linear model (G LM). A n open sour ce statistical analysis program (R3.0.2) was used to run the Poisson regression analysis (R Core Team, 2013). This procedure was used to

PAGE 61

61 determine differences in student success in reading and math on the 2013 FCAT in schools where prin cipal tenure and turnover vary . T he number of principal changes for each school over the past 10 years was tested against student achievement on the 2013 FCAT test. A Poisson regression model was initially used, offsetting for the number of years a schoo l has been up and running in the last 10 years . This determined any significant relationshi p between principal turnover rate, percentage of minority students, percentage of SWD, and percentage of economically disadvantaged students, student achievement, and sc hool typ e. Next, was estimated . After observing high correlations among the predictors (specifically Reading score, Math score, Minority rate and SES rate as measured by Free and Reduced Lunch) , there was concern over multicollinearity (See Table 3 1). With m ulticollinearity in mind, four different Qua si Poisson regression models were run (Reading, Math, Minority rate and SES rate) . The Poisson model was determined to be an unrealistically conservative model and the Quasi Poisson model was deemed a better fit for this study. Poisson and Quasi Poisson Re gression A Poisson regression analysis was conducted to evaluate the null hypothes is that the predictor variables ( percentage of minority students, percentage of SWD, percentage of economic ally disadvantaged students, reading FCAT scores, math FCAT scores or school type ) do not significantly predict principal turnover rate in this school district . Poisson regression is a generalized linear model (GLM) for count data. The data collected for this study was count data (ie. percent minority students, percent l ow SES students, etc.). Weisberg (2005) states, The simplest GLMs for count data assume a Poisson distribution for the random component ( p.379 ) . Weisberg (2005) Like counts, Poisson variates can tak e any nonnegative integer value

PAGE 62

62 ( p.379 ) . T he Poisson distribution is unimodal and skewed to the right ov er the possible values 0,1,2,.. That is, The Poisson distribu tion has a positive mean. The GLMs for the Poisson mean can use the identity link, but it is more common to model take any real number value. According to Weisberg (200 5), A Poisson loglinear model is a GLM that assumes a Poisson distribution for Y and uses the log link function p. 379 ) . For a single explanatory variable x, the Poisson loglinear model has the form In other words, Weisberg (2005 ) states, the typical Poisson regression model expresses the log outcome rate as a linear function of a set of predictors (p. 391). In this study, the linear function was not as visible because several of the predictors were highly correlated. These include the FCAT reading scores, FCAT math scores, minority rate and SES rate, as determined by Free and Reduced Lunch. All of these four variables were found to be statistically associated with the outcome (rate) but since they are highly correlated, they could not give an accurate finding. Therefore, this needed to be corrected . After this step, and accomm odating the small sample size, four Quasi Poisson regressions were determined to be the m ost appropriate analyses for th is study. Regression 1 used . The reading model formula:

PAGE 63

63 glm(formula = Prin.Trnovr ~ Reading.Score + Disab..Rt + school.type, family = quasipoisson, data = data, offset = log(years)) the independent variable. The math formula: glm(formula = Prin.Trnovr ~ Math.Score + Disab..Rt + school.type, family = quasipoisson, data = data, offset = log(years)) . The minority rate f ormula: glm(formula = Prin.Trnovr ~ Disab..Rt + Minority.Rt + school.type, family = quasipoisson, data = data, offset = log(years)) Regression 4 used as the independent variable. The SES formula: glm(formula = Prin.Trnovr ~ D isab..Rt + Lunch.Stat + school.type, family = quasipoisson, data = data, offset = log(years)) Specific school data can be viewed in Appendix C. The Quasi Poisson regression model assumes variance is equal to the mean , and t his was found to correct f or under dispersion. Poisson and Quasi Poisson Regression Assumptions Poisson regression assumes all variables follow a bell shaped curve (Fox & Weisberg, 2011) . Many mental test scores such as the FCAT are known to follow a normal distribution (FLDOE, 2 013) . It is assumed that the relationship between the independent and dependent variables is linear (Fox & Weisber g , 2011) . High sample size makes Poisson regression robust against violations of this assumption (Fox & Weisberg, 2011) .

PAGE 64

64 This study assumed that variables were measured reliably and without error. The FCAT is a valid and relia ble assessment instrument (FLDOE, 2013) . V a lidity and reliability data were provided by the FLDOE Testing and Assessment Department. The multicollinearity assumption w as addressed by creating a correlation matrix to determine how each variable correlated with the others (Fox & Weisberg, 2011) . Multicollinearity assumes the variables are not extremely correlated with one another at the .7 or higher r value (R Core Team, 2013) . A generalized variance inflation factor (GVIF) function from the ( Fox & Weisberg, 2011 ) package was used to estimate the severity of multicollinearity. Lastly, the few extreme values were not excluded from the data since they were not deemed to be statistical outliers. According to Tabachnick and Fidell (2007), when the sample size is small, including or excluding specific data points that are not clearly outliers , may significantly influence the regression line a nd the correlation coef ficient. This is why the outliers were included in the data analysis. Summary Because of the No Child Left Behind accountability measures, school districts are searching for every advantage to maximize student knowledge and performance. Understanding how principal turnover affects student achievement could provide many districts with explicit knowledge to help make informed decisions that will benef it their students . This study identified school and student characteristics that lend th emselves to frequent principal turnover . Using the FCAT resul ts as calculated for AYP use s data across all this selected Florida c ounty p ubli c s chools, providing a district wide picture of academic success. Using the student and school characteristics, r esults of this study can help schools determine a re alistic timeline for

PAGE 65

65 improving academic achievement once a new principal is hired , and identify predictors of high turnover schools. Results of this study can also help schools improve their current succ ession management plans, and hiring procedures.

PAGE 66

66 Table 3 1. GVIF: Predictors Flagged for Multicollinearity GVIF Df GVIF^(1/(2*Df)) Reading.Score 7.639576 * 1 2.763978 Math.Score 5.146682 * 1 2.268630 Disab . Rt 1.441418 1 1.200591 Minority.Rt 5.828898 * 1 2.414311 Lunch.Stat 10.413148 * 1 3.226941 S chool.type 1.531606 2 1.112466 *Multicollinearity is the undesirable situation where the correlations among the independent v ariables are strong. In other words, multicollinearity misleadingly inflates the standard errors. Thus, it makes some variables statistically insignificant while they should be otherwise significant. For example, this table shows that reading score, math s core, minority rate , and lunch stat were flagged for multicollinearity . Therefore, four models were run.

PAGE 67

67 CHAPTER 4 FINDINGS The primary purpose of this study was to determine if there is a relationship between principal turnover rate , and school and st udent characteristics on selected Florida p ublic s chools. Specifically, the school and s tudent characteristics examined were percentage of students with disabilities , perc entage of minority students, percentage of low socioeconomic students , scho ol type, and student achievement as determined by FCAT reading and math scores. Student achievement data on the 2013 math and reading FCAT was gathered for the 47 participating schools from the annual AYP reports publicly viewable on the Florida De partment of Education website. Both t he AYP reports and the local school websites provided the percentages of economically disadvantaged, minority, and students with disabilities for each school. Principal turnover rates were collected through the district offic e and Data Warehouse. Assumptions This study utilized correlation and Quasi Poisson regression analysis to determine if a significant relationship existed between principal turnover rate, percentage of minority students, percentage of students with disabil ities , school type, percentage of economically disadvantaged studen ts, and FCAT reading and math scores. Several assumptions had to be met before Quasi Poisson regression analysis coul d be conducted. This study began with a Poisson regression, and ended u p finding a Quasi Poisson regression was a mo re appropriate fit due to multi collinearity. O ne assumption of this study was that the variables were measured reliably and were free from error. As stated in c hapter 3, the FLDOE provide d valid and reliable da ta through the math and reading FCAT . As data was entered into Microsoft Excel, it was

PAGE 68

68 cross checked for consistency with the AYP reports. The principal turnover data was verified through the district office and entered into the spreadsheet . Some of the v ariables in this study were highly correlated , and had a generalized variance inflation factor (GVIF) greater than ten. For instance, SES data had a GVIF of 0.859, and was flagged for multicollinearity. Anothe r assumption w as over and under dispersion. It is common , within Poisson regressio n analysis, to run into an over dispersion concern (R Core Team, 2013) . For a correctly specified model, the Pearson chi square statistic and the deviance, divided by their degrees of freedom, should be close to or equal to one (Cameron & Trivedi, 1990) . When their values are much larger than one, the assumption of binomial variability may not be valid , and th e data are said to exhibit over dispersion ( Cameron & Trivedi, 1990 ) . Under d ispersion, ratios less than one , occurs less often in practice (Cameron & Trivedi, 1990). The Pearson chi square for this study indicated that over di spersion was not a problem ( see Table 4 1). The result of the over dispersion test was that the over dispersion parameter was 0.341, whic h (because it is less than 1) indicates over dispersion is not a problem. Under dispersion was a concern, so Quasi Poisson regression was run to cor rect for this . Table 4 1 . The CORR Procedure Reading score Math score Disab. Rate Minority Rate Low SES Reading score 1 Math score 0.85868 1 <.0001 Disab. Rate 0.24303 0.33252 1 0.0998 0.0224 0.408 Minority Rate 0.77467 0.79945 0.12355 1 <.0001 <.0001 0.408 Low SES 0.87869 0.8249 0.27014 0.89504 1

PAGE 69

69 <.0001 <.0001 0.0663 <. 0001 Hypothesis Testing Results Pearson correlations were performed for each predictor , and principal turnover rate were performed. Four variables were flagged for high corre lation, which indicated a multi collinearity issue: reading score, math score, mi nority rate , and low SES . The Low SES was the variable most commonly correlated with the other variables. P ercentage of SWD had a weak negat ive relationship with principal turnover rate. Four individual models were run using Quasi Poisson regression mod els. Reading Model FCAT reading scores were run with 4 predictors. F indings showed that the interc ept factor was 0.14716 ( see Table 4 2 ). When FCAT reading data were exponentiat ed, the intercept rate was 0.863 with the p value of .0102. This indicates significant statistical difference . A n increase in reading score is associated with a decrease in principal turnover because the coefficient is negative. The number indic ates that the odds of principal turnover , are multiplied by 0.863 for schools that only differ by 1 standard deviation for reading FCAT scores. Table 4 2 . Reading Model Reading Model Estimate Std. Error t value Pr(>ltl) Intercept 1.10166 0.07171 15.362 <2e 16 Reading Score 0.014716 * 0.05468 2.691 0.0102 Disability Rate 0 .02225 0.05653 0.394 0.6969 School Type 2 0.10594 0.13757 0.77 0.4456 School Type 3 0.24559 0.13271 1.851 0.0713 Math Model The FCAT math scores were run with the Quasi Poisson regression model. The intercept f inding from this model was 0.17119 . Wh en math data were exponentiated,

PAGE 70

70 the intercept rate was 0.843 with the p v alue of 0.0044. This indicates a significant statistical difference . An increase in math score is associated with a decrease in principal turnover rate because the coefficient is n egative. This indicates that the odds of principal turnover are multiplied by 0.843 for schools that only differ by 1 standar d deviation for math FCAT score ( see T able 4 3 ) . Table 4 3 . Math Model Math Model Estimate Std. Error t value Pr(>ltl) (In tercept) 1.10349 0.07075 15.596 <2e 16 Math Score 0.17119 * 0.05686 0.679 0.5007 Disability Rate 0.03858 0.0568 0.679 0.5007 School Type 2 0.09963 0.13564 0.735 0.4667 School Type 3 0.25687 0.13073 1.965 0.0561 Minority Model Minority rate was run with the Quasi Poisson regression model. The intercept finding from this model was 0 .1482 ( see Table 4 4 ) . When minority data were exponenti ated, the intercept rate was 1.1597, with the p value of 0.0073. This indicates significant statistical diffe rence . An increase in minority rate is associated with an increase in principal turnover. Table 4 4 . Minority Rate Model Minority Model Estimate Std. Error t value Pr(>ltl) (Intercept) 1.11732 0.072183 15.479 <2e 16 Disability Rate 0.007376 0. 056028 0.132 0.8959 Minority Rate 0.148263 * 0.052573 2.82 0.0073 School Type 2 0.109862 0.137357 0.8 0.4283 School Type 3 0.329466 0.134279 2.454 0.0184 Low SES Model Low Quasi Poisson regr ession model. The intercept finding from this model was 0.1696 ( see Table 4 5 ) .

PAGE 71

71 When low SES data was exponentiated , the intercept value was 1.1848 , with a p value of 0.00312. This indicates a significant statistical difference . An increase in low SES p opulations is associated with an increase in principal turnover. Table 4 5 . Low SES Model Low SES Model Estimate Std. Error t value Pr(>ltl) (Intercept) 1.12414 0.07113 15.804 <2e 16 Disability Rate 0.01991 0.0554 0.359 0.72114 Low SES Ra te 0.16958 * 0.05406 3.137 0.00312 School Type 2 0.1164 0.13476 0.864 0.39263 School Type 3 0.34211 0.13161 2.599 0.01283

PAGE 72

72 CHAPTER 5 DISCUSSION Summary of Findings N ull hypotheses were tested using Pearson Correlations and the Quasi Poisson regression m odel , using R software and Microsoft Excel. This study used correlational research , and Quasi Poisson regression model to determine if there was a statistically significant relation ship between school and student characteristics and principal turnover rat e . The data showed significant statistical differences in reading and math FCAT scores, minority rate, and low SES rate. Review of the Research Question s RQ1 : Is there a statistically significant relationship between selected school and student character istics , and principal turnover? This research question led to the following sub research questions: RQ1a : Is there a statistically significant relationship between principal turnover rate , and percentage of students with disabilities? RQ1b : Is there a st atistically significant relationship between principal turnover rate , and percentage of minority students? RQ1c : Is there a statistically significant relationship between principal turnover rate , and percentage of low socio economic status (low SES) stud ents? RQ1d : Is there a statistically significant relationship between principal turnover rate , and reading FCAT scores? RQ1e : Is there a statistically significant relationship between principal turnover rate , and school type? RQ1 f : Is there a statisti cally significant relationship between principal turnover rate , and math FCAT scores? Review of the Hypothesis and Null Hypothesis H1 : Selected school and student characteristics are significant predictors of principal turnover rate. H1a : SWD is a statis tically significant predictor of principal turnover rate.

PAGE 73

73 H1b : Percentage of minority rate is a significant predictor of principal turnover rate. H1c : Percentage of low SES (economically disadvantaged ) students is a statistically significant predictor of principal turnover rate. H1d : Reading FCAT scores are statistically significant predictors of principal turnover rate. H1e : School type is a statistically significant predictor of principal turnover rate. H1 f : Math FCAT scores are statistically signifi cant predictors of principal turnover rate. Null Hypotheses H 0 1 : Selected school and student characteristics are not significant predictors of principal turnover rate. H 0 1a : SWD is not a statistically significant predictor of principal turnover rate. H 0 1 b : The percentage of minority students is not a statistically significant predictor of principal turnover rate. H 0 1c : The percentage of low SES students is not a statistically significant predictor of principal turnover rate. H 0 1d : Reading FCAT scores are not statistically significant predictors of principal turnover rate. H 0 1e : School type is not a statistically significant predictor of principal turnover rate . H 0 1f : Math FCAT scores are not statistically significant predictors of principal turnover rate . Discussion of Finding in Light of Relevant Literature The following subheadings will discuss the findings in light of relevant li terature. The findings include significant statistical differences in the reading and math FCAT models as well as the minori ty type and low SES models.

PAGE 74

74 Reading FCAT and Principal Turnover Findings This study found statistical difference between principal turnover rate and reading FCAT scores. An increase in reading FCAT scores is associated with a decrease in principal turnov er. This is consis tent with relevant literature. Loeb et al. (2010) and Papa et al. (2002 ) report that generally speaking, when principals change schools (at their own request), they tend to transfer to higher achieving schools. There is a link between low performing schools and principal turnover rate. When schools do not meet AYP , the state government intervenes in an effort to restructure the school , often replacing the principal . levels of academic achievement (Blazer, 2010). Miller (2009) studied the relationship between tests at the elementary level and end of grade tests at the middle school level) , and principal f lux in North Carolina schools. P rincipal transitions were found to be associated with changes in student achievement. the scores d uring the previous principal , transition period. Principal transitions play a significant role in reading test scores. According to findings, the null hypothesis is rejected (H 0 1d: Reading FCAT scores are not stati stically significant predictors of principal turnover rate ), and the hypothesis is accepted (H1d: Reading FCAT scores are statistically significant predictors of principal turnover rate ) . Math FCAT and Principal Turnover Findings This study found signifi cant statistical difference between math FCAT scores and principal turnover. An increase in math FCAT scores is associated with a decrease in

PAGE 75

75 principal turnover. Partlow (2007) Ohio study indicated that the only variable influencing principal turnover was the percentage of students passing the math achievement test. In other words, the principal retention rate is affected positively when students are achieving in the area of mathematics . Ogawa and Hart (1985) and Hart and Ogawa (1987) conducted stud ies on the relationship between leadership stability and student performance. Their research identifies significance between principal and superintendent tenure , and student achievement. This study confirms their findings. Thus, the null hypothesis ( H 0 1f : Math FCAT scores are not statistically significant predictors of principal turnover rate ) is rejected, and the hypothesis ( H1f : Math FCAT scores are statistically significant predictors of principal turnover rate ) is accepted. Minority Rate and Principa l Turnover Findings Significant s tatistical difference was found between principal turnover rate , and the minority rate of a school. Current research finds that minority rate is often a predictor of low achieving schools, which according to this study is also a predictor of high principal turnover ( Horng, 2009; Lankford, Loeb , & Wycock, 2002; Peske & Haycock, 2006). This study confirms that an increase in minority rate is associated with an increase in principal turnover in one selected Florida school dis trict. The null hypothesis ( H 0 1b: The percentage of minority students is not a statistically significant predictor of principal turnover rate ) is rejected. The hypothesis ( H1b: Percentage of minority rate is a significant predictor of principal turnover r ate ) is accepted. Low SES and Principal Turnover Findings Significant statistical difference was found between an increased l ow SES population and principal turnover. An increase in l ow SES populations is associated

PAGE 76

76 with an increase in principal turnove r. However, some literature (Branch et al., 2012 ) found that principal impact was nearly twice as large in high poverty schools as in low poverty schools , when using value added model scores to measure this influence . The Branch (2012) study also indicate d large , high poverty schools more so than large , low poverty schools, was more greatly impacted (Branch et al., 2012; Clark et al., 2009; Loeb et al., 2010). Likewise, similar results were found for high minority and low performing schools (Branch et al., 2012). Other research is not as encouraging. I n general, i t has been long known the more affluent a family is, the better the children of that family perform in school (Donald, 2012). poverty. If it is anchored to issu es of race, then you have to address issues of race, not just a new reading as cited in Brobeck, 2011, para.17). T he relationship between family earnings and children's test scores has typically been expected not to fluctua te over time . However, one Stanford study indicates this is not occurring (Donald, 2012). Reardon (2008) of Stanford University found an alarming increase in student test scores, between families in the 90 th percentile of income (approximately $160,000) a nd families in the 10 th percentile of income (approximately $17,500). This indicated a 40 % increase, and nearly doubled the black white achievement gap (Duke, 2008). Naturally, t here is not just one reason for the widening gap ( Brobeck, 2011 ; Duke, 2008 ; Reardon, 2011 ). Quite often , when there is a large trend occurring, the reasons are multi faceted ( Brobeck, 2011; Duke, 2008; Reardon, 2011 ), and there are no simple answers on how to fix it. Murphy ( as cited in Brobeck, 2011) suggests that a large scale societal attack is needed; not just an educational attack. He links health policy, welfare policy, social policy , and

PAGE 77

77 transportation policy to the real underlying issue of low SES schools. This study indicates that low SES schools are linked to high pri ncipal turnover, which according to current research is linked to low teacher morale, high teacher turnover, and low student achievement. Thus, the null hypothesis ( H 0 1c: The percentage of low SES students is not a statistically significant predictor of p rincipal turnover rate ) is rejected. The hypothesis ( H1c: Percentage of low SES students is a statistically significant predictor of principal turnover rate ) is accepted. Study Limitations and Recommendations for Further Research This study examined rel ationship s among principal turnover rate and percentage of minority students, percentage o f students with disabilities, percentage of low SES students , and student achievement in reading and math as measured by the Florida Comprehensive Assessment Test in one Florida county . This study was consistent with other studies of this size. Implications The primary purpose of this study was to research the relationship between school and student characteristics , and determine if they were significant predictors of principal turnover rate. The findings have implications for policymakers, principals, superintendents, district boards, and researchers. The most important finding of the study is that the relationships between some selected school characteristics do have significant implications for principal turnover. P olicymakers, principals, superintendents, district boards, and researchers would benefit from knowing school and student characteristics that lend themselves to high principal turnover . The implicati on for principal preparation programs is that colleges need to provide future administrators with the skills required to be a successful leader ,

PAGE 78

78 beginning with the first day on the job . Principals should be prepared to enter schools pal turnover risk , issues they typically would encounter. I t is more important to have a quality leader in place than to be overly concerned about prin cipal turnover . Principal preparation programs would b enefit both aspirin g principals and schools and students, to intern under a successful principal who has proven that academic success is possible even with high percentages of students with disabilities , minority students, and economically disadvantaged st udents in their schools (Darling Hammond, LaPointe, Meyerson, Orr, & Cohen, 2007) . Alternatively , in light of the No Child Left Behind Schools in Need of Improvement restructuring, it would be beneficial to observe a principal who has successfully turned a round this type of school. S uccessful principals could also be used as class speakers for principal preparation classes . Currently, there are some programs specifically for preparing principals to turnaround schools. These include programs such as the University of Virginia Turnaround Program, Harvard Institute for School Leadership: Leadership for Large scale Improvement, University of Illinois: Urban Education Leadership, and New Leaders for New Schools, based in Washington, D.C. Principa l preparation programs can prepare principals by training them on establishing a culture of high expectations, identifying opportunities for innovation, developing strategic plans, and incorporating a strong mentorship program. The implication for superi ntendents and local school boards is a need to improve the principal hiring process, and the need to retain quality principals. The number of principal changes in a school may not impact student achievement, according to this

PAGE 79

79 study . However , the literatur e indicates that having an effective principal at a school improves school culture, teacher morale, teacher retention, and ultimately, imp rove student achievement. I believe r esults of this study imply that hiring quality principals each time the position comes open is more important than how often the position is open. It also indicates the importance of having a principal in the same school for five to seven years for fully effective implementation of innovations. The superintendent s and school board s sh ould keep these implications in mind when they are hiring a principal for a position at a school with unique and challenging characteristics. They should make every attempt to provide the stability of a five to seven year commitment for that principal and the school. Limitations This study examined archival data only; therefore, any significant find ings and conclusions made in this study are restricted to similar Florida public schools. F indings and conclusions can only be applied to other schools that have similar demographic characteristics. Schools selected for this study were public schools in operation during the 2012 2013 school year , whose history of principal employment could be traced to the 2002 2003 school year via central office. Schools built after 2002 03 were not included in this study. Several schools in this Florida county were not eligible to participate in this study due to either opening after the 2002 03 school year , or not being in continuous operation for the 10 year period. Private schools and charter schools were omitted from the study. School and student characteristics , and the relationship to principal turnover was the focus of this study. There are many reasons for a principal leaving a school ;

PAGE 80

80 however, t he variety of reasons (removal, retirement, transfer, illness, death, or promotion) was not part of this study. Principals change positions for both positive and negative reasons. Positive reasons could include promotion and opportunities for higher levels of leadership. Negative reasons could include removal for being deemed ineffective by superiors, or principals leaving because of unsatisfactory conditions (Miklos, 1988). This study did not include rea sons for departure. The focus of this study was quantitative. A mixed study design implementing qualitative methods to gather data ( such as principal leadership styles and reasons for principal turnover ) could i ncrease the amount of data gathered . It cou ld offer more insight as to what causes principals to leave or stay in a position . C haracteristics of the principals could add an interesting piece to this study. Gender, age, race, ethnicity, languages spoken, leadership style, years teaching, and princ ipal tenure could all be qualitative pieces that could potentially influence turnover and performance. Principal turnove r data was collected over a 10 year period, while FCAT data was only collected for the 2012 2013 school year. Expanding t he study to re flect a longe r trend in FCAT scores may provide different outcomes. The literature indicates that there is a drop in achievement the year after a new principal arrives, and gradually increases as the principal establishes tenure at that school . This study was limited as it only looked at principal turnover rates over a 10 year period. Student achievement was measured by success on the FCAT. The FCAT was the standardized test designed specifically to assess student mastery of the Sunshine State Standards; th erefore, generalizations outside the state of Florida may not be valid.

PAGE 81

81 The literature indicates that principals are key to student achievement . However, showing evidence of this is complex, as there is not a direct relationship . According to Hanson (2003 he basic problem involves the number of intervening variables that stand between the direct cause (stability of the principal) and the effect (test p. 28) . Some of the extraneous, influential variables are instructional methods of teachers, teacher motivation, organizational structure of the district, parental involvement, availability of resources, and so forth . There was uncontrolled variance and causal inference (i.e., the amount of inference readers draw from the data is not intended to show cause and effect). However, what was expected to be a limitation the problem of multicollinearity (some predictors being too closely related to each other) Correlations and Quasi Poisson regression. Recommendations B ased on findings of this study, the following are recommendatio ns for further research. In order to provide expanded information regarding principal impact on academic achievement, r esearch is needed to compare student test scores the year before and the ye ar after a change in principal. This could offer more insight as to principal impact in this particular county. This study should be re plicated in school districts in surround ing counties of Florida, to increase the sample size and the number of low ac hieving schools studied. Many schools in surrounding regions have higher percentages of minority and low SES students , which may yield different results. This could extend the data on student characteristics and principal turnover rates .

PAGE 82

82 Expanding this stu dy to include qualitative data is highly recommended. Student, teacher, and principal insights and perceptions about principal turnover would be beneficial. This information could positively influence hiring processes, contract duration, and succession p lanning for school districts. Knowing that teacher and student morale affect culture, which in turn influences student achievement, school districts can attempt to limit the adverse effects of principal change. Qualitative research should be conducted on the extent to which NCLB, RttT, Common Core, and AYP influence principal decision making. It could be that the increases in accountability, and the ways principals are evaluated , are mainstreaming principal decisions. Principal decision maki ng may also be influenced by student characteristics, specifically socioeconomic status . It would be interesting to conduct research on this topic to see how much influence these mandates and characteristics have , and their effect , on student achievement. A dditional research is also needed that focuses on schools that do n ot meet Adequate Yearly Progress in this particular county . This could provide evidence linking variables in these schools specifically, and whether they indicate high or low principal turnover rates. Research is needed at each school level individually (elementary, middle, high) , to determine the relationship between principal turnover rate and student achievement. Students at different age levels are affected by different things . It c ould be that one level of school is more highly effected by change specifically a change in leadership. Research should be conducted to determine if there is a relationship between principal longevity and tur nover rate. Prior studies show that principal s may stay at a ipals

PAGE 83

83 nearing retirement may be less interested in changing schools, and when given the option , will remain at one school. Research is needed in areas o f specifically high migrant worker population s to see if there is a relationship between principal turnover rate and student achievement. It could be that student turnover is causing school grades to appear low. Students whose parents are migrant farmers , ( this county has an area known for migrant population ), may be affected by the constant upheaval in their lives switching schools as the season s change. We may find that principals at such schools are viewed as low performing, and thus transferred out o f the schools when the school is unable to make AYP , and must be restructured under ESEA 1116 (b). Conclusion This study addressed the question, is there a relationship between select school and student characteristics , and principal turnover in one Florid a county? Many they are in a school for less than five years, and then transferred , terminated, or burned out by the workload. The review of literature indicates that this is not ideal for school culture or student achievement. The findings of this study indicate that a decrease in FCAT reading and math scores is a predictor of principal turnover, as is an increase in minority rate and low SES population. This study adds t o the body of literature , which helps to lessen turnover. It assists in providing teachers and students with the effective school leadership they need . It does so by equipping school districts with knowledge of what characteristics of students and school s lend themselves to high turnover rates. School districts can use the findings of this study to move forward in adequately d efining the job of the principal , p roviding high quality training for aspiring school leaders , hiring

PAGE 84

84 selectively , and e valuating principals and giving them the on the job support they need ( The Wallace Foundation, 2012) . Prior to the last ten years, little research has been collected on principal turnover. The increase in accountability has caused a great stir in the edu cational leadership community. There is cause for optimism , however , as this prior neglect is being eradicated , and new research is being conducted. School leadership is being strengthened by recent findings and new experiences. Principal turnover data will lend itself to more effective hiring and firing processes, as well as better preparation for aspiring leaders. Effective leaders in every school are needed if our national educational goals are to be achieved , and if we are to remain competitive with the rest of the world, educationally .

PAGE 85

85 APPENDIX A COUNTY PRINCIPAL TURNOVER, TEN YEAR SPAN Figure A 1. Percent of schools' principal turnover rates: TEN YEAR PERIOD 1 2 3 4 5 6 7 24% of schools had 2 principals in a ten year period 24% had 3 principals 28% had 4 principals

PAGE 86

86 APPENDIX B FLORIDA COUNTY MAP By United States Census Bureau ([1]) [ Public domain], via Wikimedia Common

PAGE 87

87 APPENDIX C SCHOOL CHARACTERISTICS: STUDENT ACHIVEMENT Figure C 1. Elementary school FCAT Reading, Math Scores and number of principals. Figure C 2. Middle school Reading, Math Scores and number of prin cipals.

PAGE 88

88 Figure C 3. High schools Reading, Math Scores and number of principals.

PAGE 89

89 APPENDIX D DATA FILE

PAGE 90

90

PAGE 91

91 APPENDIX E RESEARCH PERMISSIONS

PAGE 92

92 LIST OF REFERENCES Abelson, M. , & Baysinger, B. (1984). Optimal and dysfuncti onal turnover: T oward an organizational level model. The Academy of Management Review , 9 , 331 341. Anthes, K. (2002). School and district le adership. No Child Left Behind Policy B rief (Report No. ECS GP 02 02). Retrieved from Education Commission of the S tates, Denver, C O , website: http://www.ecs.org/clearinghouse/34/62/3462.pdf. Ary, D., Jacobs, L. C., Razavieh, A. & Sorens en, C. (2006). Introduction to r esearch in e ducation ( 7 th ed.). Belmont, CA: Thomson Wadsworth. Baker, B., & Cooper, B. (2005). Do p rincipals with stronger academic backgrounds hire better teachers? Policy implications for improving high poverty schools. Educational Administration Quarterly , 41 (3), 413 448. Baker, B., Orr, M. , & Young, M. (2005). Academic d rift, i nstitutional p roducti on, and p rofessional d istribution of g raduate d egrees in e ducational l eadership . Educational Administration Quarterly . 43 (3), 279 318. Bass, B. , & Stodgill, R. (1990). h andbook of l eadership : Theory, resear ch, and managerial applicati ons ( 3 rd ed.) . The Free Press. Retrieved from: http://tocs.ulb.tu darmstadt.de/22997466.pdf Berrong, D. (2012). The r elationship b etween p rincipal t urnover and s tudent a chievement in r eading/ e nglish l anguage a rts and m ath g rades s ix t hrough e ight . Libert y University, Doctoral Dissert ations and Projects. Retrieved from: http://digitalcommons.liberty.edu/doctoral/520 Beteille, T., Kalogrides, D., & Loeb, S. (2012). Stepping stones: Principal career paths and school outcomes. Social Science Research , 41 (4), 904 919. See more at: http://cepa.stanford.edu/content/stepping stones principal career paths and school outcomes#sthash.HbO3qZaU.dpuf Blazer, C. (2010). Principal turnover. Information capsule . Volume 0914. Research Services, Miami Dade County Public Sc hools. Bouchard, E. D., Cercone, L., Hayden, H., Riggens Newby , C.G., & Zarlengo , P . (2002). Acknowledgements. Chronicles: A h istory of the d evelopment of the p l eadership n etwork. Addressing p rincipal l eadership c hallenges . Brown University . Retrieved from: http://www.brown.edu/academics/education alliance/sites/brown.edu.academics.education alliance/files/publications/chrncl_pln.pdf

PAGE 93

93 Bower, J. (2007). Solve the s uccession c risis by g rowing i nside o utside l eaders . Harvard Business Review. R etrieved from: https://hbr.org/2007/11/solve the succession crisis by growing inside outside leaders/ar/1 Boyd, D., Grossman, P., Lankford, H., Loeb, S., & Wyckoff, J. (200 8). Measuring effect sizes: The effect of measurement error . Working Paper prepared for the National Conference on Value Added Modeling . University of Wisconsin Madison April 22 24, 2008. Branch, G. F., Hanushek, E. A. , & Rivkin, S. G. (2012). Estimating the effect of leaders on public sector productivity: The case of school principals . Center for Analysis of Longitudinal Data in Education Research. Retrieved from : http://www.caldercenter.org/upload/Hanushek_Estimating the Effect of Leaders.pdf . See more at: http://www.centerforpubliceducation.org/Main Menu/Staffingstudents/The Principa l Perspective at a glance/The Principal Perspective References.html#sthash.oifnST3S.dpuf Brewer, D. J. (1993). Principals and s tudent o utcomes: Evidence from U.S. h igh s chools. Economics of Education Review , 12 (4), 281 292 . Brobeck, K. (2011). Turning ov er turnaround: A conversation with Joseph Murphy . Ideas in Action. Vanderbilt University Publications. October 28, 2011. Retrieved from: http://vanderbilt.edu/magazines/ideas in action/2011/10/turning over turnaround/ Brown, M.C. (1982). Administrative succession and organizational performance the succession effect. Administrative Science Quarterly, 27 , 1 16. Bruggink, P. (2001). Principal succession and school improvement: The relationship between the frequency of principal turnover in Florida public schools from 1990 91 and school performance indicators in 1998 99 . Unpublished doctoral dissertation, Florida State University. Bryk, A. S., Sebring, P. B., Allensworth, E., Luppescu, S., & Easton, J. Q. (2010). Organizing schools for improvement: Lesson s from Chicago . Chicago , IL : University of Chicago Press. Burke, L. M. (2013). . National Review Online. Retrieved from: http://www.nationalreview.com/articles/344897/why there s backlash against common core lin dsey m burke Burkhauser, S., Gates, S., Hamilton, L., & Ikemoto, G. (2012). First y ear p rincipals in u rban s chool d istricts: How actions and working conditions relate to outcomes. RAND Corporation . Retrieved from: http://www.rand.org/content/dam/rand/pu bs/technical_reports/2012/RAND_TR11 91.pdf

PAGE 94

94 Burns, J. M. (1978). Leadership . New York : Harper & Row Publishers , Inc . Calkins, A., Guenther, W., Belfiore, G., & Lash, D. (2007). The turnaround challenge: ve student achievement lies in our worst performing schools . Mass Insight Education & Research Institute . Cameron, A. C. , & Trivedi, P. K. (1990). Regression based t ests for o verdispersion in the Poisson m odel. Journal of Econometrics , 46 , 347 364. Cano le, M. , & Young, M. (2013). Standards for educational leaders: An analysis . Council of Chief State School Officers , Washington, DC . Retrieved May 31, 2014 http://www.ccsso.org/Documents/Analysis%20of%20Leadership%20Standards Final 070913 RGB.pdf Cash, J. (2008). Fastball leadership. Leadership , 37 (3), 22 25. Charan, R. (2005, February). Ending the CEO succession crisis. Harvard Business Review . Retrieved from: http://hbr.org/2005/02/ending the ceo succession crisis/ar/1 Chard, D., Vaughn, S., & Tyler, B. (2002). A s ynthesis of r esearch on effective interventions for building reading fluency with elementary students with l earning d isabilities . J ournal of Learning Disabilities, 35 (5), 386 406. http://www.wce.wwu.edu/depts/sped/forms/kens%20readings/ins truction/instruct %20effective%20interventions%20for%20building%20reading%20fluency%20ch ard%202002.pdf Clark, D., Martorell, P. , & Rockoff, J. (2009). School p rincipals and s chool p erformance. National Center for Analysis of Longitudinal Data in Education Research . Retrieved from http://www.caldercenter.org/upload/Working Paper 38_FINAL. pdf Clotfelter, C., Ladd, H., Vigdor, J., & Wheeler, J. (2006). High poverty schools and distribution of teachers and principals . P aper presented at the UNC Conference on High Poverty Schooli ng in America, Chapel Hill, NC. Coelli, M., & Green, D. A. (2012). Leadership effects: School principals and student outcomes. Economics of Education Review, 31 (1), 92 109. Cohen, J., Cohen, P., West, S. G., & Aiken, L. S. (2003). App lied multiple regression/correlation analysis fo r the behavioral sciences ( 3 rd e d.). Mahwah, NJ: Lawrence Erlbaum Associates. Cohen, J., McCabe, L., Michelli, N. M., & Pickeral, T. (2009). School climate: Research, policy, practice, and teacher education. The Teachers College Record, 111 (1), 180 213.

PAGE 95

95 Collier County Public Schools. (2012). Collier leadership evaluation model: Observation and evaluation forms and procedures for leadership practice . Retrieved from: http://www.collierschools.com/hr/docs/CLEM. pdf Collins, J. , & Porras, J. (1994). Built to last: Successful habits of visionary c ompanies . New York: HarperCollins. Collins, J. (1995 a ). Change is good but first, know what should never change. Fortune . Retrieved from: http://www.jimcollins.com/art icle_topics/articles/change is good.html Collins, J. (1995 b ). Building companies to last. Inc. special issue: The s tate of small b usiness. Retrieved from: http://www.jimcollins.com/article_topics/articles/building companies.html Collins, J. (2001). Good to Great . New York: HarperCollins. Collins, J. (2007). Level 5 leadership. The Jossey Bass reader on educational leadership , 2 , 27 50. Conger, J. A. & Fulmer, R. M. (2003). Deve loping your leadership pipeline. Harva rd Business R eview , 81 , 76 84. Corti ella, C. (2007a). No child left b ehind and students with disabilities. The Exceptional Parent , 37 (9), 70 73. Cortiella, C. (2007b). Rewards and roadblocks: How special education students are faring under no child left behind . The Center for Development & Learning, Metairie, LA. Retrieved from http://www.cdl.org/resourcelibrary/pdf/ncldrewardsandroadblocks.pdf Council of Chief State School Officers [CCSSO] . (1996). Interstate school leaders licensure consortium: Standards for school leaders . Washington, DC : Author. Council of Chief State School Officers [CCSSO] . (2008a). Educational leadership policy standards : ISLLC 2008. Washington, DC: Author. Retrieved May 13, 2014, from http://www.ccsso.org/Documents/2008/Educational_Leadership_Policy_Standard s_2008 .pdf. Council of Chief State School Officers [CCSSO] . (2008b). Performance expectations and indicators for education leaders . (N. M. Sanders & K. M. Kearney, Eds.). Washington, DC: Author.

PAGE 96

96 Crary, D. (2005, July). chool . NBC News. Retrieved from: http://www.nbcnews.com/id/8630313/ns/us_news education/t/migrant workers children savor summer school/#.VF5WavnF So Darling Hammond, L., LaPointe, M., Meyerson, D., Orr. M. T., & Cohen, C. (2007). Preparing school leaders for a changing world: Lessons from exemplary leadership development programs . Stanford, CA: Stanford University, Stanford Educational Leadership Institute. Denis, D. & Denis, D. K. (1995). Performance changes following top management dismissals. The Jour nal of Finance , 50 , 1029 1057. Developing Leadership Capacity through Leadership Behaviors , Principal Planning Guide, SIRC . Dhuey, E. , & Smith , J. ( 2011 ). How school principals influence student learning . Evanston, IL: Society for Research on Educational Effectiveness . Dillon, S. (2011 , February 8 ). U . S . plan to replace principals hits snag: Who will step i n ? The New York Times . New York. Retrieved from www. nytimes .com/2011/02/08/.../08education.html Donald, B. (2012 , February ). Stanford study finds wi dening gap between r ich and poor s tudents : Analysis of standardized tests over a 50 year period shows the achievement gap between rich and poor has grown steadily. Stanford Report , Retrieved May 31, 2014 from: http://news.stanford.edu/news/2012/february/r eardon achievement gap 021312.html Drucker, P. ( 1954 ). The p ractice of m anagement . New York: Harper & Brothers . Drucker, P. (1967). The e ffective e xecutive . New York: Harper and Row. Drucker, P. (2002). The effective executive: The definitive guide t o getting the right things done . New York: Harper Collins. Duke, D. (2005). Keys to sustaining successful school turnaround . Darden Curry Partnership for Leaders in Education. University of Virginia. Duke, D. (2008). Diagnosing school decline. Phi Delta Kappan , 89 (9), 667 671. Duke, D., Carr, M., & Sterrett, W. (2012). The school i mprove ment p lannin g handbook: Getting f ocused for turnaround and t ransition . R&L Education .

PAGE 97

97 Dyson, L. (2010). Unanticipated effects of children with learning disabilit ies on their families. Learning Disability Quarterly, 33 (1), 43 55. Eberts, R. , & Stone, J. (1988). Student achievement in public schools: Do principals make a difference? Economics of Education Review . 7 (3) , 291 299. Education Sector. (2011). A portrait of school improvement grantees . Washington, DC : American Institutes for Research. Retrieved from http://www.educationsector.org/publications/portrait school im provement grantees Education World. (2013). Meeting the educational needs of migrant students . Retrieved from: http://www.educationworld.com/a_curr/curr347.shtml Eilers, L . (2012). Essential l eadership e lements in i mplementing c ommon c ore s tate s tanda rds . Delta Kappa Gamma Bulletin . Retrieved from: http://www.readperiodicals.com/201207/2724762271.html#ixzz33PkTmwcp Elfers, A., Plecki, M., & Knapp, M. (2006). Teacher mobility: Looking more closely at Peabody Journal of Education , 81 (3 ). Florida Counties Map. Retrieved April 7, 2014 from: http://commons.wikimedia.org/wiki/File%3AFlorida_counties_map.png Florida Department of Education [FLDOE]. ( 2014 ) . www.fldoe.org Florida Senate Bill Summary 0736. Retrieved from: http://www.flsenate.gov/Committees/BillSummaries/2011/html/0736ED Fox, J. , & Weisberg, S. (2011). An {R} c ompanion to a pplied r egression , (2 nd ed.) . Thousand Oaks , CA: Sage Publications . http://socserv.socsci.mcmaster.ca/jfox/Books/Companion Fullan, M. (1999). Change forces the sequel . London: Falmer. Fullan, M. (2001). The new meaning of educational change (3rd ed.). New York: Teachers College Press. Fuller, E. (2012, July 16). Examining principal turnover . [blog article]. Retrieved from http://shank erblog.org/?p=6196 Fuller, E., Young, M. D., & Orr, T. (2007). Career pathways of principals in Texas . Paper presented at the Annual Meeting of the American Educational Research Association, Chicago, IL.

PAGE 98

98 Gates, S., Ringel, J., Santibanez, L., Guarino, C . , Ghosh Dastidar, B. , & Brown, A. (2005). Mobility and turnover among school principals. Economics of Education Review , 25 , 289 302. Gates, S., Ringel, J., Santibanez, L., Ross, K. , & Chung, C. (2003). Who is leading our schools? An overview of school a dministrators and their c areers . RAND Santa Monica. Gill, J. (2013). Make room for the principal supervisors . New York: The Wallace Foundation . Gilson, T. (2008). Educational leadership: Are we busy yet? American Secondary Education, 36 (3), 84 97 . Gree n, P. (2003). The undocumented: Educating the children of migrant w orkers in America. Bilingual Research Journal , 27 (1) , 51 71. Grusky, O. (1960). Administrative succession in formal organizations. Social Forces , 39 , 105 115. Hallinger, P. , & Heck, R. effectiveness: A review of empirical research, 1980 1995. Educational Administration Quarterly 32 , 5 44. Hallinger, P. , effectiveness: 1 980 1995. School Effectiveness and School I mprovement , 9 , 157 191. Hanson, E. (2003). Educational administration and organizational behavior ( 5 th ed.). New York: Allyn & Bacon. Hart, A. , & Ogawa, R. (1987). The influence of superintendents on the academ ic achievement of school districts. The Journal of Educational Administration , 25 (1), 72 84. Heck, R. H. , & Hallinger, P. (1999). Next generation methods for the study of leadership and school improvement. In J. Murphy & K. S. Louis (Eds.), Handbook of re search on educational administration (2 nd ed.) pp. 141 162 . San Francisco: Jossey Bass. Heitin, L. (2013). How does the U.S. stack up a gainst other n ations? Education Week: Teacher . Retrieved from http://www.edweek.org/ew/section/multimedia/pisa_scores_c omparison.html

PAGE 99

99 Herzog, L. (2014). Common core critics speak out at forum in Ridgewood . Retrieved from http://www.northjersey.com/news/education/common core critics speak out at forum in ridgewood 1.1025976#sthash.0umsBY4F.dpuf Hill, S. (2008). Where wil eaders? Harvard Business Review , 86 (1), 123 12 8 . Retrieved from http://europepmc.org/abstract/med/18271324 Horng, E. (2009). Teacher tradeoffs: D conditions and student demographics. Ame rican Educational Research , 46 , 690 717. Hull, J. ( 2012). The principal perspective: A full report . The Center for Public Education . Retrieved from: http://www.centerforpubliceducation.org/principal perspective ISLLC S tandards . (2014). School Leadershi p Briefing. School Administration Publishing, LLC. Retrieved from http://www.schoolbriefing.com/isllc standards/ Johnson, E., Humphrey, M., Mellard, D., Woods, K., & Swanson, H. (2010). Cognitive processing deficits and students with specific learning dis abilities: A selective meta analysis of the literature. Learning Disability Q uarterly , 33 (1), 3 18. Kannapel, P. J., & Clements, S. K., with Taylor, D., & Hibpshman, T. (2005). Inside the black box of high performing high poverty schools . Lexington, KY: P richard Committee for Academic Excellence. Krathwohl, D. (1998). Methods of educational and social science research: An integrated approach ( 2 nd ed.). New York: Longman. Kleiber, C. , & Zeileis, A. (2008). Applied e conometrics with R. New York: Springer Verlag. ISBN 978 0 387 77316 2. URL http://CRAN.R project.org/package=AER Lankford, H., Loeb, S., & Wycko , J. (2002). Teacher s orting and the plight of urban schools: A descriptive a n alysis. Educational Evaluation and Policy Analysis , 24 ( 1 ) 37 62 . La sh way, L. (2003). Distributed l eadership. Research Roundup , 19 ( 4 ) . Retrieved from: http://eric.ed.gov/?id=ED477356 Leithwood, K. , & Riehl, C. (2003). What w e k now a bout s uccessful s chool l eadership . A report by Division A of AERA. National College for Sc hool Leadership. Leithwood, K., Seashore Louis, K., Anderson , S., & Wahlstrom, K. (2004). Ho w l eadership i nfluences s tudent l earning . Minneapolis, MN: Learning from Leadership Proje ct .

PAGE 100

100 Leithwood, K., Louis, K. S., Wahlstrom, K., Anderson, S., Mascall, B., & Gordon, M. (2009). How successful leadership influences student learning: The second installment of a longer story. In Second international handbook of educational change (pp. 611 629). Springer Netherlands. Leithwood, K., & Mascall, B. (2008). Coll ective leadership effects on student achievement . Educational Administrative Quarterly , 44 (4), 529 561. Levine, A. ( 2005). Educating school leaders . The Education Schools Project . Retrieved from http://www.e dschools.org/pdf/Final313.pdf Levy, A., Fields , E., & Jablonski, E. (2006, October). Overview of research: What we turnover . In NCTAF Symposium on the Scope and Consequences of K 12 Science and Mathematics Teacher Turnover. Loeb, S., Kalogrides, D., & Béteille, T. (2012). Effective schools: T eacher hiring, assignment, development, and retention . Education Finance and Policy , 7 , in press. Loeb, S., Kalogrides, D., & Horng, E. (2010). Principal preferences and the unequal distribu tion of principals across schools. Education Evaluation and Policy Analysis , 32 , 205 229. Loeb, S., & Reininger, M. (2004). Public policy and teacher labor markets: What we know and why it matters . East Lansing, MI: The Education Policy Center at Michiga n State University. Martz, E. (2013). Enough is e nough! Handling m ulticollinearity in r egression a nalysis. The Minitab Blog . Retrieved from: http://blog.minitab.com/blog/understanding statistics/handling multicollinearity in regression analysis Marzan o, R. J. (2003). What works in schools . Alexandria, VA: Association for Supervision and Curriculum Development. Marzano, R. J. (2013). Marzano school leadership evaluation model . Washington State Teacher/Principal Eval uation Project. Retrieved from htt p://tpep wa.org/the model/framework and rubrics/leadership frameworks/marzano school leadership evaluation model/ Mattison, R. (2008). Characteristics of reading disability types in middle school students classified ED. Behavior Disorders , 34 (1), 27 41. Mayhew, R. (2014). Employee t urnover v s. a ttrition . Chron: Small Business . Retrieved from http://smallbusiness.chron.com/employee turnover vs attrition 15846.html

PAGE 101

101 McAdams, R. P. (1997). A s ystems a pproach to s chool r eform. Phi Delta Kappan , 79 (2), 138 42 . McEwan, E.K. (2003). 10 traits of highly effective principals: From g ood to g reat p erformance. Thousand Oaks, CA: Corwin Press . McLaughlin, M. J. (2010). Evolving interpretations of educational equity and students with disabilities. Exceptional Childr en, 76 (3), 265 278. Metropolitan Life Insurance Company . (2013, February 21). MetLife Survey of the American Teacher: Challenges for School Leadership (2012) . New York: MetLife. Retrieved from at www.metlife.com/teachersurvey . Meyer, K. (2004). No c hild l eft b ehind: A national perspective. Spectrum: The Council of State Governments , 5 7 . Meyer, M. , & M acmillan, R. (2011). Principal succession and the m icropolitics of educators in schools: Some incidental results from a larger s tudy. Canadian Journal of E ducational Administration and Policy . N(117). Meyer, M. J., Macmillan, R. B., & Northfield, S. (2009). Principal succession and its impact on teacher morale. International Journal of Leadership in Education , 12 (2), 171 185. Miklos, E. (1988). Administrat ive selection, career patterns, succession, and socialization. In N. J. Boyan (Ed.), Handbook of r esearch on e ducational a dministration ( pp. 53 76 ) . New York: Longman. Miller, A. (2009). Principal turnover, student achievement and teacher retention. Educ ation Research Section. Princeton, NJ: Princeton University. Retrieved from:http://www.ers.princeton.edu/Miller.pdf Mitgang, L. (2003). Beyond the pipeline: Getting the principals we need, where they are needed . New York : The Wallace Foundation . Mobley, W. (1982). Some unanswered questions in turnover and withdrawal research. The Academy of Management Review , 7 , 111 116. presidents. The Chronicle of Higher Education , 55 (13), B7 B8. Murphy, J. (2008 a ). Turnaround insights from the organizational sciences: A review of the empirical evidence and the development of a staged model of recovery with potential implications for the PK 12 education sector. Leadership and Policy in S chools , 7 (4), 331 357.

PAGE 102

102 Murphy, J. (2008 b ) The place of leadership in turnaround schools: Insights from organizational recovery in the public and private sectors . Journal of Educational Administration , 46 (1), 74 98. Murphy, J., Elliot, S., Goldring, E., & Porter, A. (2006). Learning centered leadership: A conceptual foundation . Vanderbilt University. New York: The Wallace Foundation. Ogawa, R. , & Hart, A. (1985). The effect of p rincipals on t he instructional performance of s ch ools. Journal of Educationa l Administration , 23 (1), 59 72 . Orr, M. T., King, C., & LaPointe, M. (2010). Districts developing leaders: Lessons on consumer actions and program approaches from eight urban districts . Boston: Education Development Center, Inc. Papa, F., Lankford , H., & Wyckoff, J. (2002a). The attributes and career paths of principals: Implications for improving policy . New York: Teacher Policy Research Center , University at Albany, SUNY. Partlow, M. (2007). Contextual f actors r elated to e lementary p rincipal t urnover . Planning and Changing , 38 ( 1/2 ) . Peske, H. , & Haycock, K. (2006). Teaching inequality: How poor and minority students are shortchanged on teacher quality: A report and recommendations by the Education Trust . Education Resources Information Center. Retr ieved from http://eric.ed.gov/?id=ED494820 Porter, A., Goldring, E., Elliott, S., Murphy, J., Polikoff, M., & Cravens, X. (2008). Setting performance standards for the VAL ED: Assessment of principal eadership . Online Submission. Portin, B., Kapp, M., Da riff, S., Feldman, S., Russel, F., Samuelson, C., & Ling Yeh, T. (2009). Leadership for learning improvement in urban schools: Executive summary . Center for the Study of Teaching and Policy , University of Washington. Portin, B., Schneider, P., DeArmond, M., & Gundlach, L. (2003). Making sense of leading schools: A study of the school principalship . Center on Reinventing Public Education , University of Washington. Retrieved from http://www.wallacefoundation.org/knowledge center/school leadership/principal training/Pages/Making Sense of Leading Schools Study of School Principalship.aspx R Core Team . (2013). R: A language and environment for statistical computing. R Foundation for Statistical Computing , Vienna, Austria. URL http://www.R project.org/.

PAGE 103

103 Rainw ater, L., & Smeeding, T. (1995). Doing p oorly: The r eal i ncome of American c hildren in a c omparative p erspective. Luxembourg Income Study , 127 . Retrieved from http://eric.ed.gov/?id=ED388743 Rand Education. (2004). The careers of public school administra tors . Research Brief: Rand Education. Santa Monica, CA. Rashotte, C. A., Torgesen, J. K., & Wagner, R. K. (1997). Longitudinal studies of phonological processing and reading. Journal of Learning Disabilities, 27 (5), 276 286. Ready, D., & Conger, J. (2007 ). Make your company a talent factory . Retrieved from http://www.martinmeister.cl/wp content/uploads/2013/07/make your compagny a talent factory.pdf Reardon, S. (2011). The widening academic achievement gap between the rich and the poor: New evidence and possible explanations . Retrieved from http://cepa.stanford.edu/content/widening academic achievement gap between rich and poor new evidence and possible#sthash.6azbIm7o.dpuf Riddile, M. (2012). he common core . Reston, VA: National Association Secondary School Principals . Ringel, J., Gates, S., Chung, C., Brown, A., & Ghosh Dastidar, B. (2004). Career paths of school administrators in Illinois: Insights from an analysis of state data . Santa Moni ca, CA: Rand Education . Rivers, A. , & Sattler, C. (n.d.). School i mprovement and r estructuring under NCLB [PowerPoint slides]. Retrieved from http://www.fldoe.org/bsa/title1/pdf/sirestructure.pdf Robinson, V. M., Lloyd, C. A., & Rowe, K. J. (2008). The i mpact of leadership on student outcomes: An analysis of the differential effects of leadership types. Educatio nal A dministration Q uarterly, 44 (5), 635 674 . Rodgers, H. , & Payne, L. (2007). Child p overty in the American s tates: The i mpact of w elfare r ef orm, e conomics, and d emographics . Policy Studies Journal . Retrieved from http://onlinelibrary.wiley.com/doi/10.1111/j.1541 0072.2007.00204.x/abstract Ronfeldt, M., Loeb, S., & Wyckoff, J. (2013). How teacher turnover harms student achievement. American Ed ucational Research Journal , 50 (1), 4 36. Ross, M. (1998). The American college president . Washington, DC: American Council for Education (ACE).

PAGE 104

104 Rothwell, W. (2010). Preface. Effective succession planning: Ensuring leadership continuity, and building tal ent from within . New York: AMACOM. Roza, M. (2003). A matter of definition: Is there truly a shortage of school principals ? Center on Reinventing Public Education , University of Washington. Salmonowicz, M. (2009). Meeting the challenge of school turna round: Lessons from the intersection of research and practice. Phi Delta Kappan , 91 (3). Retrieved from http://www.pdkintl.org/kappan/k_v91/docs/k0911sal.pdf Samuels, C. (2012) Study: Principal t urnover b odes p oorly for s chools . Education Week . 31 (23), 10 . Seashore Louis, K., Leithwood, K., Wahlstrom, K. L. , & Anderson, S. E. (2010). Learning from leadership: Investigating the links to improved student learning . St. Paul, MN: University of Minnesota, Center for Applied Research and Educational Improvement . Schaeffer, L. (2002). The l eadership j ourney. Harvard Business Review , 80 (10), 42 47 . Sheppard, R. (2010). Determining factors that influence high school principal turnover over a five year period . (Doctoral dissertation). University of North Texas. Retrieved fro m http://digital.library.unt.edu/ark:/67531/metadc28476/m2/1/high_res_d/dissertatio n.pdf Sparks, D. (2013). Strong teams, strong schools. Journal of Staff Development , 34 (2), 28 30. Stogdill, R. M. (1974). Handbook of leadership: A survey of the literature . New York: Free Press. Stuit, D. (2010). Are bad schools immortal? The scarcity of turnarounds and shutdowns in both charter and district sectors . Washington, DC: Thomas B. Fordham Institute . Tabachnick, B. G. , & Fidell, L. S. (2007). Us ing multivariate statistics ( 5 th ed.). Boston, MA: Allyn and Bacon. Torgesen, J. K. (1989). Cognitive and behavioral characteristics of children with learning disabilities: Concluding comments. Jou r nal of Learning Disabilities , 22 (3),166 168. Tucker, J. ( 2010 , March 26 ) . 5 S.F. School p rincipals u nder f ire. San Francisco Chronicle .

PAGE 105

105 The Wallace Foundation . (2011). Request for proposals: Building a better principal pipeline to boost student achievement: A Wallace demonstration project with districts and p rincipal training programs . New York: Author. The Wallace Foundation . (2012). The school principal as leader: Guiding schools to better teaching and learning . New York: Author. The Wallace Foundation . (2013). Summary. Studies find high churn among school principals . New York: Author. Retrieved from http://www.wallacefoundation.org/knowledge center/school leadership/principal training/Documents/Six Districts Begin the Principal Pipeline Initiative.pdf Thurlow, M. L., Sinclair, M. F., & Johnson, D. R. (200 2, July). Students with disabilities who drop out of school: Implications for policy and practice. Issue Brief, 1(2). Minneapolis, MN: University of Minnesota, Institute on Community Integration, National Center on Secondary Education and Transition. Usda n, M., McCloud, B. , & Podmostko, M. (2000). Leadership for student learning: reinventing the principalship . Washington, DC: Institute for Educational Leadership . Retrieved from http://www.iel.org/programs/21st/reports/principal.pdf U .S. Department of Edu cation . (2001). Digest of education statistics . Washington, DC. U . S . Department of Education . (2010). Guidance on fiscal year 2010 school improvement grants . Washington, DC. Wahlstrom, K. L., & Louis, K. S. (2008). How teachers experience principal leade rship: The roles of professional community, trust, efficacy, and shared responsibility. Educational Administration Quarterly, 44 (4), 458 495. Waters, T., Marza no, R. J., & McNulty, B. (2003). Balanced leadership: What 30 years of research tells us about t he effect of leadership on student achievement . ( A working paper ) . Retrieved from ERIC database (Accession No. ED481972). Weinstein, M., Jacobowitz, R., Ely, T., Landon, K., & Schwartz, A. E. (2009). New schools, new leaders: A study of principal turnover and academic achievement at new high schools in New York City . New York: The Institute for Education and Social Policy, New York University. White, R. , & Cooper, K. (2011). Principals in succession: Transfer and rotation in educational administration . Do rdrecht: Springer. Weisberg, S. (2005). Multiple regression. Applied Linear Regression (3 rd ed.) , 47 68. New York: Wiley/Interscience.

PAGE 106

106 Wiersema, M. (2002). Holes at the top: Why CEO firings backfire. Harvard Business Review . Retrieved from https://hbr. org/product/holes at the top why ceo firings backfire/an/R0212E PDF ENG

PAGE 107

107 BIOGRAPHICAL SKETCH Sarah Burd grew up in Easton, Pennsylvania, and received her Bachelor of Science degree from Samford University (Birmingham, AL) in 2000, with a major in early ch ildhood and elementary education. She completed her Master of Science degree in elementary education also at Samford University in 2001. In fall 2007, she began a part time doctoral program whil e working full time as an early childhood education teacher. She graduated in August 2009 with an Education Specialist degree in educational leadership from the University of Florida (Gainesville) before attaining her Doctor of Education degree in educatio nal leadership in August 2014. Sarah has taug ht early childhood, elementary school, and high school. She currently teaches English at Naples High School. Sarah is married to Evan Burd. They are the proud parents of triplet boys: Everett, Jack and Owen, al l age 4.