<%BANNER%>

Relationships among Principals' Beliefs about Data-Driven Decision Making, Principal and School Characteristics, and Stu...

Permanent Link: http://ufdc.ufl.edu/UFE0022833/00001

Material Information

Title: Relationships among Principals' Beliefs about Data-Driven Decision Making, Principal and School Characteristics, and Student Achievement in Elementary Schools
Physical Description: 1 online resource (129 p.)
Language: english
Creator: White, Vicki
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: data, elementary, leadership, mediational, principal, quantitative, student
Educational Administration and Policy -- Dissertations, Academic -- UF
Genre: Educational Leadership thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: The role of the elementary school principal has changed as a result of increased accountability requirements, and principals have embraced data-decision making in order to make more informed decisions regarding student achievement. Much of the available research regarding the use of data-driven decision making has addressed its use by teachers to improve instruction. Less research focuses on its use by principals to effect student achievement. The purpose of the study was to examine the relationships among principal characteristics and school demographics, principals' beliefs about the use of data-driven decision making, and student achievement. Specifically the intent of the dissertation was to determine the mediating effects of data-driven decision making on student achievement. This census study addressed principals at public elementary schools within the state of Florida. The quantitative study utilized a web-based survey of principal beliefs about data-driven decision making. The student achievement data examined through the study utilized 2008 FCAT and NRT tests for Florida elementary schools. A series of multiple regression analyses were conducted to determine the relationship between the antecedent, outcome, and proposed meditational variables. Results showed that the principals' beliefs regarding the use of data-driven decision making do not act as a mediator for student achievement. The results of the study indicated that principals? in Florida elementary schools believe in the use of data-driven decision making within their schools, and they believe that the quality of the decision making within their schools has improved through its use. The results of the factor analysis indicated that four key constructs were present in Florida schools; beliefs regarding the use of data-driven decision making by teachers to affect student achievement, beliefs regarding data-driven cultures, beliefs regarding the systems that incorporate data-driven decision making, and beliefs regarding collaboration among teachers using data-driven decision making. A strong negative correlation was found between the number of students on free and reduced lunch and student achievement.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Vicki White.
Thesis: Thesis (Ph.D.)--University of Florida, 2008.
Local: Adviser: Quinn, David.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0022833:00001

Permanent Link: http://ufdc.ufl.edu/UFE0022833/00001

Material Information

Title: Relationships among Principals' Beliefs about Data-Driven Decision Making, Principal and School Characteristics, and Student Achievement in Elementary Schools
Physical Description: 1 online resource (129 p.)
Language: english
Creator: White, Vicki
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: data, elementary, leadership, mediational, principal, quantitative, student
Educational Administration and Policy -- Dissertations, Academic -- UF
Genre: Educational Leadership thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: The role of the elementary school principal has changed as a result of increased accountability requirements, and principals have embraced data-decision making in order to make more informed decisions regarding student achievement. Much of the available research regarding the use of data-driven decision making has addressed its use by teachers to improve instruction. Less research focuses on its use by principals to effect student achievement. The purpose of the study was to examine the relationships among principal characteristics and school demographics, principals' beliefs about the use of data-driven decision making, and student achievement. Specifically the intent of the dissertation was to determine the mediating effects of data-driven decision making on student achievement. This census study addressed principals at public elementary schools within the state of Florida. The quantitative study utilized a web-based survey of principal beliefs about data-driven decision making. The student achievement data examined through the study utilized 2008 FCAT and NRT tests for Florida elementary schools. A series of multiple regression analyses were conducted to determine the relationship between the antecedent, outcome, and proposed meditational variables. Results showed that the principals' beliefs regarding the use of data-driven decision making do not act as a mediator for student achievement. The results of the study indicated that principals? in Florida elementary schools believe in the use of data-driven decision making within their schools, and they believe that the quality of the decision making within their schools has improved through its use. The results of the factor analysis indicated that four key constructs were present in Florida schools; beliefs regarding the use of data-driven decision making by teachers to affect student achievement, beliefs regarding data-driven cultures, beliefs regarding the systems that incorporate data-driven decision making, and beliefs regarding collaboration among teachers using data-driven decision making. A strong negative correlation was found between the number of students on free and reduced lunch and student achievement.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Vicki White.
Thesis: Thesis (Ph.D.)--University of Florida, 2008.
Local: Adviser: Quinn, David.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0022833:00001


This item has the following downloads:


Full Text

PAGE 1

1 RELATIONSHIPS AMONG PRINCIPALS BELI EFS ABOUT DATA-DRIVEN DECISION MAKING, PRINCIPAL AND SCHOOL CH ARACTERISTICS, AND STUDENT ACHIEVEMENT IN ELEMENTARY SCHOOLS By VICKI CONRAD WHITE A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2008

PAGE 2

2 2008 Vicki Conrad White

PAGE 3

3 This dissertation is dedicated to my husband, my children, and my mother. I also dedicate this endeavor to the many people whos e collective efforts both large a nd small have shaped my life, allowing me to fulfill a life goal set twenty-seven years ago.

PAGE 4

4 ACKNOWLEDGEMENTS I would like to thank m y family, friends, co lleagues, and committee members for helping me to fulfill this dream. It would not have been possible without your support, encouragement, patience, and guidance. My husband has always been my biggest supporter. He has not only encouraged me through all of my endeavors, but he has always believed in me even when I had doubts about my success. Throughout our marriage he has provided me with love, patien ce, understanding, and emotional and financial support. He has always w illingly carried more than his share of the load when I needed to focus on school and work. No one could ask for a better husband, supporter, and friend; and I am thankful for every da y that he has been a part of my life. I also have three wonderful children who have given me much love and joy, and who have sacrificed a great deal throughout my doctoral studies. My son Douglas James, now a sophomore in college at the University of Florida, provi ded me with many hours of stimulating conversation and thought-provoking questions about my studies. My oldest daughter Margaret Ellen, a freshman at the University of Miami, has al ways listened and provide d unlimited amounts of encouragement even when I was tired, cranky, and overwhelmed. Finally, my youngest daughter Jennifer Lynn gave me many hugs and much love and more importantly, patience when I was late to pick her up or unable to attend important school events. I also have been fortunate in my life to ha ve the support of a loving family. My parents taught me early not only about th e importance of obtaining an excel lent education, but also about the joy that comes from life l ong learning. They provided me with so many opportunities to learn, always encouraging me to pursue my dream s. My mother has always urged me to set aggressive goals and then encouraged me to me et them. I would not be who I am today without her love and support. I have also received a great deal of encourag ement from both my sisters. In

PAGE 5

5 particular I must acknowledge my sister Conni who has been my closest friend and confidante throughout this journey. My return to university studies would not ha ve been complete wit hout the support of Dr. Jim Doud. His gentle encouragement allowed me to grow, helping me to overcome my insecurities. Dr. Doud was a wonderful role model who listened and guided me through his teaching, thought-provoking conversations, and em phasis on reflection. He gave me many hours of his time, and I am grateful for the opportunity to have known him. I know that throughout my career I will remember and reflect on the many thi ngs that he has taught me about leadership, the art of being a principal, and about education. Dr. Doud encouraged me to open my mind to the possibilities. My committee members also provided me with a great deal of support, each in their own way. Dr. David Quinn, my committee chair, also provided me with a great deal of gentle guidance and encouragement. He opened my mi nd to the power of data-driven decision making, and its role within the school. Dr. Quinns honest y and insightfulness helped me to work through issues and complete my studies. I must also thank Dr. David Mill er who taught me to love and respect statistics, a feat I never expected to accomplish. Dr. Millers extensive knowledge, love of teaching, patience, and willingness to give his time helped me to develop a love for quantitative research. I would also like to acknowledge Dr. Kat hy Gratto who provided me with encouragement and who was always willing to st ep in when needed. Her professionalism and willingness to give her time did not go unnoticed or unappreciated. Finally, I would also like to acknowledge Dr. Fran Vandiver who I believe embodies the many qualities that make an outstanding principal. Her dedication to her pr ofession and to the suppor t and development of teachers everywhere is reflected in all that she does.

PAGE 6

6 Acknowledgement must also be given to Ma rilyn Travis, without whom this journey would not have occurred. Mrs. Tr avis provided me with an opportuni ty to return to the field of education. She nurtured and supported me as I grew to become a better teacher and a school administrator. Marilyn Travis not only guided me but she provided opportunities to develop my own beliefs, strengths, and inte rests. I can never thank her en ough for all that she has done for me in my career and in my life. I must also acknowledge Dr. Connie Sorice, w ho has acted as both a friend and mentor throughout my doctoral studies. Dr Sorice not only listened, but she guided my studies and provided encouragement and resources every step of the way. Acknowledgements must also be given to my friend, James Hendricks. Mr. Hendricks taught me about printing, mass mail ing, and how to make efficien t and cost-effective decisions when conducting a survey of this magnitude. There are so many others who have helped me along the way. The support of the entire UF community, my friends at school and at home, and my coworkers are also much appreciated. In closing, as we so often see throughout al l aspects of life, the accomplishments of one person reflect the underlying efforts of the many people who touched their lives along the way. I am fortunate to have received so much love and support from so many.

PAGE 7

7 TABLE OF CONTENTS page ACKNOWLEDGEMENTS .............................................................................................................4 LIST OF TABLES .........................................................................................................................10 LIST OF FIGURES .......................................................................................................................11 ABSTRACT ...................................................................................................................... .............12 CHAP TER 1 INTRODUCTION AND OVERVIEW OF THE RESEARCH ............................................. 14 Introduction .................................................................................................................. ...........14 School Accountability ......................................................................................................... ...14 Role of the Public School Principal ........................................................................................15 Data-Driven Decision Making ................................................................................................ 17 Statement of the Problem ...................................................................................................... ..18 Purpose of the Study .......................................................................................................... .....19 Research Hypotheses ..............................................................................................................19 Instrumentation ............................................................................................................... ........20 Definition of Terms ................................................................................................................20 Delimitations and Limitations ................................................................................................ 21 Delimitations ................................................................................................................. ..21 Limitations ................................................................................................................... ....21 Significance of the Study ........................................................................................................21 Summary ....................................................................................................................... ..........22 2 REVIEW OF THE LITERATURE ........................................................................................23 Introduction .................................................................................................................. ...........23 Accountability ................................................................................................................ .........23 Use of Standardized Testing in Florida ..................................................................................25 Principal Leadership ...............................................................................................................26 Professional Learning Communities ....................................................................................... 28 The Principal As the Instructional Leader .............................................................................. 29 Principal Leadership and Da ta-Driven Decision Making ....................................................... 30 Barriers to Successful Data -Driven Decision Making ............................................................ 31 The Relationship Between Data-Driven Deci sion Making and Prin cipal Leadership ........... 32 Principal Leadership Standards ..............................................................................................34 Florida Principal Leadership Standards ...........................................................................35 Chicago Competencies for Data-Driven School Improvement ....................................... 35 Introduction to Data-Driven Decision Making .......................................................................36 Data-Driven Decision Making and Instructional L eadership .......................................... 36 Data-Driven Decision Making and School Improvement ............................................... 37

PAGE 8

8 Data-Driven Decision Making and Culture .....................................................................38 A Model for Data-Driven Decision Making ................................................................... 38 Data-Driven Decision Making Tools .............................................................................. 39 Communicating Through Da ta-Driven Dialogue ............................................................39 Extending Data-driven Decision Making to Data Mining .............................................. 40 Other Data-Driven Decision Making Tools .................................................................... 41 Limitations ................................................................................................................... ...........42 Summary ....................................................................................................................... ..........42 3 METHODLOLOGY ............................................................................................................... 44 Introduction .................................................................................................................. ...........44 Research Questions ............................................................................................................ .....44 Context of the Study ...............................................................................................................44 Participants .................................................................................................................. ...........45 Institutional Review Boar d Procedure and Approval ......................................................45 Population .................................................................................................................... ....46 Principal Respondents Personal Characteristics ..............................................................47 Instrumentation ............................................................................................................... .47 Student Achievement Data ..................................................................................................... 48 Mediational Model ............................................................................................................. .....48 Analysis of Hypothesis 1 ................................................................................................. 51 Analysis of Hypothesis 2 ................................................................................................. 52 Analysis of Hypothesis 3 ................................................................................................. 52 Data Analysis ..........................................................................................................................53 Summary ....................................................................................................................... ..........54 4 RESULTS AND ANALYSIS OF DATA .............................................................................. 55 Introduction .................................................................................................................. ...........55 Analysis of the Survey Instrument ......................................................................................... 56 Reliability Analysis .......................................................................................................... ......58 Analysis and Quantitative Results .......................................................................................... 58 Question 1 ........................................................................................................................58 Analysis ...........................................................................................................................59 Question 2 ........................................................................................................................60 Analysis ...........................................................................................................................60 Hypothesis 1 .................................................................................................................. ..60 Analysis of Hypothesis 1 ................................................................................................. 60 Hypothesis 2 .................................................................................................................. ..62 Analysis of Hypothesis 2 ................................................................................................. 62 Hypothesis 3 .................................................................................................................. ..64 Summary of Results ................................................................................................................66

PAGE 9

9 5 DISCUSSION, RECOMMENDATI ONS, AND CONCLUSIONS ......................................90 Discussion .................................................................................................................... ...........90 Study Purpose .........................................................................................................................90 Target Population ....................................................................................................................91 Summary and Discussion of Results ......................................................................................91 Factor Analysis ................................................................................................................92 1. What are the beliefs held by elementary school principals in Florida with respect to data-driven decision making? ..................................................................................96 2. Do Florida elementary school principa ls beliefs about data-driven decision m aking mediate the effect of principal characteristics and school demographics on student achievement? .............................................................................................. 99 Implications and Recommendations .....................................................................................101 Principals Hold Strong Beliefs Rega rding Data-D riven Decision Making ................... 102 Culture is an Important Part of Data-Driven Decision Making ....................................102 Supporting Data-Driven Deci sion Making for Teachers ............................................... 103 Closing the Gap ............................................................................................................. 104 Recommendations for Future Research. ...............................................................................104 Summary ....................................................................................................................... ........105 APPENDIX A PERMISSION TO USE THE DATA-DRI VEN DECISION MAKING R EADINESS PRINCIPAL SURVEY .........................................................................................................107 B REQUEST FOR PRINCIPAL EMAIL AND RESIDENTI AL ADDRESSES ....................108 C INSTITUATIONAL REVIEW BOARD APPROVAL ....................................................... 109 D INITIAL CONTACT LETTER ............................................................................................110 E INITIAL EMAIL LETTER ..................................................................................................111 F FOLLOW-UP EMAIL LETTER ..........................................................................................112 G FOLLOW-UP INFORMED CONSENT LETTER .............................................................. 113 H THANK YOU LETTER ....................................................................................................... 114 I IRB FOLLOW-UP AP PROVAL LETTER ..........................................................................115 J IRB APPROVAL TO SEND EMAIL TO VOL USIA COUNTY PRINCIPALS ................ 116 K DATA-DRIVEN DECISION MAKING SURVEY ............................................................. 117 LIST OF REFERENCES .............................................................................................................123 BIOGRAPHICAL SKETCH .......................................................................................................129

PAGE 10

10 LIST OF TABLES Table page 4-1 Total Variance Explained ..................................................................................................694-2 Rotated Factor Matrix ........................................................................................................704-3 Table of Data-Driven Decision Making Factors and Related Survey Questions .............. 714-4 Principal Beliefs Regarding Data-Driven Decision Making .............................................724-5 Frequency of Responses Regarding Pr incipal Beliefs Regarding Data-Driven Decision Making ................................................................................................................734-6 Descriptive Statistics for Antecedent Variables for Hypotheses 1. ................................... 754-7 Regression Coefficients, St andardized Regression Coefficients, t-test Statistics and Partial Correlations for Each of the F our Factors Addressed in Hypothesis 1. .................764-8 Descriptive Statistics Regarding Data -Driven Decision Making Factors and FCAT NRT and FCAT SSS Scale Scores for Grad es 3, 4, and 5 in Reading and Math .............. 774-9 Regression Model for Data-Driven D ecision Making Factors and FCAT NRT and FCAT SSS Scale Scores for Grades 3, 4, a nd 5 in Reading and Math for Hypothesis 2..........................................................................................................................................784-10 Unstandardized Regression Coefficients, Standardized Regression Coefficients, t-test Statistics, and Partial Correlations for Hypothesis 2. ........................................................ 794-11 Descriptive Statistics Re garding Principal Characteristics, School Demographics, and Data-Driven Decision Making Factor s and FCAT NRT and FCAT SSS Scale Scores for Grades 3, 4, and 5 in Reading and Math ..........................................................814-12 ANOVA Table for Data-Driven Decision Making Factors and FCAT NRT and FCAT SSS Scale Scores for Grades 3, 4, and 5 in Reading and Math While Controlling for Principal Characteristics a nd School Demographics for Step One and Two of Hypothesis 3. ......................................................................................................... 824-13 Regression Model for Data-Driven D ecision Making Factors and FCAT NRT and FCAT SSS Scale Scores for Grades 3, 4, and 5 in Reading and Math While Controlling for Principal Characteristics a nd School Demographics for Step One of Hypothesis 3................................................................................................................... ....84

PAGE 11

11 LIST OF FIGURES Figure page 3-1 Mediational Model (Baron and Kenny, 1986) ................................................................... 493-2 Proposed Mediational Model ............................................................................................. 504-1 Scree Plot for Maximum Likelihood Estimation Analysis. ............................................... 68

PAGE 12

12 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy RELATIONSHIPS AMONG PRINCIPALS BELI EFS ABOUT DATA-DRIVEN DECISION MAKING, PRINCIPAL AND SCHOOL CH ARACTERISTICS, AND STUDENT ACHIEVEMENT IN ELEMENTARY SCHOOLS By Vicki Conrad White December 2008 Chair: David Quinn Major: Educational Leadership The role of the elementary school principal has changed as a result of increased accountability requirements, and principals have embraced data-decision making in order to make more informed decisions regarding student achievement. Much of the available research regarding the use of data-drive n decision making has addressed its use by teachers to improve instruction. Less research focuses on its use by pr incipals to effect student achievement. The purpose of the study was to examine the relationsh ips among principal characteristics and school demographics, principals beliefs about the us e of data-driven decision making, and student achievement. Specifically the intent of the dissert ation was to determine the mediating effects of data-driven decision making on student achievement. This census study addressed principals at pub lic elementary schools within the state of Florida. The quantitative study util ized a web-based survey of prin cipal beliefs about data-driven decision making. The student achievement data ex amined through the study utilized 2008 FCAT and NRT tests for Florida elementary schools. A series of multiple regression analyses were conducted to determine the relationship betw een the antecedent, outcome, and proposed

PAGE 13

13 meditational variables. Results showed that the principals beliefs rega rding the use of datadriven decision making do not act as a mediator for student achievement. The results of the study indicate d that principals in Florida elementary schools believe in the use of data-driven decision ma king within their schools, and they believe that the quality of the decision making within their schools has improve d through its use. The results of the factor analysis indicated that four key constructs were present in Flor ida schools; beliefs regarding the use of data-driven decision making by teachers to affect student achievement, beliefs regarding data-driven cultures, beli efs regarding the system s that incorporate data -driven decision making, and beliefs regarding collaboration among teach ers using data-driven decision making. A strong negative correlation was found be tween the number of students on free and reduced lunch and student achievement.

PAGE 14

14 CHAPTER 1 INTRODUCTION AND OVERVIEW OF THE RESEARCH Introduction I believe that the transform ing movement that raises the serv ing quality of any institution, large or small, begins with the initiative of one individual personno matter how large the institution or substantial the m ovement.--Robert Greenleaf, 1980, p. 3 On January 2, 2002, President Bush signed into law the No Child Left Behind Act of 2001 (U.S. Dept. of Ed., 2001). This act was to precip itate massive changes in the field of education, particularly in the leadership role of the public school principal with respect to school accountability and achievement for all childre n (Lunenberg and Ornstein, 2004). Seven years later and nearly half-way to the 2014 deadline mandating proficienc y for all students on state standardized tests, the role of the principal has changed, yet imp rovement in student achievement is questionable (Darling-Hammond, 2007). It is important for edu cators not only to develop an understanding of how the role of the principal has changed, but also to determine whether or not these changes are related to an increase in stude nt achievement for elementary school students. One of the changes adopted by many school prin cipals is the use of data-driven decision making as part of school leadership. Questions em erge as to the extent to which public school principals are using these skil ls and whether or not the use of data-driven decision making practices affect student achievement. School Accountability School accountability has been a political issu e for the past twenty-five years, but the focus on student mastery of standards and impr oved academic performance for all students has grown exponentially since the turn of the century and the implem entation of the NCLB Act in 2002. Under NCLB, the Federal government expande d its control, tying funding to student achievement. Public schools are required to demons trate adequate levels of student proficiency if

PAGE 15

15 they are to receive resources and funds from the Federal government (U.S. Department of Education, 2001). Under NCLB, scho ols are required to report st udent academic achievement as measured through standardized testing for a va riety of racial and ec onomic subpopulations as well as for students with disabilities. Schools ar e also obligated to meet state standards for academic improvement in math and reading. To meet the needs of students and to comp ly with the requirements of NCLB, public schools in the state of Florida have enacted a series of reforms designed to improve academic achievement in reading and math. Standards will only be met when schools provide an academic environment focused on meeting the needs of students, improving instructional practices by teachers, providing safeguards that ensure succ ess for all students, instituting practices for feedback and assessment, developing relationships with key stakeholders, and corrective action and problem solving (Darling-Hammond et al., 1993). Much of the reform effort in the state of Florida has manifested itself through the schoo l improvement plan, a plan which must be developed, implemented, and monitored in every public school on a yearly basis. Role of the Public School Principal A significan t portion of the burden for school accountability has fallen on the shoulders of the public school princi pal, the person who is most visible to the general public. In order to support the increased focus on accountability, the role of the principal has evolved to include that of leader, mentor, learner, pol itician, supervisor, advocate, an d manager (Matthew and Crow, 2003). Gone are the days of the principal, who sits in the office, supervising employees and monitoring the day to day operations of the school The successful elementary school principal takes an active role with resp ect to the foremost business of the school, that of educating students. Michael Fullan identifies the school principal as a change agent (Fullan and Striegelbauer, 1991). The many role s of the elementary school prin cipal represent a change in

PAGE 16

16 focus from management of school operations to that of a school reformer responsible for designing an academic environment that meets th e needs of all students regardless of their background, abilities, or curre nt level of performance. How exactly does one create an academic e nvironment that meets the needs of all students? There is a great deal of research that supports a variet y of activities and conditions that are present in successful schools. Educational experts support im plementation of a variety of strategies including building pa rtners with family and community members, professional learning communities, collegiality, shared-decisi on making between administration and faculty, action research, targeted professional development, and a positive healthy school culture (Peterson and Deal 1999; DuFour, 2004; Reeves, 2004; Schmoker, 2003; Marzano, 2003). Others support the use of common assessments formative assessments, goal setting, studentcentered learning, teamwork, and student progre ss monitoring (Reeves, 2004; Schmoker, 2003; Marzano, 2003; Stiggins, 1999). In order to provid e the appropriate leadership and balance for all of these many activities, the principal must also be able to engage in systems thinking (Senge, 1990). NCLB specifically requires that the principal h ave the skills to help teachers teach and students learn. (U.S. Dept. of E d., 2001). As part of the effort to improve student achievement, principals are now taking a more active role in managing and monitoring classroom activities. In order to meet the challenges of NCLB, principals are expected to strengthen their role as the instructional leader of the school. As the instru ctional leader, the prin cipal seeks to emphasize doing the right things correctly, in cluding acting as a resource pr ovider, instructional resource, communicator and visible presence (Smith and A ndrews, 1989). As an inst ructional resource the principal provides specific classroom support to teachers by modeling instructional behaviors

PAGE 17

17 and supporting staff development and the instruc tional concerns of teachers (Marzano, et al., 2005). As active instructional le aders within the school, the pr incipal provides resources, including instructi onal materials, teaching tools, equipmen t and facilities (M arzano, et al., 2005). As the instructional leader, the principal communicates the vision of the school, focusing efforts in support of that vision (Marzano, et al., 2005). The principal mon itors student performance and alignment with curriculum. According to Richar d Elmore, the principal should also understand the effective practices that enhance curriculum, instructi on, and assessment (Elmore, 2000). Data-Driven Decision Making As principals continue to strengthen their ro le a s the instructional leader of the school, many turn to the use of data-driven decision ma king as a mechanism for understanding strengths and weaknesses within their school. They seek to make informed decisions based on actual data, rather than on intuition. At th e school level, they may seek to evaluate the school at a higher level, analyzing the culture, progr ams, operations, facilities, and st aff. Victoria Bernhardt (1998) postulates that we must first look at the system that produces results, a nd then make adjustments that focus on continuous systemic improvement (p.13). Bernhardt identifies four domains of information that should be collected and analyzed in order to create a comp lete assessment of the school. These four domains are student demogra phics, perceptions, school processes and student learning. For Bernhardt the emphasis is on impr oving student achievement but she recommends doing this by looking at the big pi cture of the school, the interr elationships between the four domains, and the details within each of the four domains. With respect to student performance, data-d riven decision making represents a systematic method of collecting student data so that administrators, teachers and parents can accurately assess student learning. They can then make decisions based on the data to improve administrative and instructional systems so as to continually promote student achievement.

PAGE 18

18 Educational leaders such as D ouglas Reeves (2004), Richard DuFour (2004), Michael Schmoker (1999), Robert Marzano (2005), a nd Richard Stiggins (1999) a ll emphasize using data to establish baselines, set goals, conduct assessments, and adjust instructional practices in order to achieve desired results. According to Sylvia Me ndezMorse (1991), when principals use data about trends in students performance to adjust the curriculum or instructional practices being used, instruction is maximized. (p.5). Statement of the Problem Because of the increased emphasis on school a ccountability, the role of the principal has evolved over the past seven years. As the prin cipals role becomes mo re complex and s/he becomes more actively involved in setting goa ls, monitoring student progress, promoting professional learning communities, providing opportunities for collaboration and professional development, and influencing classroom instructi on and alignment of curri culum with standards, the need to use data-driven d ecision making practices also b ecomes more pronounced. Much of the available research about th e relationship between data-drive n decision making and student achievement is qualitative in nature. There are a few studies that target the principals use of data-driven decision making, but th ese do not address the role of the elementary school principal nor do they focus on the use of both FCAT SSS a nd FCAT NRT data. In addition, this study also supports existing research regarding the changing role of the principal as a result of NCLB. During the past twenty years there have been several quantitative studies that address the relationship between principals characteristics and student ac hievement (Halinger and Heck, 1996). An analysis of these studie s indicated that those studies th at utilized a mediated-effects model showed a more positive effect. The Baron and Kenny (1986) model presents a framework for testing the effect of a mediati ng variable on the outcome variable.

PAGE 19

19 Purpose of the Study The purpose of the study was to exam ine the relationships among principals characteristics and school demographics, principals beliefs about the use of data-driven decision making, and student achievement. Sp ecifically the intent of the dissertation was to determine the mediating effects of data-driven d ecision making on student achievement. Question 1: What are the beliefs held by elementary schoo l principals in Florida with respect to data-driven decision making? Question 2: Do Florida elementary school principals beliefs about data-d riven decision making mediate the effect of principal characte ristics and school dem ographics on student achievement? Research Hypotheses Hypothesis 1: Principal characteristics (years ex peri ence, years at a sc hool, and level of education) and school demogra phics (student enrollment, percentage of students on Free and Reduced Lunch) are related to the princi pal beliefs about data-d riven decision-making. Hypothesis 2: Principals beliefs about the use of data -driven decision making skills are related to student achievement. Hypothesis 3: Principals beliefs about the use of data -driven decision making skills mediate the relationship between principal characteristic s (years experience, years at a school, and level of education) and school demographics (s tudent enrollment, percentage of students on Free and Reduced Lunch) on student achievement. The Baron and Kenny (1986) model presents a framework for testing the effect of a mediating variable on the outcome variable. Each of the hypotheses test a step identified in the Baron and Kenny Model. The first hypothesis tests st ep one in the model. It measures the effect of data-driven decision making (t he mediating variable) on school demographics and principal characteristics (the antecede nt variables). The second hypot hesis seeks to measure the relationship between data-driven decision ma king (the mediating variable) and student achievement (the outcome variable ). The last hypothesis regresse s student achievement (outcome

PAGE 20

20 variable) on both principal charact eristics and school demographics (antecedent variables) and data-driven decision making (the antecedent variable). Instrumentation This study used the Statewide Data-Driven Readiness Study Principa l Survey developed by Dr. Scott McLeod (2005) at the University of Minnesota. The study was conducted through a com bination of a web-based survey for the in itial contact and a follow-up mail survey to nonrespondents in order to gain a se lf-report by principals regarding their beliefs about data-driven decision making. The measurement of student achievement data utilized scale scores for third, fourth, and fifth grade FCAT and NRT data for reading and math for the 2007-2008 school year, grouped by school and grade level for each year. Definition of Terms The following definitio ns were used as part of this study. Adequate Yearly Progress (AYP) is the Federal government rating system for public schools and it can be achieved in the state of Florida when a school meets all 39 criteria. Collegiality is a condition where teachers respect each others abilities, share methods and techniques, and work together to improve student learning. Data-driven decision making is the use of data to make informed decisions with respect to student progress rather than relying on intuition or incomplete data. Florida Comprehensive Assessment Test SSS (FCAT SSS) is a criterion-referenced test that is designed to assess mastery of the Fl orida Sunshine State Standards. FCAT Comprehensive Assessm ent Test NRT (FCAT NRT) is a norm-referenced test (based on the SAT-10) that classifies or ranks students against their peers throug hout the nation. Formative assessment occurs when feedback from the learning activity is used to adjust teaching to meet the students needs. Instructional leadership is represented through those practices used by the principal to improve student learning. They include allocating resources, establishi ng clear goals and objectives, monitoring instruction and lesson plans, aligning instruction to the curriculum and providing teacher support and evaluation.

PAGE 21

21 Professional Learning Communities are organizations that promote teacher collaboration, peer observation, action research and study groups, a focus on student learning, and a commitment to continuous improvement. School accountability is the requirement for schools to be responsible for teacher and administrative actions and their effect on studen t achievement. It includes the requirement to report, explain and be answerable for success or failure of school operations. School culture is represented through the shared values, beliefs, priori ties, expectations, rituals, and norms through which a school manifests itself. Systematic progress monitoring occurs when teachers and administrators regularly to assess students academic performance on a regular basis. Delimitations and Limitations Delimitations The sam ple was limited to elementary school principals in public schools that support pre/kindergarten thr ough fifth grade. The study incorporated the use of FCAT SSS an d FCAT NRT test data for grades three, four, and five in reading and math in public elementary schools in Florida Limitations The study was lim ited to public schools and publ ic school administrators in Florida. Specifically the sample was limited to elementary schools that teach grades prekindergarten or kindergarten through grade five. Magnet schools and charter schools were not included in the study. There was no provision for open-ended questions or input on the survey. Significance of the Study There is a growing body of resear ch supporting the cha nges in the role of the principal in the elem entary school. Much of this research focuses on the role of the principal as the instructional leader or his/he r influence on school culture, prof essional development activities, and school improvement. There is very little rese arch that supports the relationship between the principals beliefs about the us e of data-driven decision making skills and student achievement. Findings from this study provide additional insight to the degree of use of data-driven decision making practices by principals and the potenti al effect on student achievement. It also

PAGE 22

22 contributes to the growing body of literature surrounding the changi ng role of the principal in public schools as evidenced by the use of specific data-driven d ecision making skills. Because data-driven decision making is curren tly a popular topic, there is a wealth of literature, but little re search to support the effectiveness of its use by principals. Much of this literature focuses on the use of data-driven decision making by teachers and the need for formative assessment, progress monitoring, profe ssional learning communiti es, and collaboration between administrators, teachers, family members, and the community. Research that focuses on its use by administrators to effect change in student achievement is less available. Summary As the deadline for 100% proficiency for all students under NCLB looms ever closer, the pressure to improve student achievem ent becomes greater. Supovitz and Klein (2003) suggest that only when we use systematic data analysis to support student achievement will we begin to meet our goal of improved learning outcomes fo r all students. Educational leaders who truly understand the structure of schools and the needs of our students embrace the use of data-driven decision making practices. Chapter 2 provides a review of the literature surrounding the use of data-driven decision making by principals and how its use can affect student achievement. Al so discussed are the standards and competencies for pr incipals that relate to data-driven decisi on making. Finally, the chapter provides additional information about the accountability requireme nts, including the use of standardized testing in the state of Florida.

PAGE 23

23 CHAPTER 2 REVIEW OF THE LITERATURE Introduction Chapter 1 provided an overview of the st udy, including a statem ent of the problem, the significance of the study, and the associated research questions. The purpose of the study was to determine the relationship between principal ch aracteristics, school demo graphics, and student achievement in relation to the principals beli efs about the use of da ta-driven decision making skills. Chapter 2 continues with a discussion of school accountability and its effect on the leadership practices of school principals. The chap ter discusses the role of the principal as the instructional leader and the associated need for data-driven decision making within schools. Key attributes of data-driven decisi on making are also presented within the chapter. Also included in the chapter are key principal leadership studi es that address the relationship between the principal and student achievement. Studies that review the principa ls use of data-driven decision making and student achievement have also been included. Finally the chapter will close with a review of the research that supports the constructs set forth in the study and a discussion of the potential issues associated with research models that analyze the influence of the principal on student achievement. Accountability In 1983 the National Commission on Excellence in Education published A Nation at Risk (National Commission on Excellence in Educat ion, 1983). The report implied that the public education system in the United Sates was failing to meet the needs of corporate America. The commission made thirty-eight recommendations including 4 years of English, 3 years of mathematics, 3 years of science and one-half ye ar of computer science for high school students; the establishment of standards a nd expectations for all students; standardized testing at key

PAGE 24

24 transitional levels; seven hour sc hool days in a 200-220 day school year; teaching standards; and programs to support the gifted, socio-economi cally disadvantaged, minority and language minority students and handicapped. Nearly twenty years later, President George W. Bushs education agenda was approved and implement ed. The new agenda for education included testing and school accountability mandates, placi ng the focus on curriculum and instructional practices. The No Child Left Behind Act of 2001 (U.S. Department of Ed., 2001) ties Federal funding to accountability for result s in student achievement. The NCLB Act requires that states test students against state standards in readi ng, science, and math (U.S. Dept. of Ed., 2001). Educational experts supporte d the call for accountability, but focused on the underlying needs of students and teachers wi thin the schools rather than st andardized testing. As early as 1993, Linda Darling-Hammond and J on Snyder called for schools to develop learner centered accountability to put in place in structional practices and syst ems that support feedback and assessment in order to keep students from falling through the cracks (Darling-Hammond and Snyder, 1993). In 1998, Black and Wiliam conducted an analysis of 280 research articles, identifying the need to use formative assessment to raise standards of achievement (Black and Wiliam, 1998). Richard Stiggins advocat ed placing an emphasis on assessment for learning rather than assessment of learning, and the resulting access to more frequent evidence of student mastery of standards (Stiggins, 1999). While assessment of learning focuses on meeting accountability requirements, assessment for lear ning seeks to use assessments to promote additional learning by actively i nvolving the student in the process. Instead of focusing on standardized test results alone educational experts suggested that schools should focus on developing a holistic accountabil ity that emphasizes teacher pr actices, assessment, feedback and collaboration, curriculum, and leadership (Reeves, 2003). A more productive approach for

PAGE 25

25 student accountability would be to develop world-class standards, curricula, and assessments that help to improve teaching (Darling-Hammond, 20 07). Another approach proposed by Willard Daggett is the use of data to provide a more rigorous and relevant cu rriculum (Daggett, 2000). Accountability involves acceptance of respons ibility for student learning by all school staff. In 2005, Kannapel and Clements identified eight items that are an integral part of the culture for high poverty high performing schools. Th ey include a belief that all students can learn, collaboration across the school setting, teacher acceptance of responsi bility for success of failure, school, staffing, communica tion with parents, a caring staff, ongoing assessments that contribute to individualized in struction, and a curriculum that is aligned with instructional practices and assessments (K annapel and Clements, 2005). Use of Standardized Testing in Florida The state of Florida m eets the requiremen ts set forth by the Federal government and NCLB by conducting two types of st andardized testing for public school students in grades three through ten in reading, mathematics, science, and writing; criterion-referenced and normreferenced. The purpose, structure, scoring method, and content are different for each of the two tests. The school grading system for the state of Florida is based on a cr iterion-referenced test, and the test results are used to determine school grades (Florida Dept. of Education, 2005). Additional funding is provided to schools that meet state requirements and receive an A on the school report card. The Florida Comprehensive Assessment Test of Sunshine State Standards (FCAT SSS) is a criterion-referenced test designed to assess ma stery of the Florida standards for excellence in education (Florida Department of Educa tion, 2005). The FCAT SSS compares student performance against a predetermined performance level. Students are scored at levels one through five. A score of level one indicates that th e student has not mastered the state standards.

PAGE 26

26 Students with scores of level thr ee through five are considered to have mastered the standards to an adequate level or greater. The second standardized test used in the state of Florid a is the FCAT NRT, a norm referenced test that is used to classify or rank students against thei r peers throughout the nation. In 2005, Florida adopted Stanford 10 which was developed by Harcourt Assessment, Inc. The Stanford 10 is aligned with both stat e and national standard s. Test items are classified either as basic unde rstanding or thinking skills. About 15-20% of the Stanford 10 questions fall into the basic category that emphasizes simple recall and identification. The remainder of the questions target thinking skil ls such as analyzing, synthesizing, classifying, sequencing, compare and contrast, evalua tion, predicting, hypothesizing, and drawing conclusions (Florida De pt. of Education, 2005). Principal Leadership One of the goals of the NCLB Act of 2001 is to increas e accountability for school principals. The publication of school data via th e school report card and the Adequate Yearly Progress (AYP) report provides parents and community members with insight into a schools progress with respect to student achievement for all subpopulations. The school report card profiles progress based on student performance and growth over a years time in math, reading, science, and writing. The AYP report specifies th e schools progress based on 39 criteria. School data is disaggregated based on racial affiliati on, socioeconomic status, and classification as a student with a disability. The pr incipal is responsible for diss eminating this information to parents and key stakeholders. According to the U.S. Department of Educa tion, principals are re sponsible for program selection, curriculum, arrangement professional development, and allocation of school resources (U.S. Department of Education, 2001 ). Specifically the law states that principals must have "the

PAGE 27

27 instructional leadership skills to help teachers teach and students learn," and "the instructional leadership skills necessary to help students meet challenging state student academic achievement standards" (Title II, Section 2113 (c)). Because the principal is the primary administrator at the school level, the principal becomes the primary focus for public scrutiny. This visibility has contributed to the changes in the role of the pr incipal which has evolved during the past twenty years to become that of leader, supervisor, manager, mentor, learner, advocate and politician (Mathews and Crow, 2003). The increased complexity of the principals role and the increased accountability has intensified the need for the principal to become a master of change. The words of Joel Barker (2005) are particularly true with respect to pub lic schools in America, No one will thank you for taking care of the present if you have neglected the future. Jerry Patterson (1993) argues that leadership is about changing a nd that school changes should incl ude an openness to diversity, active participation, learning from mistakes, controversy, and de legation The school principal acts as a change agent (Fullan and Striegelbauer, 1991) where change is a process, not an event (Fullan, 2001, p.5). According to Fu llan (2001), real school change is nearly impossible because the system itself is reluctant to change and th ere are no definitive answ ers to its problems and dilemmas. Successful school lead ers view change as opportuniti es, balancing the relationship between the group and the individual, and always understanding that chan ge cannot be mandated because it must be internalized at the local level (Fullan, 2001). It is important for school leaders to practice the five disciplines for creating a learning organizati on within the school, that of personal mastery, shared vision, mental models, team learning, and systems thinking (Senge, 1990).

PAGE 28

28 Professional Learni ng Communities School leadership must embrace change by creating and supporting professional learning communities that are focused on mission, vision, and continuous improvement; results-oriented; and collaborative in nature (DuFour, 2003). The professional learning community offers a vehicle through which to effect maximum change, and principals must evaluate the extent to which students are learning the intended outcomes and how they are giving both students and teachers the support necessary to improve le arning (DuFour, 2002). Richard DuFour (2004) defines a professional learning community as one where teachers work together to answer three key questions: 1) What do we want students to learn? 2) How will we know when each student has learned it? and 3) How will we respond when a student experiences difficulty in learning? Organizations that use the professional le arning community approach promote teacher collaboration, peer observation, action research and study groups, a focu s on student learning, and a commitment to continuous improvement. Teachers have a clear sense of the mission and vision for the school, and they share beliefs and values. They are life-l ong learners who are results-oriented. In order to develop clear answer s to the questions, teachers must be able to set goals and assess student progress. Through collaboration they are ab le to adjust instruction to better support the needs of the students. As results-oriented life-long learners, they are able to seek out new methods and ideas in order to meet their students learning needs. Finally, Michael Schmoker (2006) writes that learning communities must be more rigorous, and must meet regularly in order to be effective. For schools to be successful school administrators must ensure that teachers teach the same standards during th e same time period using the same assessments (Schmoker, 2003).

PAGE 29

29 The Principal as the Instructional Leader Peterson and Deal (1999) argue that the role of the principa l has changed from that of manager to visionary. Principals must act as a guide and a coach for professional development, a manager who provides time and resources, and an in structional leader; and they must also inspire collaboration. As visionary leaders principals must create a connec tion to others through a shared vision and be able to adapt to extrem e pressures for change (Bennis, 2003). Research shows that the teacher is the single most important factor that affects student achievement (Marzano, 2003). Teacher effectiven ess is influenced by their choices in instructional strategies, the ability to design effective class curriculu m, and their classroom management skills (Marzano, 2003). Therefore, it is cr itical that the principa l take an active role in maximizing the instructional capacity within the school, focusing on their role as that of instructional leader. Principals who promote inst ructional leadership understand that effective teaching practices affect the curriculum, inst ruction and assessment (Elmore, 2000). As the instructional leader of the school, the princi pal is responsible for monitoring the use of instructional techniques, deliv ery of the curriculum, and cl assroom management techniques (Marzano, 2005). As the instructiona l leader, the principal must also take an active role in the implementation and monitoring of the curricul um (Schmoker, 2005). Smith and Andrews (1989) identified four key roles that th e principal plays with respect to teachers: (1) resource provider, (2) instructional resource, (3) comm unicator, and (4) visible presence. As an active instructional leader, the prin cipal models instructional behaviors and supports staff development (Marzano et al., 2005 ). They make curriculum and instructional changes that maximize student le arning (Schmoker, 1999). According to Reeves, principals can maximize instruction by providing focus, refi ning strategic planning, and creating an environment that supports holis tic accountability or the system atic monitoring of instruction,

PAGE 30

30 curriculum, and leadership (Reeves, 2003). As an instructional lead er responsible for accountability, it is critical that the principa l focus on teaching, leadership, curriculum, and parent involvement (Reeves, 2004). Principal Leadership and Data-Driven Decision Making The use of data to support and influence decision-m aking has been a key component of sound business practices for centuries certainly in the United States since the days of Frederick Taylor in 1881. Only recently have school administ rators attempted to cons istently apply its use in educational environments. Da ta-driven decision making represen ts a system of data-driven practices designed to collect and interpret info rmation necessary to make informed decisions (McLeod, 2005). These data-driven practices can be used to addr ess not only issues related to student achievement, but also ot her school management issues a ssociated with running a school. From a school administrator pers pective data-driven d ecision making represents a tool which can be used to shape school improvement, growth, and change. As part of the role of instru ctional leader, the administrato r must also influence teachers to use data to refine and adjust teaching pract ices, resulting in improved student performance (Schmoker 2005; McLeod, 2005). School leaders must develop and use effective strategies for data collection and analysis, and must help teach ers to understand and work with data to improve learning in the classr oom (Creighton, 2007). Another perspective for principal leadership is represented by Panettieri (2006), who identifies five important means for implementi ng data-driven practices. In addition to data collection, a school wide emphasis on outcome assessments, progress monitoring and feedback, and teacher ownership of outcomes, successful data driven decision-making relies on the administrators ability to build a learning organization. Accord ing to Joe Kitchens (2005), one of the most important factors is to provide teachers with real time access to student data. One must

PAGE 31

31 ensure that the technology is integrated, allowing teachers and administrators to use the data to predict AYP, and make adjustment during the year to intercede. The emphasis should be to leverage data to drive o ngoing student improvement. Barriers to Successful Data-Driven Decision Making Educational leaders have also determ ined several barrier s to successful data-driven decision making. A recent study by Ingram (2004) identified five major areas of concern for teachers. The first barrier was a mistrust of the data by both teachers and ad ministrators. If datadriven practices are not a part of the school culture, this distrust is more li kely to occur. Time and resources always have an effect on the ability to affect change. Teachers al so indicated that they needed more professional development in orde r to effectively use data-driven decision making. The lack of consistency for measurement was also a concern. Teachers also felt that it took too much time to collect and analyze information in order to make decisions. Perhaps the most interesting barrier identified by Ingram was teacher efficacy. Ma ny teachers still believe that learning is the responsibility of the student, and that teach ers are only responsible for instructional delivery. Felix (2005) suggests that teache rs are also affected by a lack of training, but her research also identified several other barriers that affect a schools ability to continuously improve. They are interoperability of data systems, an absen ce of clear priorities, outdated technology, a failure to collect data uniformly, low quality data, lack of training for data collection, and a perception that data-driven decisi on making is too complex. She recommends that schools use teams to assist with implementation. She also emphasizes the importance of administrative leadership to ameliorate poor conditions.

PAGE 32

32 The Relationship Between Data-Driven D ecision Making and Principal Leadership There are several s tudies that address the relationship between data-driven decision making and principal leadership. Jonathan Supov itz and Valerie Klein (2003) conducted a study addressing the use of student performance data to influence school improvement. Several leadership attributes came out of the study. They stated While examples of inventive data use came from both formal and informal leadership within the schools we examined, in most cases, the principal was the driving force behind strong data usethe principals constant emphasis on data that turned the data fr om numbers on a page into acti on in the classroom. (Supovitz & Klein, 2003, p. 36). Specifically, there are several dissertation studi es that address the relationship between principal leadership and student achievement The results of these studies are mixed. For her dissertation study, Dr Carla Van Fossen Mathews (2002) conducted a qualitative study of six middle school administ rators in Virginia. The purpos e of the study was to determine how administrators reacted to an identified need for change based on data, what influenced their decisions, and how they assessed the decisions that were made. Results indicated that the administrators who participated in the study reacted positively to the data and need for change. When made aware of the need for change, the principals became aware of the other problems that influenced the need, and they began to collaborate with team members and to create systematic processes in order to effect change One interesting result of the study is that principals do not always follow through to assess their decisions. Dr. Cathryn Anderegg (2007), a st udent at Pepperdine Univers ity, studied the use of datadriven decision making practices for teachers and ad ministrators in the state of Alaska. She also explored the effect of staff development on both teachers and administrators. She concluded that although teachers and administrators continue to focus on intuition to make decisions, 91% of

PAGE 33

33 school administrators analyze da ta to make policy and program decisions. She also concluded that teachers who have had annua l staff development in data-driven decision making are more likely to engage in collaborative activities. They benefitted from the ongoing support from their peers, and worked more close to a ssess results and plan interventions. In a study of principal perceptions regardi ng implementation of professional development for educational reform, elementary school prin cipals reported a greater incidence when reviewing and analyzing student work and building partnerships for learning (Patten, 2006). Dr. Leanne Bettesworth (2006) at the Univers ity of Oregon conducted a mixed-method quasiexperimental study of the effect of staff devel opment in statistics on the use of and efficacy for data-driven decision making for administrators. Three seminars were given to thirty-one participants in the study. Each module included a pretest, PowerP oint presentation, practice, and a post test. Results indicated that administrators did learn how to us e data as part of the decisionmaking process, but that their confidence was low. The lack of confidence affected efficacy, limiting the overall use of da ta-driven decision making. Recently Sluser (2006) at Montana State Un iversity conducted a survey studying the relationship between the administrator use of data-driven decision making and student achievement in high school mathematics. Result s of the study indicated that Montana school administrators who participated in the study had a higher perception of their ability to engage in data-driven decision making. However, results did not support any significant relationship between student achievement and the use of technology for data-driven decision making. In another dissertation study, John Arnold ( 2007) conducted a quantitative study in South Carolina examining the relationship between the schools capacity to use data-driven decision making and student achievement. The data for the study included 267 survey responses from

PAGE 34

34 middle school principals. Specifically the study was designed to determine if there was a relationship between the data-driven decision ma king and improvement in student achievement. Student data used for the study was represen ted through the 2006 Absolute and Improvement Indices and report card data. Re sults of the study showed a w eak, yet significant correlation between the school capacity to use data-driven decision making and student improvement. Finally Susan Hutton (2007) conducted a quantitative study comparing Virginia principals use of student achievement data in the decision making pro cess. Electronic survey responses for 452 principals were analyzed and compared with current literature. Huttons study resulted in four data-driven decision making domains including analyzi ng data, reporting and communicating through data, using data for scho ol improvement, and creating a data-friendly culture. Principal Leadership Standards The National Association of Ele mentary Schoo l Principals (NAESP, 2004) incorporates the use of data into its standa rds for instructional l eadership for elementa ry school principals. Standard Five requires the use of multiple sources of data as diagnostic tools to assess, identify and apply instructional improvement. The suggested strategies include use of a variety of data sources to measure performance, analysis using a variety of strategies, using data as tools to analyze student weaknesses and make adjustme nts to instruction, benchmarking against other schools with similar demographi cs, and creating a data-drive n school environment. Other organizations that have created similar strate gies include American Association of School Administrators (AASA, 2004) and the Intersta te School Leaders Lice nsure Consortium (ISLLC, 1996).

PAGE 35

35 Florida Principal Leadership Standards The Florida Princip al Leadership standards focus on the principal as the instructional leader and the use of data for effective deci sion making. These standards measure performance of principals at three levels; entry level prin cipal, effective princi pal, and high performing principal. The ten standards include vision, in structional leadership, managing the learning environment, community and stakeholder partne rships, decision making strategies, diversity, technology, learning accountability and assessment, human resource development and ethical leadership, and their effects on continuous school improvement and student achievement (Florida Dept. Of Education, 2005) With respect to data-driven decision making, the Florida Standard 8.0 Learning, Accountability, and Assessment comes closest to supporting these practices. Specific standards for entry level principals include the following: 8.1 Uses data to assess and monitor school improvement 8.2 Uses multiple sources of data to inform decisions and improvement processes 8.3 Monitors and assesses student progress 8.4 Monitors and assesses the progress of activities 8.6 Develops and demonstrates skills in eval uating instructional stra tegies and materials 8.7 Understands how to use diagnostic tools to assess, identify, and apply instructional improvement 8.8 Works with staff to identify strategies for improving student ach ievement appropriate to the school population Chicago Competencies for Data -Driven School Improvement All of the above practices also support the Chicago Com peten cies for data-driven school improvement identified by the Chicago School district and Dr. Sco tt McLeod (2005). Although still in the preliminary stages of development, these standards identify particular data-driven

PAGE 36

36 practices for school administrators in the domains of essential concepts, collecting and analyzing summative assessment data, setting measurable goals, collecting and analyzing frequent formative assessment data, making changes, da ta transparency and safety, technology, and alignment for results. In particular, the Chica go competencies focus on the use of summative data to establish baselines, setting measurable goals, conducting frequent formative assessments, establishing professional learning communities, and making instructional and organizational changes based on formative and summative data (McLeod, 2005). Successful school administrators must also use multiple measures to assess student learning, including the concept of triangulation. Introduction to Data-Driven Decision Making When educational researchers address the c oncept of data-driven decision m aking, they are most often referring to instructional methodologies, curr iculum, and student learning. Educational leaders such as Douglas Reeves, Richard DuFour Michael Schmoker, Robert Marzano, and Richard Stiggins all emphasize using data to establ ish baselines, set goals, conduct assessments, and adjust instructi onal practices in order to achie ve desired results (Reeves, 2003; DuFour, 2003; Schmoker, 2003; and Stiggins, 1999). Data-Driven Decision Making and Instructional Leadership For administrators the focus is on how they function in their role as the instructional leader for the school and how they influence teach ers to use data to refi ne and adjust teaching practices, resulting in improved student performance. Schools mu st set goals with supportive feedback and assessment systems (Marzano, 2003). Teachers and educational leaders must be able to analyze and inte rpret data accurately in order to unde rstand which instructional strategies can address student weaknesses (Schmoker 2003). Schmoker (2005) advocates the use of constructive data analysis from multiple sources, including common assessments. He specifically

PAGE 37

37 suggests analysis of student COHORT data in or der to look at school programs, instructional practices, and curriculum. Schmoker (2006) also advocates the teaching of a concise set of standards taught on a relatively common schedule. Classroom assessments should be used to improve student achievement, self-monitor, reflect those targets that support the standard s, promote student success, encourage student improvement, and use success as a motivation for learning (Stiggins, 2003). By conducting frequent assessments, teachers are able to mon itor student progress and make adjustments to instruction in order to improve student pe rformance (Reeves, 2006). Safer and Fleischman (2003) emphasize the importance of systematic progress monitoring where teachers measure progress towards goals on a reoccurring basis. By establishing baseline data and creating measurable goals we are providing the framework for successful data-driven decision making practices. Through formative assessment, teacher collaboration, and professional learning communities we can monitor progress towards these goals. Schmoker (2003) advocates the use of S.M.A.R.T. goals that are specific, measurable, attainable, results-oriented and tim ely. Marzano (2003) emphasizes the importance of the schools ability to se t goals with supportive feedback and assessment systems. Challenging goals and effective feedback means that a school has a method of assessment that provides detailed information on sp ecific learning goals for specific students on a timely basis. (Marzano, 2003, p. 35). Data-Driven Decision Making an d School Improvement Victoria Bernhardt (1998) propos es first looking at the syst em that produces results, and then making adjustments that focus on conti nuous systemic improvement (p.13). Bernhardt identifies four domains of information that should be collected and analyzed in order to create a complete picture of the school. These four do mains are student demographics, perceptions,

PAGE 38

38 school processes and student learning. For Bernha rdt the emphasis is also on improving student achievement, but she looks at both the broad pi cture of the school and the interrelationships between the four domains. Data-Driven Decision Making and Culture For any change to be effective it m ust beco me anchored in the culture of the school (Kotter, 1996). According to Noyce, Perda, and Traver (2003), data-d riven practices must become a part of the culture. They suggest that when schools us e data to make decisions, there becomes an institutionalized willingness to use numbers systematically to uncover patterns and answer questions about polic y, methods, and outcomes. (p52) A Model for Data-Driven Decision Making Dr. Scott McLeod (2005) at the Universi ty of Minnesota developed a m odel incorporating the use of data-d riven decision making that enco mpasses the work of Michael Schmoker (2003), Richard DuFour (2003), and D ouglas Reeves (2003). According to McLeod (2005), the five essential components of data -driven decision making include establishing baseline goals, setting measurab le instructional goals, conducting ongoing formative assessment, making adjustments to instruction, and implem enting professional learning communities. The model also places an emphasis on data safety an d transparency in the school setting (McLeod, 2005). Another perspective is represented by Panett ieri (2006), who identifies five important means for implementing data-dri ven practices. They include da ta collection, a school wide emphasis on outcome assessments, progress monitoring and feedback, teacher ownership of outcomes, and administrator ability to build a learning organization.

PAGE 39

39 Data-Driven Decision Making Tools Over the past ten years, a wide variety of tools have been developed to support schools in their quest f or data. These tools range from tec hnical tools to ideas a nd practices that support data-driven decision making .Some of the tools are very expensive computer programs while others utilize simple Excel spreadsheets. According to Joe Kitchens (2005), the most important factor is to provide teachers with real time access to student da ta. One must ensure that the technology is integrated in such as way that teachers and administrators can use the data to predict AYP, and make adjustment during the year to intercede. The emphasis should be to leverage data to drive ongoing st udent improvement. (Kitchens, 2005). Communicating Through Data-Driven Dialogue Data-driven dialogue is a tool that can be used facilitate data-driven decision m aking. It focuses on using both quantitative and qualitative student performance data. According to Kathy Dale (2005), when school personnel communicate using data-driven dialogue, it results in collaborative planning and group problem solv ing. The process enables group members to develop a shared understanding of problems by util izing a variety of resources. Leaders and group members analyze data and th en reflect and inspect results to identify problems and create solutions. The emphasis shifts from the actual product, a decision, to the process of using information to make a meaningful decision. Nancy Love (2004) also advocates the use of data-driven dialogue to improve decisionmaking. She recommends that schools create data teams to review teaching and student learning. Administrators should be able to use data-driven dialogue to faci litate collaborat ion and inquiry. The process of data-driven decision making promotes the use of data to provide feedback for continuous school improvement, i nquiry and data literacy.

PAGE 40

40 Extending Data-driven Decision Making to Data Minin g Data mining is the process of analyzing data to determine the relationships between data and school improvement. It includes uncovering patterns and looking at data through multiple levels. Schools that are succe ssful at data mining are gene rally successful at using the information to make decisions. Todd McIntyre (2005) defined three levels fo r schools who engage in data mining. Stage one schools only collect data in or der to meet mandated requirements. Information is stored, but not used. Stage two schools begin to understand how they can benefit from data mining. Schools in this stage begin to use formative assessment in order to support instructional activities. Schools often use information to ta rget those students with the gr eatest educational needs, those whos passing or failing have the greatest influence on the school grade. For example, a school might use the data to monitor its students in the bottom 30%, and make adjustments to resources and teaching methodologies in order to pass federal requirements. McIntyre makes a key point when he states that data analysis that focuse s on improving efficiency works mostly at the edges of the problem, and eventually the school will pull all of the slack out of the system. (McIntyre, 2005). Schools in stage two focus on meeting the immediate needs of groups of students, but are not looking forward to anticipate individual student needs for th e future. As the percentage needed to pass AYP continues to rise, it become s more difficult to anticipate the needs of students without moving to stage three. In stage three, schools are able to use data analysis techniques to support ongoing student achievement. In stage three, schools not only focu s on the needs of student s at the lower quartile, but they are interested in achievement for all students. At this level the school focuses on developing an individual learning profile for each and every student Teachers work with parents and students to work together to set learning go als, and progress against the goals is monitored.

PAGE 41

41 When the student becomes actively involved in the process of progress monitoring, assessment for learning occurs as well. NCLB has caused many schools to focus on grou ps of students, specifically through their subpopulations. Schools in stage three focus on achieving a years worth or more of growth for all students, regardless of subpopul ation or achievement level. A gain in individual student achievement is compared against their expected target, not against th e goal for a subpopulation. Other Data-Driven Decision Making Tools Data-driven decision m aking does not requir e extensive investments in hardware, software, or complex computer systems. It does how ever require a significant investment in time and resources. According to Dr. Scott McLeod (2 005), it is most important to provide teachers with good baseline data. Schools can develop generic data templates that their teachers can use to track student achievement. These templates can be developed using MicroSoft Excel. Dr. McLeod has developed several temp lates, including formative assessments in math, science, and reading; attendance; discipline; and classroom engagement (McLeod, 2005). It is important to note that spreadsheets can also be used for fo rmative assessment at the classroom and student level as well as to analyz e yearly assessment data. The Annenberg Foundation (National Educati onal Association, 2003) has created the Inquiry Cycle to conceptualize data-driven decision making. Stages in this circular system include establishing desired outcomes, defini ng questions, collecting and organizing data, deriving meaning from the data taking action, and evaluating. The North East Florida Educational Consortium (2005) has developed a data analysis tool that can be used to target and improve clas sroom instruction. The Di saggregate Data Assess Review Target (D.A.R.T.) model allows school s to analyze and interpret assessment data, dropping down to analyze perfor mance at the strand level.

PAGE 42

42 Historically, research has questioned the imp act of the principal on student achievement. Research by Philip Halinger and Ronald Heck in 1996 explored research studies that examined the contribution of the princi pal within the schoo l during the period from 1980 through 1995. Results of the research indicated that approximately fort y empirical studies were conducted during the fifteen year period, and that principals do have an e ffect on the school and on student achievement. The result is indirect and small, yet significant. The report also suggested that studies that used a mediated-effects model to ad dress principal leadersh ip influences show a consistent pattern of positive indirect effects (Halinger and Heck, 1996). Many of these studies use multiple regression analysis to identify the interaction effects to measure the strength of indirect effects. Limitations Because app lication of data-driven decision making in schools is so new, there are a limited number of studies that addr ess its use in the public school se tting. The bulk of the studies that have been conducted are qualit ative in nature. The few studies that are empirical in nature do not use a mediated-effects model. They are cons istent with the findings of Halinger and Heck (1996); many of these studies show a small indir ect relationship between principal leadership and student achievement at best. Quantitative research that seeks to understand the impact of data-driven decision making practices by principals on student achievement is even less available. Summary Supovitz and Klein (2003) postula te that only when we use sy stem atic data analysis to support student achievement will we begin to m eet our goal of improved learning outcomes for all students. Educational leaders who truly understand the structur e of schools a nd the needs of our students embrace the use of data-driven de cision making practices. It is important for

PAGE 43

43 educational leaders to understand the importance of using data to make informed decisions that result in improved student achievement. As we move forward into the future, we must continue the emphasis on data-driven decision making, and continuously improve until we enable all schools and children to benefit from its strengths. This chapter has addressed the available research regarding principal leadership and the use of data-driven decision making in schools. Chapter 3 will provide a summary of the met hodology used for the study of the effect of principals beliefs about the us e of data-driven decision making skills on student achievement in elementary schools in Florida.

PAGE 44

44 CHAPTER 3 METHODLOLOGY Introduction Chapter Three provides an overview of the research m ethodology associated with the study. It includes a statement of the problem, the context of the study, the research questions addressed through the study, and the population that was studied. Also di scussed will be the method used for data collection, inst rumentation, and data analysis. The purpose of the study was to examine the relationships among principal characteristics and school demographics, principals beliefs a bout the use of data-dri ven decision making, and student achievement. Specifically th e intent of the dissertation was to determine the mediational effects of data-driven decision making on the rela tionship between principal characteristics and school demographics on student ach ievement. This study is important for several reasons. It adds to the existing body of knowledge about data-driven decision maki ng. It also contributes to the understanding of the relationship be tween a principals beliefs about the use of data in decisionmaking and student achievement. Finally, it seeks to continue the study of the relationship between principal characterist ics and student achievement. Research Questions Question 1: What are the beliefs held by elem entary school principals in Florida with respect to data-driven decision m aking? Question 2: Do Florida elementary school prin cipals beliefs about data-driven decision making mediate the effect of principal charact eristics and school demographics on student achievement? Context of the Study This study represents a census study that addr essed all principals at public elem entary schools within the state of Florida. The data used for the study was co llected during the spring and summer of 2008. The student achievement da ta used for the study included 2008 FCAT SSS

PAGE 45

45 and NRT testing data for Florida elementary sc hool students. Specifically, the study used testing data for reading and math for students in grades three, four, and five. According to the Florida Department of E ducations published list, there are currently 1,890 elementary schools in the state of Florid a (2008). The original list was reviewed to eliminate charter, specialized, and alternative sch ools, and only those schools that teach grades prekindergarten or kindergarten th rough fifth grades were include d in the study. Principals from all schools in all districts that teach prekindergarten or kindergarten through grade five were asked to participate in the study. Brevard and Clay counties were eliminated because all of their elementary schools include sixth grade. Participants The following discussion presents the proced ure used to obtain perm ission to conduct the study and to use the survey inst rument. It also includes a disc ussion of the population and the data collection procedures that we re used as part of the study. Institutional Review Board Procedure and Approval Data for this study were obtained through self -report by principals via a survey on datadriven decision m aking. The survey was based on the Statewide Data-Driven Readiness Study Principal Survey developed by Dr. Scott McLeod at the University of Minnesota (2005). Prior to the initiation of the study, perm ission to use the survey was obtained from Dr. McLeod. The email requesting permission and th e resulting response is included in Appendix A. A request was also sent to the Florida Department of Education requesting residential ad dresses and emails for Florida elementary school prin cipals (Appendix B). Before beginning the research study, approval was obtained from the University of Floridas Institutional Review Board (UFIRB) (Appendix C). The initial contact letter, email, follow up email, informed consent letter, and thank you email are included in Appendices D, E, F, G and H of this document. The follow up

PAGE 46

46 mailing required a different format for the surv ey, and a separate IRB approval was obtained (Appendix I). Finally, an additional email was required for Volusia Count y Principals (Appendix J). Population The initial target population fo r the survey included all principals who work at public elem entary schools in Florida. The preliminary list downloaded from the Department of Education website identified 1,890 elementary sc hools within the 67 distri cts in the state of Florida. Upon review, the list was pared down to eliminate combination, exceptional student, vocational, charter, specialized, and alternative schools. Also e liminated were schools that did not teach prekindergarten or kinde rgarten through grade five. By us ing this model, all schools in Brevard and Clay counties were eliminated. The li st was further pared down to eliminate schools that did not have a principal re sidential or email address. A preliminary letter was sent via the U.S. Postal Service to the remaining 1,478 principals inviting them to participate in the study (Appendix D). A follow up email using SurveyM onkey was sent one week later providing the email link to the web-based survey (Appendix E). Only 1,309 of the emails were received due to the opt out function in SurveyMonkey or undelivera ble addresses. At this time, Broward County administrators indicated that they required ad vanced permission for a study to be conducted within their school district. Although some principals from Broward County responded, the remaining principals were eliminated from the follow up list. Two weeks later, another reminder email (Appendix F) was sent to 1,058 principals. The response rate from the email survey was 415, or 28%. Two weeks later an additional mailing was sent to 928 principals at their home address (Appendices G and K). The response rate from the mailing was 89, giving a total response rate of 504 or 34%. At this time, seve ral surveys were eliminated for a variety of reasons including no school identified, no school data available, incomplete survey (1 question

PAGE 47

47 answered), the respondents worked less than tw o months at the school, or the school was not open during the 2007-2008 school year. The final re sponse rate was 32% or 471 principals. The target rate of 70% for the survey was not met, however this response rate does compare with other dissertation surveys of this magnitude. Principal Respondents Personal Characteristics Of the 471 principals that responded, the m ean time that they spent working as a principal at their school was 4.6 years. The mean time work ed as a principal was 8.34 years. Finally the level of education was 2.54 on a scale of one to four. The level of education ranged from 1 representing a bachelors degree, 2 representing a masters degree, 3 representing an Educational Specialist degree, and 4 represen ting a doctorate degree. The mean enrollment was 669 students. The mean for the percentage of stude nts on free and reduced lunch was 54.6 %. Instrumentation Appendix K provides an exam ple of the survey used in the study. The survey used for the study was comprised of 82 items that address principal beliefs about assessments, acting upon data, support systems, school culture, support sy stems and other demographic information. This survey was selected because the areas identified in the survey most closely align with those elements that have been identified with princi pal beliefs about the use of data-driven decision making. Specifically, these elements address co llecting data, analyzing data, reporting data, using data for school improveme nt, and communicating data (Ame rican Association of School Administrators, 2002). The survey used a six point Likert Scale for the first 77 questions in the survey. Participants were asked to rate their answers as disagree strongly, disagr ee moderately, disagree slightly, agree slightly, agree m oderately, or agree strongly. Th e last six questions used both open-ended and multiple choice questions

PAGE 48

48 The first section of the survey addressed princi pal beliefs about the t ypes of data that are available for use within the school, specifically state assessments, other yearly assessments, common periodic assessments, and other periodic assessments. The second section of the survey addressed beliefs about the use of data-driven decision making by faculty and staff within the school. The third section of the survey asked questions about principal beliefs regarding the support systems within the school, including technology, technical support, resources, and professional development. The fourth section of the survey examined beliefs about the school culture with respect to the use of data to affect studen t achievement and school improvement. The final section of the survey asked for demogr aphic information. It included information about the school, demographic characteristics of the school, and princi pal characteristics. Student Achievement Data The student achievem ent data used for the study was represented by the scale scores for each grade level (grade three, grade four, and grade five) on the FCAT SSS and NRT math and reading tests. In 2005, the state of Florida adop ted the Stanford-10 (SAT-10) Achievement Test Series for norm-referenced testing of elementa ry school students. Flor ida also tests student mastery of the Sunshine State Standards with the Florida Comprehensive Assessment Test. Test data for both tests were used as part of this study. The scale scores by subject, district, school, and grade level were downloaded from the Florida Department of Education website. Respondents whose schools did not report any FCAT data were eliminated from the analysis. Mediational Model This study utilized the mediational m ode l developed by Baron and Kenny (1986). The model proposes the use of mediating variables to determine th e degree to which they can account for the relationship between an antecedent vari able and an outcome variable. The following illustration depicts the relationship between the variables.

PAGE 49

49 Figure 3-1. Mediational M odel (Baron and Kenny, 1986) The Baron and Kenny model utilizes three vari ables, with two paths leading to the outcome (dependent) variable from the predictor variable. Path a represents the relationship between the independent variab le and the mediator. Path b represents the impact of the independent variable on th e mediator. Finally, Path c represents the direct relationship between the antecedent or independent variable and the outcome or dependent variable. According to Baron and Kenny, in order for a variable to function as a mediator, the antecedent or predictor variable should have a significant positive relation ship with the presumed mediator. Variations in the mediator should account for variations in th e outcome variable. Finally when both are held constant, there should be no significant relati onship between the antecedent and the outcome variable. The mediator variable is then considered to be the stronger variable. This study was designed to examine the relatio nship between principal characteristics and school demographic variables, beliefs about the use of data-driven deci sion making, and student achievement. In this study the principal char acteristics and school demographic variables represent the antecedent variab les. Student achievement represents the outcome or dependent variable. Principal beliefs about the use of data-driven decision making represent the mechanism through which the principal characteristics and school demographic vari ables affect student achievement. The model below depicts the Bar on and Kenny Model as ap plied to this study.

PAGE 50

50 Figure 3-2. Proposed Mediational Model The following represents a discussion of the research questions, the associated hypothesis, and the application of the Baron and Kenny model. Question 1 What are the beliefs held by elem entary school principals in Florida with respect to datadriven decision making? Analysis for Question 1 Question 1 was analyzed through descriptive st atistics for principal beliefs about datadriven decision m aking, including each variables measure of central tendency and variance. Question 2 Do Florida elementary school principals beliefs about data-driven decision making mediate the effect of school demographics and pr incipal characteristics on student achievement? Hypothesis 1 Princip al characteristics (years experience, years at a school, and level of education) and school demographics (student enrollment, percentage of students on free and reduced lunch) are related to the principals beliefs about the use of data-driven de cision-making skills.

PAGE 51

51 Null Hypothesis There are no statistically significant differe nces between the principals self-reported beliefs about the use of data -driven decision making skills with varying demographic characteristics (experience, year s in the position, level of edu cation) and school demographics (student enrollment, percentage of students on free and reduced lunch). Analysis of Hypothesis 1 To test this f irst condition of the Baron and Kenny Mediational mode, the relationships among the principal characteristics, school demogr aphics, and principals self-reported beliefs about the use of data driven decision maki ng were analyzed through a series of multiple regression analyses. Principal ch aracteristics and school demographics represent the antecedent variables. Specifically, principa l experience, years at the school, school enrollment, level of education, and percentage of students on free and reduced lunch we re treated as quantitative, continuous variables. The self-reported principals beliefs about data-driven de cision making represented the mediating variable. The structur al model was as follows: E (Y1) = a + b1x1 + b2x2 +b3x3 +b4x4 + b5x5; where Y1= principals beliefs about da ta-driven decision making x1= principal experience, x2= years at the school, x3= level of education, x4= school enrollment, x5= percentage of students on free and reduced lunch. Hypothesis 2 Princip als beliefs about the use of data-d riven decision making are related to student achievement. Null Hypothesis There are no statistically significant differe nces between the principals self-reported beliefs about data-driven decision making and student achievement.

PAGE 52

52 Analysis of Hypothesis 2 The second phase of the m ediational model re quired testing of the relationship between the principals self-reported beliefs about data-driven decision making (the mediating variable) and student achievement (the ou tcome variable) when controlling for the e ffects of principal characteristics and school demographics. Again this was measured through a regression analysis. The estimated structural model was E (Y2) = a + b6Y1 where Y2= student achievement and Y1= principal self-reported use of data-driven decision making skills. Hypothesis 3 Princip als beliefs about the use of data -driven decision making skills mediate the relationship between principal char acteristics (years experience, years at a school, and level of education) and school demographi cs (student enrollment, percentage of students on free and reduced lunch) and student achievement. Null Hypothesis There are no statistically significant differen ces between principals self-reported beliefs about data-driven decision maki ng and student achievement. Analysis of Hypothesis 3 The last condition of the m edi ational model was tested through a two-step regression analysis. Student achievement was regre ssed on principal char acteristics and school demographics. Specifically principal experience, y ears at the school, school enrollment, level of education, and percentage of students on free and reduced lunch we re treated as quantitative, continuous variables. The structural model to be tested for this analysis was as follows: E (Y2) = a + b1x1 + b2x2 +b3x3 _b4x4 + b5x5 +b6x6 +b7Y1; where Y2= student achievement, x1= principal experience, x2= years at the school, x3= level of education, x4= school size, x5= percentage of students on free and reduced lunch and Y1=principal use of data-dri ven decision making skills.

PAGE 53

53 Data Analysis The following analysis was conducted using SPSS statistical software. In addition to the descriptive statistics th at were developed for all variables, each hypothesis was analyzed through the approp riate statistical analysis techniques. This study examined the relationship between the principals beliefs about data-driven decision making and student achievement. The st udy explored the mediational role of datadriven decision making when controlling for the e ffects of principal char acteristics and school demographics. Specific principal ch aracteristics included years of experience, level of education, and years as principal at thei r school. School demographic variables included student enrollment and percentage of students on free and reduced lunch. Principal char acteristics and school demographics represented the antecedent vari ables and student achieve ment represented the outcome variable when studying their relationships Also, the principal ch aracteristics and school demographic variables were treated as antecedent variables when exploring their relationship with the mediating variable, data-driven decisi on making. Finally, multiple regression analyses were used to analyze principal characteristic s, school demographic variables and data-driven decision making as antecedent variables and st udent achievement as the outcome variable. The preliminary analysis using descriptive st atistics for all variables included the means, standard deviations, and frequencies. The analys is followed with a series of multiple regression analysis to determine the relationship between the antecedent variables and the outcome variables. In particular, the analysis focused on the associat ion between data-driven decision making and student achievement, while controlling for school demographics and principal characteristics. In this study, regression coeffici ents were used to analyze the relationships among the principal characteristics, school demographic variables, principa ls use of data-driven decision making and student achievement.

PAGE 54

54 Summary This chapter addressed the research m ethodol ogy used to study the relationship between principal use of data-driven decision making and student ach ievement. It addressed the population, research questions, and survey inst rument. The Baron and Kenny (1986) model was used for this study. The chapter also included the specific data analysis for each research question.

PAGE 55

55 CHAPTER 4 RESULTS AND ANALYSIS OF DATA Introduction Chapter 4 provides a discussion of the results of the study and an analysis of the data. The purpose of the study was to exam ine the relati onship between data-driven decision making by elementary school principals and school level student achievement. The study was designed to test whether the principals beliefs act as a me diator between school demo graphics and principal characteristics and student achievement on the FC AT criterionreferenced and norm-referenced tests. The antecedent variables for principal char acteristics included the number of years as a principal, the number of years at the school, and the level of edu cation. The antecedent variables for school demographics included the enrollment and number of students on free and reduced lunch. The outcome variables for student achieve ment included scales scores for the FCAT Sunshine State Standards (crite rion-referenced test) and the FC AT norm-referenced test (SAT10) for reading and math for grades three, four, and five. Principal self-r eported answers to the data-driven decision making survey were treated as the mediating variable. An analysis of the survey instrument was conducted using a maximum likelihood estimation (MLE) analysis to develop factors and a reliability test for consistency. Descriptive statistics were performed on all variables. Th e mediating relationship of principal beliefs regarding data-driven decision making was exam ined using the mediation model proposed by Baron and Kenny (1986). Linear re gression was performed on the principal characteristics and school demographics, the four factors from the su rvey, and student achievement scales scores for each test, subject, and grade leve l. Finally a two-step regressi on analysis was conducted on the antecedent, mediating, and outcome variables in order to determine whether a mediating effect was present.

PAGE 56

56 The specific research questions and hypothesis presented in the study are as follows: Question 1: What are the beliefs held by elem entary school principals in Florida with respect to data-driven decision making? Question 2: Do Florida elementary school prin cipals beliefs about data-driven decision making mediate the effect of school demographics and princi pal characteristics on student achievement? Hypothesis 1: Principal characte ristics (years experience, year s at a school, and level of education) and school demogra phics (student enrollment, number of students on free and reduced lunch) are related to the principals beliefs about the use of data-driven decisionmaking skills. Hypothesis 2: Principal beliefs about the use of data-driven decision making are related to student achievement. Hypothesis 3: Principals beliefs about the use of data-driven d ecision making skills mediate the relationship between principal char acteristics (years experience, years at a school, and level of education) and school de mographics (student enrollment, number of students on free and reduced lunch) and student achievement. Analysis of the Survey Instrument Because the topic of data-driven d ecision maki ng is relatively new, there are few survey instruments available to address principal belief s on the subject. The State-wide Data Driven Readiness Study Principal Survey was determined to be the best fit for the purpose of the study. A factor analysis was conducted on the instrument after the data collec tion phase was complete in order to measure the ability of the survey to reflect the various beli efs held by elementary school principals and to reduce the data to a smaller set of factor s. The initial survey was pared down to eliminate questions about state assessment, other yearly assessments, common periodic assessments, and other periodic assessments as thes e questions were considered to be outside of the control of the elementary school principal. Fifty-seven of the remaining 63 questions were measured using a six-point Likert scale rang ing from strongly agree to strongly disagree. The remaining five questions asked for the name of the school, the annual enrollment, the

PAGE 57

57 percentage of students on free and reduced lunch, years worked as a principal at the school, years worked as a principal an d level of education. A factor analysis using the MLE model wa s conducted on the 57 questions to extract relevant factors. The in itial correlation matrix was analyzed to eliminat e questions that produced high correlations (greater than .9) and areas wher e there were several significance scores greater than .05. A value of 1.204E-4 for the determinant was developed, indicating that multicollinearity was not a problem. A Kaiser-Meyer-Olkin test was conducte d to measure sampling adequacy. The KMO statistic of .932 was close to a value of 1, indicati ng that the factor analys is should yield distinct and reliable factors. The Bartle tt Test of Sphericity was highl y significant with a Chi-square value of 3121.09 with 231 degrees of freedom. The Varimax with Kaiser Normalization rotation method was used develop orthogonal factors as the questions selected on the survey are assumed to be independent of one another. The analysis of total variance s hows four factors with eigenval ues greater than 1, accounting for nearly 54.8% of the total variance. The Scree pl ot (Figure 4-1), analys is of total variance explained (Table 4-1), and rotated component matrix indicate that between three and four factors underlie the structure of the fiftyseven questions in the survey. As indicated in Table 4-2, the four factors included 23 items that pertained to the underlying structure of the survey. The first factor incorporated nine questions that pertained to principal beliefs regarding the use of data-driven decision making by teachers to improve student achievement. The questions focused on principal be liefs about how teachers use data to affect curriculum, instruction, student performance, and school goals. The second factor contained another six questions that related to principal beliefs about data-d riven cultures. The third factor

PAGE 58

58 included six questions about s upporting systems for data-driven decision-making within the school. These systems include availability and us e of multiple data sour ces, staff meetings, allocating of resources, and professional deve lopment. Although the final factor only included two questions, they were deemed to be important to the study. Factor four supported principal beliefs about collaboration among teacher s using data-driven decision-making. Reliability Analysis Each of the four extracted f actors was analyzed for internal consistency using Cronbachs alpha test. F actors 1, 2, and 4 all tested in the a cceptable range with valu es greater than .7 (.86, .77, and .76 respectively). The third factor, beliefs regarding supporting system s, resulted in an alpha of .4147 which was significantly below the .7 threshold. By eliminating question 70, the alpha was raised to .77, bringing th e third factor into the acceptab le range. Table 4-3 provides the final list of questions and factors used in the study. Analysis and Quantitative Results The analyses perform ed in this study incl uded descriptive statis tics, correlation, linear regression, and two-step multiple regression anal yses. Each question will be presented followed by a discussion of the related hypothesis and stat istical analysis. The initial response rate provided an n of 478; however this rate may vary depe nding upon the statisti cal test performed. The public elementary schools used in the sample did not always participate in both the NRT and SSS for both reading and math at all grade levels. In the multiple regression analysis, a listwise case elimination method was used. Question 1 What are the beliefs held by elem entary school principals in Florida with respect to datadriven decision making?

PAGE 59

59 Analysis The Principal Beliefs About Data-driven D ecision Making survey m easured principal beliefs about data-driven decision making using a Likert scale ranging fr om Strongly Disagree to Strongly Agree (1 to 6) on 22 questions. Each item was analyzed using the variables measure of central tendency and variance. Table 4-4 provides a list of the items in order from the strongest beliefs to the least st rong beliefs held by principals. Each question also provides the mean and variance associated with that particular item or belief. The item means ranged from 5.79 to 4.43 with all means indicating slight, m oderate, or strong agreement with the item number. A further analysis of the items which ranked in the top 10 indicated that 6 of the 10 items related to beliefs about developing a da ta-driven decision maki ng culture. Three items related to beliefs about teachers use of data -driven decision making. The final item supported beliefs regarding collaboration among teachers using data-driven decision making. Table 4-5 provides a list of the top 10 ranked items, in cluding frequencies for each response and the percent valid. The remaining 12 items represented those beliefs that principals felt least strongly about. Of the 12 items in this category, six related to principal beliefs that teachers use data-driven decision making to make decisions regarding st udent achievement, repr esenting 75% of that category. The means for these items ranged fr om 4.43 (slightly ag ree) to 5.14 (agree moderately). Five of the remaining items in this category related to supp orting systems for data driven decision making. These items included use of multiple data sources, teacher input into data management, use of staff meetings to discuss progress, using data for professional development, and allocation of re sources. The remaining item in the least strongly held beliefs category refers to the belief that when teacher s meet with each other, they usually focus on improving student achievement.

PAGE 60

60 Question 2 Do Florida elem entary school principals beliefs about data-driven decision making mediate the effect of principal characteristics and school demographics on student achievement? Analysis The purpose of this study was to determ ine if the principals beliefs regarding data-driven decision making have a mediating effect on st udent achievement. The Baron and Kenny (1986) model presents a framework for testing the e ffect of a mediating variable on the outcome variable. Each of the hypotheses tests a step identified in the Baron and Kenny Model. The first hypothesis tests step 1 in the model. It measures the effect of data-dri ven decision making (the mediating variable) on school de mographics and principal char acteristics (t he antecedent variables). The second hypothesis seeks to measure the relationship between data-driven decision making (the mediating variable) and student achievement (the outcome variable). The last hypothesis regresses student achieve ment (the outcome variable) on both principal characteristics and school demographics (the antecedent variab les) and data-driven decision making (the mediating variable). In order for a mediating effect to be presen t, the regression of the outcome variable on the antecedent vari able should be non-significant. Hypothesis 1 Principal characteristics (years experience, ye ars at a school, and le vel of education) and school demographics (student enrollment, percentage of students on free and reduced lunch) are related to the principals beliefs about th e use of data-driven decision-making skills. Analysis of Hypothesis 1 The first step of the Baron and Kenny (1986) m ediation model requires a test of the relationship between the antecedent variable s (principal characteristics and school demographics) and the mediating variable (principal beliefs regarding data-driven decision

PAGE 61

61 making). In this analysis, the antecedent variab les were treated as quantitative, continuous variables. A regression analysis was conducted, regressing each of the four factors identified through the factor analysis agai nst all of the principal charac teristics and school demographic variables. The four factors were beliefs regarding the use of data-driven decision making by teachers to improve student achievement, belie fs regarding a data-driven culture, beliefs regarding supporting systems, and beliefs regarding collaboration among teachers using datadriven decision making. The antecedent variables te sted were student enro llment, percentage of students on free or reduced lunch, number of years as a principal, number of years as a principal at the school, and the level of education. Tabl e 4-6 provides descriptive statistics for the antecedent variables. None of the four factors proved statistically significant. Fact or 1 (beliefs regarding the use of data-driven decision making by teachers to improve instruction) resulted in an r2 of .011, F (5, 404)=.870, p=.501. Factor 2 (beliefs regarding a data-driven cultu re) showed an r2 of .002, F (5,404)=1.149, p=.334. Factor 3 (beliefs regarding s upporting systems) resulted in an r2 of .022, F (5,404)=1.827, p =.107. Finally, Factor 4 (beliefs rega rding collaboration among teachers using data-driven decision making) indicated an r2 of .020, F (5,404)=1.614, p=.155. All of the r2 values were small, indicating that a very lim ited amount of variability in the outcome can be accounted for by the predictors. The results indicate that the first step in the mediation model did not test significant, and the overall mediati onal model did not fit th is population. Table 4-7 provides the regression analysis for each of the four factors identifie d through the survey. The following represents the estimated structur al model for each of the four factors: Factor 1 Beliefs regarding use of data-d riven decision making by teachers to improve instruction= 46.687 + -.00152 (enrollment) +-.0218 (percentage of students on free and reduced lunch) + .03233 (years as a principal at this school) +-.0266 (years as a principal) +.181 (level of education).

PAGE 62

62 Factor 2 Beliefs regarding a data-dri ven culture=33.875 +-.000987 (enrollment) +-.00188 (percentage of students on free and reduced lunch) + .0214 (years as a principal at this school) +-.0262 (years as a principal) +.255 (level of education). Factor 3 Beliefs regarding supporting sy stems = 24.470 + -.00136 (enrollment) + -.000388 (percentage of students on free and reduced l unch) + -.0269 (years as a principal at this school) +-.0355 (years as a principal) +.503 (level of education). Factor 4 Beliefs regarding collaboration among teachers usin g data-driven decision making = 10.207 + -.000453 (enrollment) + .00505 (percentage of students on free and reduced lunch) + .02459 (years as a principal at this school) +-.0144 (years as a principal) +.05953 (level of education). Hypothesis 2 Princip als beliefs about the use of data-d riven decision making skills are related to student achievement. Analysis of Hypothesis 2 The second condition of the Baron and Kenny m odel tests the relationship between the mediating variable (principals beliefs about the use of data-driven decision making skills) and the outcome variable (student achievement). Fo r the purposes of the st udy, the four factors identified through the factor anal ysis were used to represent da ta-driven decision making skills. Student achievement was measured through th e schools scale scores on the 2008 FCAT SSS (criterion referenced test) and FC AT NRT (norm-referenced test) fo r grades three, four, and five in math and reading. Table 4-8 provides the descriptive statistics for e ach of the variables. The linear regression analysis proved significant for some areas as depicted in Table 4-9. The ANOVA table shows that the F ratio is significant for all areas tested, indicating that a model using data-driven decision making factors does improve the ab ility to predict the outcome variable of student achievement. The r2 value however is quite low, with between 2% and 6% of the variability in the outcome accounted for by the predictors.

PAGE 63

63 The estimated structural model for the data-d riven decision making factors are as follows. Specific parameters are detailed in Table 4-10. SSS Reading Grade 3 =306.177 + 1.151 (beliefs rega rding the use of da ta-driven decision making by teachers) + -.181 (beliefs regardi ng a data-driven culture) + -.210 (beliefs regarding supporting systems) + -3.136 (beliefs regarding collaboration by teachers). SSS Reading Grade 4 =304.321 + 1.057 (beliefs rega rding the use of da ta-driven decision making by teachers) + -.0787 (beliefs regardi ng a data-driven culture) + -.089 (beliefs regarding supporting systems) + -2.758 (beliefs regarding collaboration by teachers). SSS Reading Grade 5 =299.969 + .930 (beliefs rega rding the use of data-driven decision making by teachers) + -.147 (beliefs regardi ng a data-driven culture) + -.589 (beliefs regarding supporting systems) + -1.722 (beliefs regarding collaboration by teachers). SSS Math Grade 3 =321.204 + 1.319 (beliefs rega rding the use of data-driven decision making by teachers) + -.335 (beliefs regardi ng a data-driven culture) + -.166 (beliefs regarding supporting systems) + -3.049 (beliefs regarding collaboration by teachers). SSS Math Grade 4 =300.360 + 1.148 (beliefs rega rding the use of data-driven decision making by teachers) + -.114 (beliefs regardi ng a data-driven culture) + .02377 (beliefs regarding supporting systems) + -2.405 (beliefs regarding collaboration by teachers). SSS Math Grade 5 =320.348 + .917 (beliefs regarding the use of data-driven decision making by teachers) + -.202 (beliefs regardi ng a data-driven culture) + .05947 (beliefs regarding supporting systems) + -2.247 (beliefs regarding collaboration by teachers). NRT Reading Grade 3 =631.006 + .780 (beliefs re garding the use of data-driven decision making by teachers) + -.181 (beliefs regardi ng a data-driven culture) + -.299 (beliefs regarding supporting systems) + -1.950 (beliefs regarding collaboration by teachers). NRT Reading Grade 4 =648.131 + .646 (beliefs re garding the use of data-driven decision making by teachers) + -.0613 (beliefs regardi ng a data-driven culture) + -.164 (beliefs regarding supporting systems) + -1.842 (beliefs regarding collaboration by teachers). NRT Reading Grade 5 =665.878 + .709 (beliefs re garding the use of data-driven decision making by teachers) + -.100 (beliefs regardi ng a data-driven culture) + -.220 (beliefs regarding supporting systems) + -2.023 (beliefs regarding collaboration by teachers). NRT Math Grade 3 =624.711 + 1.022 (beliefs re garding the use of data-driven decision making by teachers) + -.260 (beliefs regardi ng a data-driven culture) + -.314 (beliefs regarding supporting systems) + -2.586 (beliefs regarding collaboration by teachers). NRT Math Grade 4 =651.164 + .340 (beliefs rega rding the use of data-driven decision making by teachers) + -.170 (beliefs regardi ng a data-driven culture) + .06999 (beliefs regarding supporting systems) + -1.766 (beliefs regarding collaboration by teachers).

PAGE 64

64 NRT Math Grade 5 =669.555 + .724 (beliefs rega rding the use of data-driven decision making by teachers) + -.118 (beliefs regardi ng a data-driven culture) + -.668 (beliefs regarding supporting systems) + -1.278 (beliefs regarding collaboration by teachers). Overall, the relationship between principa ls beliefs regarding data-driven decision making and student achievement was found to be st atistically significant. As shown in Table 410, this is particularly true for factors 1 and 4 on both tests for reading and math in all grades, except fourth and fifth grade math NRT. Hypothesis 3 Princip als beliefs about the use of data -driven decision making skills mediate the relationship between principal char acteristics (years experience, years at a school, and level of education) and school demographi cs (student enrollment, percentage of students on Free and Reduced Lunch) on student achievement. Analysis of Hypothesis 3 The last hypothesis regresses student achie vement (the outcome variable) on both principal characteristics and school demographics (the antecedent variables) and data-driven decision making (the mediating vari able). For the mediational model to be effective, the datadriven decision making factors mu st significantly predict student achievement after controlling for principal characteristics and school demographi cs. The third test for mediation was tested via a two-step multiple regression analysis. For the fi rst step in the two-step regression the outcome variable (student achievement) was regressed on all of the anteceden t variables (student enrollment, percentage of students on free and re duced lunch, number of years as a principal, number of years at the school, a nd level of education.) The second step of the regression analysis regresses the outcome variable (student achievement) on the four mediating variables (four datadriven decision making factors). The regression analysis was conduc ted for each of the twelve outcome variables. Table 4-11 provides descriptive statistics for the five antecedent variables and

PAGE 65

65 the four mediating factors. Table 4-12 provides the results of the regression analysis for step one on each outcome variable Each of the models tested significant for the percentage of students on free or reduced lunch for the first step of the analysis, with year s at the school and enroll ment also showing some effect. For the third grade SSS Reading test, the r2 of .7159 for free and reduced lunch and enrollment was statistically significant with F (5,404)=203.64, p=.000. For the fourth grade SSS Reading test, the r2 for free and reduced lunch showed a value of .722, testing significant with F (5,404)=210.15, p=.000. For fifth grade the r2 (free and reduced lunch only) was .306, with an F (5,395)=34.96, p=.000, testing significant. The trend continued for SSS Math with values for r2= .586 ( F (5, 404)=111.56, p=.000), r2=.583 ( F (5,4048)=113.02, p=.000), and r2=.513 F (5,401)=84.66, p=.000) for third, fourth, and fifth grade scores respectively. It is important to note that the model included annual school enro llment along with the percentage on free and reduced lunch for the third and fourth grade models. The NRT scores continued the same trend of including Free and Reduced in all models. School enrollment was included in the models fo r NRT reading in the third and fifth grades. Third grade NRT Reading showed values for r2 equal to .678 ( F (5,403)=169.5, p=.000) and r2=.693 ( F (5,403)=182.42, p=.000), and r2=.684 ( F (5,404)=174.63, p=.000) for third, fourth, and fifth grade scores respectively. Fi nally, the NRT math scores were r2=.655 F (5,404)=53.41, p=.000. Scores for fourth and fifth grade NRT Math returned an r2 of .663 F (5,394)=5.596, p=.000 and r2=.267 F (5,398)=29.06, p=.000 respectively. The other an tecedent variables (school enrollment, years as a principal) accounted fo r much of the variance, causing the two-step regression analysis to eliminate all four data-dri ven decision making factors due to the very small effect sizes. Because the data-driven decision making f actors were not significant, the third

PAGE 66

66 hypothesis is also false, indica ting that the overall mediationa l model does not apply to the population. Table 4-13 provides an overview of the structural components associated with the third hypothesis. Summary of Results The purpose of the analyses performed in this ch apter was to examine the relationships among principal characteristics, school demographics, principals beliefs regarding data-driven decision making, and student achievement as measured by the FCAT SSS and NRT reading and math tests for grades three, four, and five. Speci fically, the purpose of th e study was to determine important principal beliefs rega rding data-driven decision making factors and the relationship between the beliefs and student achievement. The Baron and Kenny (1986) model was used to analyze the relationship between principal char acteristics, school demographics, data-driven decision making and student achievement. The m odel was used to treat data-driven decision making as a mediating variable, such that it could mediate the relationship between the antecedent variables of school demographics an d principal characteristics and the outcome variables of student achievement as measured through the FCAT NRT and SSS math and reading scores for grades three, four, and five. For a medi ating component to be present, three tests must be significant; a relationship between the antece dent variable and the mediating variable, a relationship between the mediating variable and the outcome variable, and a relationship between the mediating variable and the outcome variable while controlling for the antecedent variables. Mediation will be strongest when th ere is no relationship between the antecedent variable and the outcome variable. Four factors were derived us ing a factor analysis study; principal beliefs regarding teacher actions to improve student achievement, principal beliefs regarding a data-driven culture, support systems for data-driven decision making, and collaborati on among teachers using data-

PAGE 67

67 driven decision making. These four factors we re tested using the Baron and Kenny (1986) model. The first test depicting the relationship between the antecedents and mediating variables failed in all twelve tests. The second test of the model, testing the re lationship between the mediating variable and outcome variable, proved significant under all twelve tests. Finally the third test required for mediation failed as well. The failure of two of the three conditions required by the model indicate that the model does not ap ply to this population and that data-driven decision making does not act as a mediator be tween principal characteristics and school demographics and student achievement as meas ured by the FCAT NRT and SSS tests for math and reading for grades th ree, four, and five. However, the findings of the analysis did indi cate that there is an indirect relationship between the four factors identi fied through the factor analysis and student achievement. Two factors proved significant in all tests; belief s regarding teacher use of data-driven decision making by teachers to influence student achieve ment, and collaboration among teachers who use data-driven decision making.

PAGE 68

68 Scree PlotFactor Number21 19 1715 13 1197531Eigenvalue10 8 6 4 2 0 Figure 4-1. Scree Plot for Maximu m Likelihood Estimation Analysis.

PAGE 69

69 Table 4-1. Total Variance Explained Initial Eigenvalues Extraction Sums of Squared Loadings Rotation Sums of Squared Loadings Component Total % of Variance Cumulative % Total % of Variance Cumulative % Total % of Variance Cumulative % 1.00 8.26 35.93 35.93 8.2635.9335.93 4.26 18.53 18.53 2.00 1.53 6.63 42.56 1.536.6342.56 4.03 17.51 36.04 3.00 1.46 6.33 48.89 1.466.3348.89 2.65 11.53 47.57 4.00 1.02 4.44 53.33 1.024.4453.33 1.32 5.76 53.33 5.00 0.98 4.24 57.57 6.00 0.88 3.83 61.40 7.00 0.82 3.57 64.97 8.00 0.80 3.50 68.47 9.00 0.71 3.10 71.57 10.00 0.70 3.04 74.61 11.00 0.65 2.85 77.46 12.00 0.61 2.66 80.12 13.00 0.57 2.47 82.59 14.00 0.53 2.31 84.90 15.00 0.49 2.11 87.01 16.00 0.47 2.04 89.05 17.00 0.44 1.90 90.95 18.00 0.42 1.81 92.76 19.00 0.39 1.70 94.46 20.00 0.38 1.64 96.10 21.00 0.31 1.37 97.47 22.00 0.29 1.28 98.75 23.00 0.29 1.25 100.00

PAGE 70

70 Table 4-2. Rotated Factor Matrix Item Factor 1 Factor 2 Factor 3 Factor 4 Teachers make changes in their instruction based on assessment results. 0.706 0.243 0.225 0.279 If teachers in my school propose a change, they bring data to support their proposal. 0.616 0.176 0.292 0.171 Teachers in this school work collaboratively to improve curriculum and instruction 0.606 0.233 0.035 0.419 Teachers conduct self-assessments to continuously improve performance 0.553 0.180 0.316 0.049 Teachers in my school use data to verify their assumptions about the causes of student behavior and performance 0.496 0.219 0.327 0.168 Teachers in my school use data from student assessment to set instructional targets and goals 0.471 0.371 0.291 0.328 Teachers in my school feel personally responsible when our school improvement goals are not met 0.469 0.148 0.255 0.160 Teachers in my school use assessment data to identify students who are not experiencing academic success 0.398 0.253 0.179 0.382 Teachers and parents communicate frequently about student performance data 0.311 0.232 0.301 0.084 Administrators model data-driven educational practices 0.156 0.785 0.184 0.113 My school adequately supports teachers' use of data to improve classroom instruction 0.196 0.690 0.182 0.054 My school's improvement goals are clear, specific, measurable, and based on student data 0.191 0.479 0.218 0.186 56. As a school we have open and honest discussions about data 0.241 0.453 0.263 0.169 Using data has improved the quality of decision-making in my school 0.197 0.422 0.175 0.204 If we constantly analyze what we do and adjust to get better, we will improve 0.076 0.317 0.029 0.102 Whole-school staff meetings focus on measured progress toward data-based improvement goals 0.090 0.294 0.617 0.204 Teachers have significant input into data management and analysis practices 0.205 0.198 0.601 0.076 Note: "Extraction Method: Maximum Likelihood. Rotati on Method: Varimax with Kaiser Normalization." Rotation converged in 7 iterations.

PAGE 71

71 Table 4-3. Table of Data-Driven Decision Maki ng Factors and Related Survey Questions Construct Number and Description Item Construct 1, Beliefs regarding the use of datadriven decision making by teachers to improve student achievement 22. Teachers in this school work collaboratively to improve curriculum and instruction. 25. Teachers in my school use assessment data to identify students who are not experiencing academic success. 29. Teachers in my school use data to verity their assumptions about the causes of student behavior and performance. 31. If teachers in my school propose a change, they bring data to support their proposal. 32. Teachers in my school make changes in their instruction based on assessment results. 36. Teachers in my school use data from student assessments to set instructional targets and goals. 38. Teachers and parents communicate frequently about student performance data. 67. Teachers conduct self-asse ssments to continuously improve performance. 75. Teachers in my school feel personally responsible when our school improvement goals are not met. Construct 2, Beliefs regarding a data-driven culture 34. My schools improvement goals are clear, specific, measurable, and based on student data. 56. As a school we have open and honest discussions about data. 60. Administrators model data-driven educational practices. 61. My school adequately suppor ts teachers use of data to improve classroom instruction. 71. Using data has improved the quality of decision-making in my school. 74. If we constantly analyze what we do and adjust to get better, we will improve. Construct 3, Beliefs regarding supporting systems 43. My school uses multiple data sources to assess the effectiveness of educational programs. 44. Teachers have significant input into data management and analysis practices. 52. Whole-school staff meetings focus on measured progress toward data-based improvement goals. 53. Student achievement data ar e used to determine teacher professional development needs and resources. 55. Student achievement data are used to determine resource allocation. Construct 4, Beliefs regarding collaboration among teachers using data-driven decision making. 21. Teacher teams in my school meet regularly to look at student data and make instructional plans. 22. When teachers in my school meet with each other, they usually focus on improving student learning outcomes.

PAGE 72

72 Table 4-4. Principal Beliefs Regard ing Data-Driven Decision Making Rank Item N Mean Std. Deviation 1 If we constantly analyze what we do and adjust to get better, we will improve 428 5.79 0.45 2 My school's improvement goals are clear specific, measurable, and based on student data 462 5.70 0.60 3 My school adequately supports teachers' use of data to improve classroom Instruction 428 5.66 0.57 4 As a school we have open and honest discussions about data 430 5.57 0.61 5 Administrators model data-driven educational practices 428 5.55 0.66 6 Teachers in my school use assessment data to identify students who are not experiencing academic success 463 5.54 0.65 7 Using data has improved the quality of decision-making in my school 428 5.45 0.71 8 Teachers in my school use data from student assessment to set instructional targets and goals 458 5.38 0.73 9 Teacher teams in my school meet regularly to look at student data and make instructional plans 463 5.36 0.83 10 Teachers in this school work collabo ratively to improve curriculum and Instruction 463 5.35 0.75 11 My school uses multiple data sour ces to assess the effectiveness of educational programs. 443 5.33 0.79 12 Student achievement data are used to determine teacher professional development needs and resources 374 5.29 0.73 13 Teachers in my school feel person ally responsible when our school improvement goals are not met 430 5.14 0.88 14 Student achievement data are used to determine resource allocation 439 5.14 1.01 15 Whole-school staff meetings focus on measured progress toward data-based improvement goals 439 4.98 0.95 16 When teachers in my school meet w ith each other, they usually focus on improving student learning 463 4.98 0.90 17 Teachers in my school make changes in their instruction based on assessment results. 462 4.95 0.84 18 Teachers and parents communicate frequently ab out student performance data 443 4.81 0.89 19 Teachers in my school use data to verify their assumptions about the causes of student behavior and performance 460 4.78 0.85 20 If teachers in my school propose a change, they bring data to support their proposal. 459 4.63 1.06 21 Teachers have significant inpu t into data management and anal ysis practices 442 4.60 1.05 22 Teachers conduct self-assessments to cont inuously improve performance 420 4.43 1.05 Valid N (listwise) 340 Note: Item responses are: 6= Strongly Agree, 5= Agree Moderately, 4= Agree Slightly, 3= Disagree Slightly, 2= Disagree Moderately, and 1=Disagree Strongly

PAGE 73

73 Table 4-5. Frequency of Responses Regarding Principal Beliefs Regarding Data-Driven Decision Making Rank Item Response Frequency Valid % 1 If we constantly analyze what we do and adjust to get better, we will improve. 6 346 80.84 5 76 17.76 4 5 1.17 3 1 0.23 Total 428 100.00 2 My school's improvement goals ar e clear, specific, measurable, and based on student data. 6 352 76.19 5 83 17.97 4 24 5.19 3 3 0.65 Total 462 100.00 3 My school adequately supports t eachers' use of data to improve classroom instruction. 6 303 70.79 5 105 24.53 4 19 4.44 3 1 0.23 Total 428 100.00 4 As a school we have open and honest discussions about data. 6 274 63.72 5 130 30.23 4 25 5.81 3 1 0.23 Total 430 100.00 5 Administrators model data-driven educational practices. 6 269 62.85 5 129 30.14 4 26 6.07 3 3 0.70 2 1 0.23 Total 428 100.00 6 Teachers in my school use assessment data to identify students who are not experiencing academic success. 6 285 61.56 5 148 31.97 4 26 5.62 3 3 0.65 2 1 0.22 Total 463 100.00 7 Using data has improved the quality of decision-making in my school. 6 234 54.67 5 165 38.55 4 22 5.14 3 4 0.93 2 2 0.47 1 1 0.23 Total 428 100.00 8 Teachers in my school use data from student assessment to set instructional targets and goals. 6 231 50.44 5 176 38.43 4 43 9.39 3 8 1.75 Total 458 100.00

PAGE 74

74 Table 4-5. Continued Rank Item Response Frequency Valid % 9 Teacher teams in my school meet regul arly to look at student data and make instructional plans. 6 249 53.78 5 151 32.61 4 51 11.02 3 7 1.51 2 4 0.86 1 1 0.22 Total 463 100.00 10 Teachers in this school work collaboratively to improve curriculum and instruction. 6 230 49.68 5 176 38.01 4 49 10.58 3 7 1.51 2 1 0.22 Total 463 100.00

PAGE 75

75 Table 4-6. Descriptive Statistics for Antecedent Variables for Hypotheses 1. Variable N Mean Std. Deviation Annual student enrollment 418 688.18 210.84 Percentage of students on Free and Reduced Lunch 425 54.82 24.76 Years as principal at the school 420 4.70 4.04 Years worked as a principal 425 8.39 6.91 Level of education 425 2.52 0.79 Valid N (listwise) 410 Note: For level of education, responses are 1=Ba chelors degree, 2= Masters degree, 3= Ed .Specialist degree, and 4=Doctorate degree.

PAGE 76

76 Table 4-7. Regression Coefficients, Standardized Regression Coeffi cients, t-test Statistics and Partial Correlations for Each of the F our Factors Addressed in Hypothesis 1. Factor Variables b SE t p Partial Corr. Teacher use of data-driven decision making to improve student achievement. (Constant) 46.687 1.633 28.587 0.000 Annual student enrollment -0.020 0.005 -0.058 -1.101 0.271 -0.055 Percentage of students on Free and Reduced Lunch -0.0228 0.012 -0.100 -1.903 0.058 -0.094 Years as a principal at the school 0.032 0.080 0.024 0.403 0.687 0.020 Years as a principal -0.027 0.047 -0.034 -0.563 0.574 -0.028 Level of Education 0.181 0.338 0.027 0.535 0.593 0.027 a. Dependent Variable: Use of data-drive n decision making by teachers to improve instruction Data-driven decision making culture (Constant) 33.875 0.794 42.664 0.000 Annual student enrollment -0.001 0.001 -0.077 -1.474 0.141 -0.073 Percentage of students on Free and Reduced Lunch -0.002 0.006 -0.018 -0.337 0.737 -0.017 Years as a principal at the school 0.021 0.039 0.033 0.549 0.584 0.027 Years as a principal -0.026 0.023 -0.069 -1.142 0.254 -0.057 Level of Education 0.255 0.164 0.077 1.555 0.121 0.077 a. Dependent Variable: Data-driven culture Supporting systems (Constant) 24.470 1.149 21.302 0.000 Annual student enrollment -0.001 0.001 -0.073 -1.405 0.160 -0.070 Percentage of students on Free and Reduced Lunch -0.000 0.008 -0.003 -0.048 0.962 -0.002 Years as a principal at the school -0.027 0.056 -0.029 -0.476 0.635 -0.024 Years as a principal -0.036 0.033 -0.065 -1.069 0.285 -0.053 Level of Education 0.503 0.238 0.105 2.118 0.035 0.105 a. Dependent Variable: A-R factor score 3 for analysis Beliefs regarding supporting systems Collaboration among teachers (Constant) 10.207 0.457 22.330 0.000 Annual student enrollment -0.001 0.000 -0.061 -1.175 0.241 -0.058 Percentage of students on Free and Reduced Lunch 0.005 0.003 0.082 1.573 0.116 0.078 Years as a principal at the school 0.025 0.023 0.066 1.095 0.274 0.054 Years as a principal -0.014 0.013 -0.066 -1.091 0.276 -0.054 Level of Education 0.059 0.095 0.031 0.629 0.529 0.031 Dependent Variable: A-R factor 1 score for analysis Beliefs regarding collaboration among teachers using D3M

PAGE 77

77 Table 4-8. Descriptive Statistic s Regarding Data-Driven Decision Making Factors and FCAT NRT and FCAT SSS Scale Scores for Grad es 3, 4, and 5 in Reading and Math Grade 3 Grade 3 Grade 3 Grade 4 Grade 4 Grade 4 Grade 5 Grade 5 Grade 5 Mean SD N Mean SD N Mean SD N SSS READING 313.93 21.20 443 318. 12 18.97 442 304.42 20.96 433 Teacher use of data-driven decision making 44.52 5.62 443 44.53 5.62 442 44.60 5.60 433 Culture 32.81 5.38 443 32.81 5.39 442 32.82 5.43 433 Supporting systems 24.41 3.80 443 24.42 3.81 442 24.45 3.77 433 Collaboration 10.33 1.55 443 10.33 1.56 442 10.35 1.54 433 SSS MATH 333.36 23.39 443 323.48 21.15 442 332.79 17.84 440 Teacher use of data-driven decision making 44.52 5.62 443 44.53 5.62 442 44.50 5.61 440 Culture 32.81 5.38 443 32.81 5.39 442 32.79 5.40 440 Supporting systems 24.41 3.80 443 24.42 3.81 442 24.40 3.79 440 Collaboration 10.33 1.55 443 10.33 1.56 442 10.33 1.56 440 NRT READING 632.33 14.14 442 651.88 12.37 441 667.87 12.66 440 Teacher use of data-driven decision making 44.52 5.62 442 44.53 5.63 441 44.54 5.62 440 Culture 32.80 5.39 442 32.80 5.39 441 32.86 5.24 440 Supporting systems 24.41 3.81 442 24.43 3.80 441 24.41 3.81 440 Collaboration 10.33 1.55 442 10.33 1.56 441 10.33 1.56 440 NRT MATH 627.30 16.97 443 644.18 14.91 429 668.42 17.71 434 Teacher use of data-driven decision making 44.52 5.62 443 44.53 5.65 429 44.55 5.64 434 Culture 32.81 5.38 443 32.82 5.30 429 32.86 5.27 434 Supporting systems 24.41 3.80 443 24.41 3.82 429 24.40 3.81 434 Collaboration 10.33 1.55 443 10.34 1.55 429 10.33 1.56 434

PAGE 78

78 Table 4-9. Regression Model for Data-Driven Decision Making Factors and FCAT NRT and FCAT SSS Scale Scores for Grades 3, 4, a nd 5 in Reading and Math for Hypothesis 2. Variable df F p r2 SSS READ Grade 3 4, 438 6.411 0.000 0.055 SSS READ Grade 4 4, 437 7.307 0.000 0.063 SSS READ Grade 5 4, 428 3.511 0.008 0.032 SSS MATH Grade 4 4, 437 6.747 0.000 0.058 SSS MATH Grade 5 4, 435 5.689 0.000 0.050 NRT READ Grade 3 4, 437 5.882 0.000 0.051 NRT READ Grade 4 4, 436 6.358 0.000 0.055 NRT READ Grade 5 4, 435 7.099 0.000 0.032 NRT MATH Grade 3 4, 438 7.051 0.000 0.061 NRT MATH Grade 4 4, 424 2.509 0.041 0.023 NRT MATH Grade 5 4, 429 3.455 0.009 0.031

PAGE 79

79 Table 4-10. Unstandardized Regr ession Coefficients, Standardi zed Regression Coefficients, ttest Statistics, and Partial Correlations for Hypothesis 2. B SE t p Partial SSS READING Grade 3 (Constant) 306.177 9.052 33.826 0.000 Use by teachers to improve instruction 1.151 0.270 0.305 4.262 0.000 0.200 Data-driven culture -0.181 0.219 -0.046 -0.827 0.409 -0.039 Supporting systems -0.210 0.310 -0.038 -0.677 0.499 -0.032 Collaboration among teachers -3. 136 0.780 -0.230 -4.023 0.000 -0.189 SSS READING Grade 4 (Constant) 304.321 8.069 37.716 0.000 Use by teachers to improve instruction 1.057 0.241 0.313 4.392 0.000 0.206 Data-driven culture -0.079 0.195 -0.022 -0.403 0.687 -0.019 Supporting systems -0.089 0.276 -0.018 -0.322 0.747 -0.015 Collaboration among teachers -2. 758 0.695 -0.226 -3.966 0.000 -0.186 SSS READING Grade 5 (Constant) 299.969 9.201 32.602 0.000 Use by teachers to improve instruction 0.930 0.272 0.249 3.418 0.001 0.163 Data-driven culture -0.147 0.220 -0.038 -0.669 0.504 -0.032 Supporting systems -0.589 0.315 -0.106 -1.867 0.063 -0.090 Collaboration among teachers -1. 722 0.796 -0.127 -2.164 0.031 -0.104 SSS MATH Grade 3 (Constant) 321.204 10.000 32.121 0.000 Use by teachers to improve instruction 1.319 0.298 0.317 4.421 0.000 0.207 Data-driven culture -0.335 0.242 -0.077 -1.384 0.167 -0.066 Supporting systems -0.166 0.342 -0.027 -0.484 0.629 -0.023 Collaboration among teachers -3. 049 0.861 -0.203 -3.541 0.000 -0.167 SSS MATH Grade 4 (Constant) 300.360 9.018 33.307 0.000 Use by teachers to improve instruction 1.148 0.269 0.305 4.269 0.000 0.200 Data-driven culture -0.114 0.218 -0.029 -0.521 0.602 -0.025 Supporting systems 0.024 0.309 0.004 0.077 0.939 0.004 Collaboration among teachers -2. 405 0.777 -0.177 -3.094 0.002 -0.146 SSS MATH Grade 5 (Constant) 320.348 7.664 41.800 0.000 Use by teachers to improve instruction 0.917 0.228 0.289 4.020 0.000 0.189 Data-driven culture -0.202 0.185 -0.061 -1.090 0.276 -0.052 Supporting systems 0.059 0.262 0.013 0.227 0.821 0.011 Collaboration among teachers -2. 247 0.659 -0.196 -3.409 0.001 -0.161 NRT READ Grade 3 (Constant) 631.006 6.051 104.282 0.000 Use by teachers to improve instruction 0.780 0.180 0.310 4.321 0.000 0.202 Data-driven culture -0.181 0.147 -0.069 -1.235 0.218 -0.059 Supporting systems -0.299 0.207 -0.081 -1.445 0.149 -0.069 Collaboration among teachers -1. 950 0.522 -0.214 -3.737 0.000 -0.176

PAGE 80

80 Table 4-10. Continued B SE t p Partial NRT READ Grade 4 (Constant) 648.131 5.283 122.686 0.000 Use by teachers to improve Instruction 0.646 0.158 0.294 4.100 0.000 0.193 Data-driven culture -0.061 0.128 -0.027 -0.478 0.633 -0.023 Supporting systems -0.164 0.181 -0.050 -0.904 0.366 -0.043 Collaboration among teachers -1. 842 0.456 -0.232 -4.039 0.000 -0.190 NRT READ Grade 5 (Constant) 665.878 5.430 122.636 0.000 Use by teachers to improve Instruction 0.709 0.161 0.314 4.406 0.000 0.207 Data-driven culture -0.100 0.134 -0.042 -0.748 0.455 -0.036 Supporting systems -0.220 0.185 -0.066 -1.191 0.234 -0.057 Collaboration among teachers -2. 023 0.465 -0.249 -4.355 0.000 -0.204 NRT MATH Grade 3 (Constant) 624.711 7.224 86.478 0.000 Use by teachers to improve Instruction 1.022 0.215 0.338 4.743 0.000 0.221 Data-driven culture -0.260 0.175 -0.082 -1.483 0.139 -0.071 Supporting systems -0.314 0.247 -0.070 -1.271 0.204 -0.061 Collaboration among teachers -2. 586 0.622 -0.237 -4.156 0.000 -0.195 NRT MATH Grade 4 (Constant) 651.164 6.557 99.312 0.000 Use by teachers to improve Instruction 0.340 0.195 0.129 1.746 0.082 0.084 Data-driven culture -0.170 0.162 -0.060 -1.050 0.295 -0.051 Supporting systems 0.070 0.225 0.018 0.311 0.756 0.015 Collaboration among teachers -1. 766 0.569 -0.184 -3.106 0.002 -0.149 NRT MATH Grade 5 (Constant) 669.555 7.744 86.466 0.000 Use by teachers to improve Instruction 0.724 0.229 0.230 3.157 0.002 0.151 Data-driven culture -0.118 0.191 -0.035 -0.617 0.537 -0.030 Supporting systems -0.668 0.264 -0.144 -2.528 0.012 -0.121 Collaboration among teachers -1. 278 0.663 -0.113 -1.928 0.055 -0.093

PAGE 81

81 Table 4-11. Descriptive Statistics Regarding Principal Characteri stics, School Demographics, and Data-Driven Decision Making Factor s and FCAT NRT and FCAT SSS Scale Scores for Grades 3, 4, and 5 in Reading and Math Grade 3 Grade 3 Grade 3 Grade 4 Grade 4 Grade 4 Grade 5 Grade 5 Grade 5 Mean SD N Mean SD N Mean SD N SSS READING 314.40 21.05 410 318. 55 18.99 410 304.92 20.86 401 Annual enrollment 688.50 205.69 410 688.50 205.69 410 689.44 207.41 401 Percentage on FRL 54.81 24.80 4 10 54.81 24.80 410 54.88 24.67 401 Years at the school 4.70 4.07 410 4.70 4.07 410 4.72 4.08 401 Years as a principal 8.39 6.96 410 8.39 6.96 410 8.43 6.91 401 Level of education 2.53 0.79 410 2.53 0.79 410 2.54 0.80 401 Use by teachers to improve instruction 44.83 5.41 410 44.83 5.41 410 44.92 5.37 401 Data-driven culture 33.62 2.63 410 33.62 2.63 410 33.65 2.63 401 Supporting systems 24.36 3.82 410 24.36 3.82 410 24.39 3.79 401 Collaboration among teachers 10.32 1.52 410 10.32 1.52 410 10.34 1.50 401 SSS MATH 333.69 23.31 410 323.83 21.16 410 333.00 17.74 407 Annual enrollment 688.50 205.69 410 688.50 205.69 410 688.36 206.34 407 Percentage on FRL 54.81 24.80 4 10 54.81 24.80 410 54.83 24.81 407 Years at the school 4.70 4.07 410 4.70 4.07 410 4.69 4.06 407 Years as a principal 8.39 6.96 410 8.39 6.96 410 8.35 6.90 407 Level of education 2.53 0.79 410 2.53 0.79 410 2.53 0.79 407 Use by teachers to improve instruction 44.83 5.41 410 44.83 5.41 410 44.81 5.40 407 Data-driven culture 33.62 2.63 410 33.62 2.63 410 33.61 2.64 407 Supporting systems 24.36 3.82 410 24.36 3.82 410 24.36 3.81 407 Collaboration among teachers 10.32 1.52 410 10.32 1.52 410 10.31 1.52 407 NRT READING 632.63 14.08 409 652.12 12.39 409 668.20 12.57 409 Annual enrollment 688.05 205.74 409 688.77 205.86 409 688.47 205.94 409 Percentage on FRL 54.79 24.83 4 09 54.76 24.80 409 54.80 24.83 409 Years at the school 4.70 4.08 409 4.65 3.98 409 4.68 4.06 409 Years as a principal 8.39 6.97 409 8.35 6.93 409 8.34 6.89 409 Level of education 2.53 0.80 409 2.53 0.79 409 2.53 0.80 409 Use by teachers to improve instruction 44.83 5.41 409 44.84 5.41 409 44.82 5.41 409 Data-driven culture 33.61 2.63 409 33.61 2.63 409 33.61 2.63 409 Supporting systems 24.36 3.83 409 24.37 3.82 409 24.35 3.83 409 Collaboration among teachers 10.31 1.52 409 10.32 1.52 409 10.32 1.52 409 NRT MATH 627.61 16.95 410 644.32 15.07 400 668.73 17.80 404 Annual enrollment 688.50 205.69 410 687.15 207.01 400 689.64 206.63 404 Percentage on FRL 54.81 24.80 4 10 54.84 24.92 400 54.62 24.78 404 Years at the school 4.70 4.07 410 4.66 3.99 400 4.65 3.97 404 Years as a principal 8.39 6.96 410 8.30 6.91 400 8.34 6.87 404 Level of education 2.53 0.79 410 2.53 0.79 400 2.54 0.80 404 Use by teachers to improve instruction 44.83 5.41 410 44.85 5.43 400 44.84 5.42 404 Data-driven culture 33.62 2.63 410 33.60 2.66 400 33.62 2.64 404 Supporting systems 24.36 3.82 410 24.38 3.84 400 24.35 3.83 404 Collaboration among teachers 10.32 1.52 410 10.34 1.51 400 10.32 1.52 404

PAGE 82

82 Table 4-12. ANOVA Table for Data-Driven Deci sion Making Factors and FCAT NRT and FCAT SSS Scale Scores for Grades 3, 4, and 5 in Reading and Math While Contro lling for Principal Charac teristics and School Demogr aphics for Step One and Two of Hypothesis 3. SSS Reading Model Grade 3 df F Sig. Model Grade 4 df F Sig. Model Grade 5 df F Sig. 1 Regression 5 203.64 0.000 1 Regression 5 210.15 0.000 1 Regression 5 34.97 0.000 Residual 404 Residual 404 Residual 395 Total 409 Total 409 Total 400 2 Regression 9 113.88 0.000 2 Regression 9 120.82 0.000 2 Regression 9 19.58 0.000 Residual 400 Residual 400 Residual 391 Total 409 Total 409 Total 400 a Predictors: (Constant), level of education, years at the school a Predictors: (Constant), level of education, years at the school a Predictors: (Constant), level of education, annual student enrollment b Predictors: (Constant), level of education, years at the school, percentage on FRL b Predictors: (Constant), level of education, years at the school, percentage on FRL b Predictors: (Constant), level of education, annual student enrollment, years the school, percentage on FRL c Dependent Variable: SSS Reading Grade 3 c Dependent Variable: SSS Reading Grade 4 c Dependent Variable: SSS Reading Grade 5 SSS Math Model Grade 3 df F Sig. Model Grade 4 df F Sig. Model Grade 5 df F Sig. 1 Regression 5 111.56 0.000 1 Regression 5 113.02 0.000 1 Regression 5 84.67 0.000 Residual 404 Residual 404 Residual 401 Total 409 Total 409 Total 406 2 Regression 9 63.04 0.000 2 Regression 9 66.23 0.000 2 Regression 9 48.02 0.000 Residual 400 Residual 400 Residual 397 Total 409 Total 409 Total 406 a Predictors: (Constant), level of education, years at this school a Predictors: (Constant), level of education, years at the school a Predictors: (Constant), level of education, percentage on FRL b Predictors: (Constant), level of education, years at the school percentage on FRL b Predictors: (Constant), level of education, years at the school, percentage on FRL b Predictors: (Constant), level of education, percentage on FRL, years at the school c Dependent Variable: SSS Math Grade 3 c Dependent Vari able: SSS Math Grade 4 c Dependent Variable: SSS Math Grade 5

PAGE 83

83 Table 4-12. Continued NRT Reading Model Grade 3 df F Sig. Model Grade 4 df F Sig. Model Grade 5 df F Sig. 1 Regression 5 169.87 0.000 1 Regression 5 182.42 0.000 1 Regression 5 174.6 3 0.000 Residual 403 Residual 403 Residual 403 Total 408 Total 408 Total 408 2 Regression 9 94.28 0.000 2 Regression 9 102.09 0.000 2 Regression 9 98.40 0.000 Residual 399 Residual 399 Residual 399 Total 408 Total 408 Total 408 A Predictors: (Constant), level of education, annual student enrollment a Predictors: (Constant), level of education, annual student enrollment a Predictors: (Constant), level of education, annual student enrollment B Predictors: (Constant), level of education, annual student enrollm ent, years at the school b Predictors: (Constant), level of education, annual student enrollm ent, years at the school b Predictors: (Constant), level of education, annual student enrollment, years at the school C Dependent Variable: NRT Read Grade 3 c Dependent Variable: NRT Read Grad e4 c Dependent Variable: NRT Read Grade 5 NRT Math Model Grade 3 df F Sig. Model Grade 4 df F Sig. Model Grade 5 df F Sig. 1 Regression 5 153.41 0.000 1 Regression 5 5.60 0.000 1 Regression 5 29.06 0.000 Residual 404 Residual 394 Residual 398 Total 409 Total 399 Total 403 2 Regression 9 86.15 0.000 2 Regression 9 3.96 0.000 2 Regression 9 16.48 0.000 Residual 400 Residual 390 Residual 394 Total 409 Total 399 Total 403 A Predictors: (Constant), level of education, years at the school a Predictors: (Constant), level of education, percentage on FRL a Predictors: (Constant), level of education, years at the school B Predictors: (Constant), level of education, years at the school, percentage on FRL b Predictors: (Constant), level of education, percentage on FRL, years at the school b Predictors: (Constant), level of education, years at the school, percentage on FRL C Dependent Variable: NRT Math Grade 3 c Dependent Variable: NRT Math Grad e 4 c Dependent Variable: NRT Math Grade 5

PAGE 84

84 Table 4-13. Regression Model for Data-Driven Decision Making Factors and FCAT NRT and FCAT SSS Scale Scores for Grades 3, 4, and 5 in Reading and Math While Controlling for Principal Characteristics a nd School Demographics for Step One of Hypothesis 3. Model Variables b SE t p Partial SSS Reading Grade 3 1 (Constant) 361.18 3.41 105.99 0.000 Annual school enrollment -0.01 0.00 -0.08-3.01 0.003-0.15 Percentage on FRL -0.73 0.02 -0.86-30.50 0.000-0.83 Years at the school 0.20 0.17 0.041.22 0.2220.06 Years as a principal 0.13 0.10 0.041.34 0.1810.07 Level of education -1.13 0.71 -0.04-1.61 0.109-0.08 2 (Constant) 357.09 8.05 44.35 0.000 Annual school enrollment -0.01 0.00 -0.08-2.92 0.004-0.14 Percentage on FRL -0.72 0.02 -0.85-29.67 0.000-0.83 Years at the school 0.22 0.17 0.041.32 0.1870.07 Years as a principal 0.13 0.10 0.041.36 0.1730.07 Level of education -1.23 0.71 -0.05-1.73 0.084-0.09 Use of D3M by teachers to improve instruction 0.14 0.16 0.040.91 0.3640.05 Data-driven culture -0.11 0.27 -0.01-0.42 0.672-0.02 Supporting systems 0.26 0.18 0.051.44 0.1520.07 Collaboration among teachers using D3M -0.50 0.47 -0.04-1.07 0.284-0.05 a Dependent Variable: SSS Reading Grade 3 SSS Reading Grade 4 1 (Constant) 355.58 3.04 117.01 0.000 Annual school enrollment 0.00 0.00 -0.05-1.91 0.057-0.09 Percentage on FRL -0.66 0.02 -0.86-30.85 0.000-0.84 Years at the school 0.21 0.15 0.051.41 0.1580.07 Years as a principal 0.05 0.09 0.020.58 0.5630.03 Level of education 0.39 0.63 0.020.63 0.5310.03 2 (Constant) 339.83 7.11 47.81 0.000 Annual school enrollment 0.00 0.00 -0.05-1.69 0.092-0.08 Percentage on FRL -0.65 0.02 -0.85-30.22 0.000-0.83 Years at the school 0.22 0.15 0.051.49 0.1370.07 Years as a principal 0.06 0.09 0.020.72 0.4740.04 Level of education 0.21 0.63 0.010.34 0.7330.02 Use of D3M by teachers to improve instruction 0.20 0.14 0.061.42 0.1560.07 Data-driven culture 0.14 0.23 0.020.61 0.5450.03 Supporting systems 0.27 0.16 0.051.71 0.0880.09 Collaboration among teachers using D3M -0.49 0.41 -0.04-1.18 0.238-0.06 a Dependent Variable: SSS Reading Grade 4

PAGE 85

85 Table 4-13. Continued Model b SE t P Partial SSS Reading Grade 5 1 (Constant) 333.01 5.35 62.21 0.000 Annual school enrollment 0.00 0.00 0.010.33 0.7400.02 Percentage on FRL -0.46 0.04 -0.54-12.18 0.000-0.52 Years at the school 0.22 0.26 0.040.86 0.3920.04 Years as a principal -0.04 0.15 -0.01-0.23 0.818-0.01 Level of education -1.83 1.10 -0.07-1.67 0.097-0.08 2 (Constant) 327.27 12.82 25.52 0.000 Annual school enrollment 0.00 0.00 0.020.35 0.7230.02 Percentage on FRL -0.45 0.04 -0.54-11.72 0.000-0.51 Years at the school 0.21 0.26 0.040.80 0.4240.04 Years as a principal -0.04 0.16 -0.01-0.23 0.817-0.01 Level of education -1.72 1.11 -0.07-1.56 0.120-0.08 Use of D3M by teachers to improve instruction 0.30 0.25 0.081.21 0.2290.06 Data-driven culture -0.08 0.42 -0.01-0.19 0.850-0.01 Supporting systems -0.25 0.28 -0.05-0.88 0.382-0.04 Collaboration among teachers using D3M 0.04 0.75 0.000.06 0.9560.00 a Dependent Variable: SSS Reading Grade 5 SSS Math Grade 3 1 (Constant) 371.48 4.59 80.96 0.000 Annual school enrollment 0.00 0.00 -0.01-0.43 0.671-0.02 Percentage on FRL -0.72 0.03 -0.76-22.25 0.000-0.74 Years at the school 0.54 0.23 0.092.39 0.0170.12 Years as a principal -0.12 0.13 -0.03-0.87 0.385-0.04 Level of education 0.43 0.95 0.010.45 0.6540.02 2 (Constant) 365.91 10.82 33.81 0.000 Annual school enrollment 0.00 0.00 -0.01-0.33 0.740-0.02 Percentage on FRL -0.71 0.03 -0.75-21.63 0.000-0.73 Years at the school 0.55 0.23 0.102.44 0.0150.12 Years as a principal -0.11 0.13 -0.03-0.83 0.405-0.04 Level of education 0.34 0.95 0.010.36 0.7180.02 Use of D3M by teachers to improve instruction 0.31 0.21 0.071.48 0.1410.07 Data-driven culture -0.41 0.36 -0.05-1.14 0.256-0.06 Supporting systems 0.28 0.24 0.051.16 0.2480.06 Collaboration among teachers using D3M -0.21 0.63 -0.01-0.34 0.735-0.02 a Dependent Variable: SSS Math Grade 3

PAGE 86

86 Table 4-13. Continued Model b SE t P Partial SSS Math Grade 4 1 (Constant) 357.23 4.15 86.09 0.000 Annual school enrollment 0.00 0.00 -0.02-0.69 0.493-0.03 Percentage on FRL -0.66 0.03 -0.77-22.50 0.000-0.75 Years at the school 0.43 0.20 0.082.12 0.0350.10 Years as a principal -0.15 0.12 -0.05-1.24 0.216-0.06 Level of education 1.35 0.86 0.051.58 0.1160.08 2 (Constant) 331.96 9.68 34.29 0.000 Annual school enrollment 0.00 0.00 -0.01-0.41 0.684-0.02 Percentage on FRL -0.65 0.03 -0.76-22.12 0.000-0.74 Years at the school 0.43 0.20 0.082.14 0.0330.11 Years as a principal -0.13 0.12 -0.04-1.07 0.287-0.05 Level of education 1.10 0.85 0.041.29 0.1990.06 Use of D3M by teachers to improve instruction 0.27 0.19 0.071.43 0.1520.07 Data-driven culture 0.14 0.32 0.020.44 0.6600.02 Supporting systems 0.34 0.22 0.061.57 0.1170.08 Collaboration among teachers using D3M -0.06 0.56 0.00-0.10 0.9210.00 a Dependent Variable: SSS Math Grade 4 SSS Math Grade 5 1 (Constant) 361.71 3.76 96.10 0.000 Annual school enrollment 0.00 0.00 -0.05-1.46 0.145-0.07 Percentage on FRL -0.52 0.03 -0.73-19.67 0.000-0.70 Years at the school 0.15 0.18 0.030.82 0.4150.04 Years as a principal 0.01 0.11 0.000.10 0.9200.01 Level of education 0.88 0.78 0.041.13 0.2600.06 2 (Constant) 355.81 8.89 40.01 0.000 Annual school enrollment 0.00 0.00 -0.05-1.36 0.174-0.07 Percentage on FRL -0.51 0.03 -0.72-19.00 0.000-0.69 Years at the school 0.17 0.18 0.040.90 0.3710.04 Years as a principal 0.02 0.11 0.010.16 0.8700.01 Level of education 0.75 0.79 0.030.95 0.3410.05 Use of D3M by teachers to improve instruction 0.23 0.17 0.071.32 0.1880.07 Data-driven culture -0.20 0.29 -0.03-0.68 0.494-0.03 Supporting systems 0.29 0.20 0.061.46 0.1460.07 Collaboration among teachers using D3M -0.51 0.52 -0.04-0.98 0.329-0.05 a Dependent Variable: SSS Math Grade 5

PAGE 87

87 Table 4-13. Continued Model b SE t P Partial NRT Reading Grade 3 1 (Constant) 660.92 2.43 272.04 0.000 Annual school enrollment 0.00 0.00 -0.06-2.11 0.035-0.10 Percentage on FRL -0.47 0.02 -0.83-27.67 0.000-0.81 Years at the school 0.05 0.12 0.010.40 0.6890.02 Years as a principal 0.13 0.07 0.061.84 0.0660.09 Level of education -0.30 0.50 -0.02-0.61 0.545-0.03 2 (Constant) 664.51 5.76 115.32 0.000 Annual school enrollment 0.00 0.00 -0.07-2.16 0.031-0.11 Percentage on FRL -0.47 0.02 -0.83-26.85 0.000-0.80 Years at the school 0.05 0.12 0.020.45 0.6560.02 Years as a principal 0.12 0.07 0.061.76 0.0790.09 Level of education -0.26 0.51 -0.01-0.51 0.609-0.03 Use of D3M by teachers to improve instruction 0.14 0.11 0.051.24 0.2160.06 Data-driven culture -0.24 0.19 -0.05-1.27 0.203-0.06 Supporting systems 0.01 0.13 0.000.04 0.9660.00 Collaboration among teachers using D3M -0.20 0.34 -0.02-0.60 0.551-0.03 a Dependent Variable: NRT Read Grade 3 NRT Reading Grade 4 1 (Constant) 675.42 2.08 323.97 0.000 Annual school enrollment 0.00 0.00 -0.03-1.17 0.242-0.06 Percentage on FRL -0.42 0.01 -0.84-28.56 0.000-0.82 Years at the school 0.09 0.10 0.030.89 0.3760.04 Years as a principal 0.05 0.06 0.030.78 0.4380.04 Level of education 0.08 0.43 0.010.19 0.8460.01 2 (Constant) 670.35 4.93 135.99 0.000 Annual school enrollment 0.00 0.00 -0.03-1.09 0.275-0.05 Percentage on FRL -0.41 0.01 -0.83-27.69 0.000-0.81 Years at the school 0.10 0.10 0.030.95 0.3440.05 Years as a principal 0.05 0.06 0.030.79 0.4300.04 Level of education 0.04 0.43 0.000.09 0.9320.00 Use of D3M by teachers to improve instruction 0.11 0.10 0.051.12 0.2650.06 Data-driven culture 0.10 0.16 0.020.60 0.5460.03 Supporting systems 0.06 0.11 0.020.55 0.5820.03 Collaboration among teachers using D3M -0.47 0.29 -0.06-1.64 0.102-0.08 a Dependent Variable: NRT Read Grade 4

PAGE 88

88 Table 4-13. Continued Model b SE t P Partial NRT Reading Grade 5 1 (Constant) 694.32 2.15 323.14 0.000 Annual school enrollment 0.00 0.00 -0.07-2.22 0.027-0.11 Percentage on FRL -0.43 0.02 -0.84-28.32 0.000-0.82 Years at the school -0.01 0.11 0.00-0.10 0.920-0.01 Years as a principal 0.05 0.06 0.020.73 0.4670.04 Level of education -0.11 0.44 -0.01-0.25 0.801-0.01 2 (Constant) 699.76 5.07 138.00 0.000 Annual school enrollment 0.00 0.00 -0.07-2.33 0.020-0.12 Percentage on FRL -0.42 0.02 -0.83-27.44 0.000-0.81 Years at the school 0.00 0.11 0.000.04 0.9650.00 Years as a principal 0.04 0.06 0.020.58 0.5600.03 Level of education -0.07 0.45 0.00-0.15 0.879-0.01 Use of D3M by teachers to improve instruction 0.17 0.10 0.071.68 0.0930.08 Data-driven culture -0.25 0.17 -0.05-1.51 0.131-0.08 Supporting systems 0.05 0.11 0.010.42 0.6750.02 Collaboration among teachers using D3M -0.57 0.30 -0.07-1.94 0.053-0.10 a Dependent Variable: NRT Read Grade 5 NRT Math Grade 3 1 (Constant) 659.40 3.02 218.11 0.000 Annual school enrollment 0.00 0.00 -0.02-0.73 0.463-0.04 Percentage on FRL -0.56 0.02 -0.82-26.30 0.000-0.79 Years at the school 0.17 0.15 0.041.12 0.2620.06 Years as a principal -0.06 0.09 -0.03-0.70 0.485-0.03 Level of education -0.06 0.63 0.00-0.10 0.917-0.01 2 (Constant) 664.31 7.14 93.07 0.000 Annual school enrollment 0.00 0.00 -0.02-0.79 0.427-0.04 Percentage on FRL -0.55 0.02 -0.81-25.46 0.000-0.79 Years at the school 0.18 0.15 0.041.20 0.2320.06 Years as a principal -0.07 0.09 -0.03-0.80 0.424-0.04 Level of education 0.01 0.63 0.000.01 0.9910.00 Use of D3M by teachers to improve instruction 0.28 0.14 0.091.98 0.0490.10 Data-driven culture -0.40 0.24 -0.06-1.72 0.087-0.09 Supporting systems 0.01 0.16 0.000.09 0.9270.00 Collaboration among teachers using D3M -0.44 0.42 -0.04-1.06 0.288-0.05 a Dependent Variable: NRT Math Grade 3

PAGE 89

89 Table 4-13. Continued Model b SE t P Partial NRT Math Grade 4 1 (Constant) 651.01 4.48 145.35 0.000 Annual school enrollment 0.00 0.00 0.020.33 0.7410.02 Percentage on FRL -0.13 0.03 -0.22-4.22 0.000-0.21 Years at the school -0.17 0.22 -0.04-0.74 0.460-0.04 Years as a principal 0.26 0.13 0.122.03 0.0430.10 Level of education -0.69 0.93 -0.04-0.74 0.459-0.04 2 (Constant) 662.95 10.48 63.26 0.000 Annual school enrollment 0.00 0.00 0.010.26 0.7930.01 Percentage on FRL -0.12 0.03 -0.20-3.84 0.000-0.19 Years at the school -0.13 0.22 -0.04-0.59 0.555-0.03 Years as a principal 0.25 0.13 0.111.90 0.0580.10 Level of education -0.65 0.93 -0.03-0.69 0.488-0.04 Use of D3M by teachers to improve instruction 0.23 0.21 0.081.10 0.2710.06 Data-driven culture -0.52 0.35 -0.09-1.50 0.134-0.08 Supporting systems 0.29 0.24 0.071.25 0.2130.06 Collaboration among teachers using D3M -1.19 0.62 -0.12-1.92 0.055-0.10 a Dependent Variable: NRT Math Grade 4 NRT Math Grade 5 1 (Constant) 688.97 4.67 147.61 0.000 Annual school enrollment 0.00 0.00 0.051.16 0.2470.06 Percentage on FRL -0.35 0.03 -0.49-10.70 0.000-0.47 Years at the school 0.27 0.23 0.061.16 0.2480.06 Years as a principal -0.07 0.13 -0.03-0.51 0.610-0.03 Level of education -1.92 0.96 -0.09-2.00 0.046-0.10 2 (Constant) 695.05 11.08 62.71 0.000 Annual school enrollment 0.00 0.00 0.051.06 0.2880.05 Percentage on FRL -0.34 0.03 -0.48-10.28 0.000-0.46 Years at the school 0.26 0.23 0.061.11 0.2680.06 Years as a principal -0.08 0.14 -0.03-0.62 0.537-0.03 Level of education -1.72 0.97 -0.08-1.77 0.077-0.09 Use of D3M by teachers to improve instruction 0.31 0.22 0.091.42 0.1560.07 Data-driven culture -0.33 0.36 -0.05-0.92 0.360-0.05 Supporting systems -0.32 0.25 -0.07-1.31 0.191-0.07 Collaboration among teachers using D3M -0.12 0.64 -0.01-0.18 0.855-0.01 a Dependent Variable: NRT Math Grade 5

PAGE 90

90 CHAPTER 5 DISCUSSION, RECOMMENDATIONS, AND CONCL USIONS Discussion The role of the public school principal ha s evolved since the im plementation of the NCLB Act of 2001. Not only has th e principal become the primary focus for public scrutiny of school success or failure, but s/he has also beco me the primary person responsible for programs, resources, professional development, and staffing. The increased pressure for accountability has resulted in a need for the principal as the schoo l leader to analyze and understand all activities within the school. Many public schoo l principals have turned to data-driven decision making as a tool to provide visibility not onl y into student performance, but into the effectiveness of other school operations as well. How has this evolution in the role of the public school principal affected student achievement? Principal leadership studies during the past twenty years have yielded little information regarding the relationship between th e principals actions a nd student achievement, however research conducted by Hallinger and He ck does show that principals create a measurable effect on school improvement and student achievement (Hallinger and Heck, 1998). In addition, the question arises as to the importan ce of data-driven decision making as a tool for principals to use in order to improve student achievement. Study Purpose The purpose of the study was to exam ine th e relationship between data-driven decision making by elementary school principals and school level student achievement. In particular, the study was designed to determine the effect of th e principals beliefs a bout data-driven decision making on student achievement. This chapter will provide a discussion and interpretation of the

PAGE 91

91 results of the study as well as recommendations for future research, implications for public schools and their leaders, and conclusions. Target Population The target population for the study included pr incipals from Florida public elementary schools. The study addressed all public elementary schools in all counties in the state of Florida that are not charter or speciali zed schools and that te ach prekindergarten or kindergarten through grade five. The target populati on included 1468 public elementary schools in Florida. Public elementary schools from Brevard and Clay c ounties were eliminated because they serve kindergarten through sixth grade. The final response count for the survey and school achievement data was 471, representing 32% of the elementary school principal population. For this study, a survey was conducted to de termine principal beliefs about data-driven decision making. The State-wide Data Driven Read iness Study Principal Survey was determined to be the best fit for the purpose of the study. The survey also included questions designed to determine school demographics and principal characteristics including school enrollment, percent of students on free and reduced lunch, ye ars of experience as a principal, years of experience as a principal at that school, and le vel of education. The study incorporated FCAT NRT and FCAT SSS data for the year 2008 fr om the Florida Department of Education. Specifically grade level scale scores for each sc hool for math and read ing FCAT SSS and FCAT NRT tests for grades three, four, and five were used. Summary and Discussion of Results This study was designed to answer two research questions regarding pr incipal beliefs about data-driven decision m aking. In addition to de termining which beliefs are held by public elementary school principals in the state of Flor ida, the study also tested the relationship between the principals beliefs regardi ng the use of data-driven decisi on making and student achievement.

PAGE 92

92 Specifically, the research tested whether data-d riven decision making by principals acted as a mediator for student achievement The study also examined the effect of school enrollment, percentage of students on free and reduced lunch, years as a principa l, years as a principal at the school, and level of education on student achie vement. In addition, a factor analysis was conducted on the survey instrument. The followi ng provides a discussion and analysis of the results of the study. Factor Analysis Because data-driv en decision making is relatively new in schools, there are few models that address its use in the public school setting from a principals perspective. The results of the factor analysis of the data-driven decision making principal survey generated four themes or constructs that reflect elementa ry school principal beliefs abou t data-driven decision making in their schools. Specifically the four themes were be liefs regarding the use of data by teachers to improve instruction, beliefs regarding a data -driven culture, beliefs regarding supporting systems, and beliefs regarding collaboration am ong teachers using data-d riven decision making. These four themes represent a unique repres entation of actual beliefs held by principals about the activities that involve data-driven decision making in their schools. The results of the factor analysis not only add to the body of knowledge regarding a theory for data-driven decision making in schools, but they also pr ovide principals with a practical illustration of those particular activities which have value in a public elementary school setting. Elementary school principals who are interested in implementing or evaluatin g the prevalence of da ta-driven decision making in their schools may benefit from th e results of the factor analysis. The original data-driven decision making su rvey included 82 items. The first 20 items were eliminated from the factor analysis becau se they did not specifica lly address data-driven decision making activities within the school. The last five items were also eliminated from the

PAGE 93

93 factor analysis because they addressed school de mographics and principa l characteristics. The remaining 57 items were analyzed using the maximum likelihood estimation (MLE) model. The results of the factor analysis ut ilized 23 of the original 57 items included in the study. The factor analysis was conducted using both a principal components m odel and the MLE model. The results from the MLE model provided the best fit because it resulted in high er correlations for the resulting four factors listed a bove. One additional question was el iminated as a result of the reliability analysis. The principal components model also loaded the questions into f our factors, but the correlations were too low to be c onsidered viable. It is interesting to note that the four factors that resulted from the principle components mode l were much closer to the data-driven decision making factors found in recent literature. One possi ble reason for the difference in constructs is that these factors often em phasize data-driven decision making as part of the school improvement process, continuous improvement, and school reform (Schmoker, 2005; Reeves, 2002), not necessarily the day-to-day activities th at occur in the public elementary school. The factors identified through the principle com ponents model include beliefs regarding teacher actions to improve instruction, beliefs regarding the use of data to suppo rt school goals, use of data for self-improvement by teachers, and overa ll beliefs regarding da ta-driven cultures. In the MLE factor analysis used for this study, the first factor, beliefs regarding the use of data-driven decision making by teach ers to improve student achieveme nt, targeted the principals beliefs about the day-to-day actions of teachers. Ni ne items loaded into this construct. Items in this category included the use of data to make changes to instruction and curriculum, improve performance, set instructional goa ls and targets, communicate and identify student needs. Also included in this factor was the belief that t eachers feel personally responsible when school

PAGE 94

94 improvement goals are not met. All nine of the items in this category relate to the principals perceptions regarding the teachers use of data -driven decision making practices, indicating the perceived importance of the teachers actions an d how their use of data-driven decision making affects student achievement. Six items loaded into the second factor, belie fs regarding a data-dri ven culture. Included in this factor are beliefs about support for teacher s, data transparency, quality of decision making in the school, school improvement goals, and m odeling by administrators. This factor also included an item that addressed th e importance of analysis in orde r to improve. All six items in this category reflect the percei ved importance of incorporating data-driven decision making into the culture of the school. Data-driven decision ma king implies change, and for change to last, it must become a part of the culture. The third factor addressed in the factor analysis is the principals beliefs regarding data to provide support systems for the school. The five items that loaded into this category include beliefs about the use of data to communicate, determine staff development needs, allocate resources, and support data-driven dialogue. The cat egory also includes beliefs about the use of multiple resources to assess educational program s. The items in this category reflect the importance of integrating data-d riven decision making into the overall network of the schools operations, not just the classroom. The fourth factor addressed the principa ls beliefs regardi ng collaboration among teachers using data-driven decision making. Alt hough only two items loaded into this category, the scree plot and the correla tions supported the inclusion of this factor. The two beliefs addressed in this category include beliefs about the degree to which pr incipals believe that teacher teams meet regularly to look at student data and make instructional plans and beliefs

PAGE 95

95 about when teachers in the school meet with each other, they usually focus on improving student learning. This category reflects the principals beliefs that colla boration among teachers is key to the successful implementation of data-driven decision making in the classroom. At least three of the domains identified through the MLE fact or analysis are supported by leading educational experts. Research by Mar zano (2005), Schmoker (2005), and Stiggins (1999) emphasize the importance of the teacher with respect to student achievement The factor analysis domain of principal beliefs regarding teacher use of data-driven decision making to improve student achievement supports this research. The domain of data-driven culture is supported by Panettieri (2006), who points out the importance of not only collection of data, but also a school wide emphasis on outcome assessments, prog ress monitoring and feedback, and teacher ownership of outcomes, and the administrators ability to buil d a learning organization. School leaders must develop and use effective strategies for data collection and an alysis, and must help teachers to understand and work with data to improve learning in the classroom (Creighton, 2007). Research by DuFour (2006) and Schmoker (2005) highlights the importance of collaboration among teachers, another domain identified through the survey. Items within three of the four domains also align with a recent dissertation study by Susan Hutton (2007) that included a literature review of research relating to principal use of student achievement data in decision-making. According to Hutton, the literature fe ll into the domains of analysis of data, using data to communicate, us ing data for school improvement, and data-driven cultures. Although some portions of the factor analysis are similar to ot her studies, the constructs that resulted from this study are uni que in their relationship to each other. Each set of beliefs held by principals represents a set of important activiti es that must occur within the school for data-

PAGE 96

96 driven decision making to be successful. Yet each set of items supports other items in other categories. For example, collaboration by teach ers who meet regularly in teams to discuss student data and make instructiona l plans also reflects their use of data to set instructional goals and targets. Whereas many models for data-drive n decision making represent separate functions (AASA, 2002; Hutton, 2007; Reeves, 2002), the four domains represented through this set of constructs are interrelated. The f our themes represented through th e factor analysis represent a simple way for principals or other school admi nistrators to analyze the use of data-driven decision making within the school from a global perspective. It al so provides them with both a macro and a micro view of data -driven decision making activities. 1. What are the beliefs held by elementary school principals in Florida w ith respect to datadriven decision making? Results from the study indicate that elementary school principa ls in the state of Florida value data-driven decision making in their schools, and that they beli eve it is an effective tool to improve student achievement. The item means ranged from 4.43 to 5.79, indicating that on average principals agreed with every item in the survey. Once again principals identified the importan ce of a data-driven culture in their schools reflecting a belief that data-driven decision making must be an integr al part of the school culture in order for it to be effective. This category re presented the strongest be liefs held by principals, with means ranging from 5.79 to 5.37. All six item s were ranked seven or higher, representing 60% of items in the top 10 category. Principals also verified the importance of teacher actions with respect to data-driven decision making by ra nking beliefs about teacher use of data to identify students in need, set instructional targets, and work collaboratively to improve curriculum and instruction very high. Principals al so believed strongly that teacher teams meet

PAGE 97

97 regularly to look at stude nt data and make instructional plans, implying confidence in the use of data-driven decision making as part of the team planning process within their schools. Principals were less confident about the system s that were in place to support data driven decision making within their schools. The means fo r items in this category ranged from 5.32 to 4.98 for four of the five factors in this category. Principals indicat ed moderately strong beliefs in the use of multiple data sources, the use of data to allocate resources and determine professional development needs, and as a means for dialogue during staff meetings. These moderately high ratings by principals indicate that they believe in th e use of data to facilita te staff discussions and to make decisions that affect allocation of resources and staff development activities. The principals beliefs regarding the teachers ability to have significant input into data management and analysis practices ranked s econd from the bottom. This low ranking may indicate a belief that data-driven practices are standard ized within their school or district. Principals were least confident in their beli efs regarding the teachers use of data-driven decision making within their schools. Six of the nine items in this category fell in the bottom half of the list. Five of the six items addressed the use of data to make changes and to communicate frequently with parents. The perception may be that although teachers may use data at the beginning of an endeavor to identify student need s and set instructional goals and targets, they are less confident in the teachers continued use of data throughout the learning process to drive change. Finally, although principals believed stro ngly in collaboration by teacher teams, they felt less strongly about the belief that when teachers meet with each other they focus on improving student achievement. The principal beliefs regarding data-driven decision support the sta ndards identified by the National Association of Elementary School Prin cipals (NAESP, 2004). Standard Five ties to

PAGE 98

98 the items relating to principal beliefs regardi ng teacher use of data-d riven decision making to improve student achievement, requiring the use of mu ltiple sources of data as diagnostic tools to assess, identify and apply instructional improveme nt. The suggested strate gies include use of a variety of data sources to measure performance, an alysis using a variety of strategies, using data as tools to analyze student weaknesses and make adjustments to instruction, benchmarking against other schools with similar demographics, and creating a data-driven school environment. One interesting note is that when answering th e survey, principals i ndicated that they felt strongest about their beliefs about the importa nce of a data-driven culture. However, the standards emphasize data-driven decision making to change instruction and improve student achievement. The high rating by principals rega rding culture may indicate their understanding that in reality one must make data-driven deci sion making a part of the school culture in order for it to be long-lasting and effective. In summary, elementary school principals expressed strong be liefs about the use of datadriven decision making in their schools to support student achievement. They felt strongest about the importance of establishing a data-driven culture. They also indicated that they believe strongly that teachers use data-driven decision making to identify at-risk students and to set instructional goals and targets. Pr incipals were less confident about the teachers use of data to make continuous change based on student data. Results also supported a perception that while teachers may use data-driven decision making to ev aluate their students, they do not necessarily personalize its use to their personal professional development needs or to the school improvement activities. The challenge will be for principals to continue to drive data-decision making practices into all school operations. The ability to maximize the effectiv eness and efficiency of school

PAGE 99

99 resources becomes more important as the bar for student achievement con tinues to rise and the available resources continue to dw indle. At a global level, the pr incipal must effectively balance the needs of the students to make informed deci sions when allocating limited staff resources and materials, designing and supporting instructional programs, a nd providing effective staff development activities. Within the school, the prin cipal must expand their role as instructional leader to ensure that teachers continuously us e data to evaluate and support the needs of all students. The constructs identified thr ough the factor analysis and th e results of the additional descriptive statistics that were performed on the 22 items add to the existing body of knowledge regarding data-driven decision making in elem entary schools in Florida. One limitation associated with the study was that the principa ls beliefs were repres ented via a self-reported survey. It is assumed that members of the ta rget population had similar definitions for and understanding of data-driven decisi on making. It is also assumed th at their self-re ported answers accurately represented the principals be liefs about data-driven decision making. 2. Do Florida elementary school principals beliefs about data-driven decision making mediate the effect of principal characteri stics and school demographics on student achievement? Results of the study indicate that data-driven d ecision m aking does not mediate student achievement. Specifically, data-driven decision maki ng was not able to overcome the effects of school demographics and princi pal characteristics on student achievement. This study was designed to examine the relationship between pr incipal characteristics and school demographic variables, beliefs about the use of data-driven decision making, and student achievement. In this study the principal characteristics and school de mographic variables represented the antecedent variables. Student achievement re presented the outcome or depende nt variable. Principal beliefs about the use of data-driven decision maki ng represented the mechanism through which the

PAGE 100

100 principal characteristics and school demogra phic variables affect student achievement. According to the Baron and Kenny model (1986), in order for a variable to function as a mediator, the antecedent or predictor variable should have a significant positive relationship with the presumed mediator. Variations in the mediat or should account for vari ations in the outcome variable. Finally when both are held constant, there should be no significant relationship between the antecedent and the outcome variable. The mediat or variable is then considered to be the stronger variable. The three hypotheses were tested using a seri es of linear regression analyses. Because the results from the test of the firs t hypothesis were not significant, it can be assumed that principal beliefs about data-driven deci sion do not act as a mediator for student achievement for elementary school principals in the state of Fl orida. The results from the test of the second hypothesis indicated that all twelve of the measures of student ach ievement tested significantly against the data-decision making factors, suppor ting step two of the Baron and Kenny model. The test of the third criteria for the Baron and Kenny model indicated that each of the twelve regression models showed a significant relation ship between at least one of the antecedent variables (principal characteris tics and school demographics) and student achievement. All four of the data-driven decision making factors were excluded from the model in all but one case, the scale score for third grade NRT math. Thus, the third criteria set fort h by the Baron and Kenny model was not met. The principals beliefs rega rding the use of data-driven decision making by elementary school principals in the state of Florida do not act as a mediator for student achievement. With respect to the antecedent variables, th e percentage of students on free and reduced lunch tested significant in all of the twelve tests that were run. The partial correlation values for

PAGE 101

101 each test ranged from -.19 to -.83. In all but two cases, the values were stronger than -.5, indicating a strong negative re lationship between student ach ievement and the number of students on free and reduced lunch. A deeper look at the relationship between the percentage of students on free and reduced lunch and student achievement indicates that correlations are highest for reading for both tests The partial co rrelations for all measures of the FCAT SSS were higher than the NRT, particularly in the area of math. The models also showed that the student enrollment can also be a factor in student achie vement, particularly for reading. The analyses for third grade SSS reading, third grade NRT readi ng, and fifth grade NRT reading was significant with respect to the number of students enrolled at the school. The results of these analyses represent a cause for concern and raise the questio n as to whether we are indeed closing the gap and truly leaving no child left behind in r eading and math in the state of Florida. Although data-driven decision making does not mediate student ach ievement, results from the study do indicate that an indirect re lationship exists between data-driven decision making and student achievement. The finding that there is an indirect relationship between principal beliefs regarding da ta-driven decision making and st udent achievement supports the need for further research on leadership theory, specifically with respect to the relationships between the principal and da ta-driven decision making and the principal and student achievement.. Although no mediated effect was found, continued research regarding the effect of data-driven decision making on st udent achievement may shed a dditional light on the potential role of data-driven decisi on making within the school. Implications and Recommendations Data-driven decision m aking represents a mechanism through which to make informed decisions that are based on data ra ther than intuition. The results of the study indicate that data-

PAGE 102

102 driven decision making is present in Florid a elementary schools. The following discussion presents practical implications for the results of the study. Principals Hold Strong Beliefs Regar ding Data-Driven D ecision Making By agreeing with all 22 items on the data-d riven decision making survey, principals indicated that they believe that data-driven decision making has a place in the elementary school. They also indicated that they believe that the quality of deci sion-making within their school has improved as a result of data-driven decision ma king. Principals should continue to support and influence the use of data-drive n decision making in their schools so that they can better understand the ever changing academic environment within their school and the needs of their students. Elementary school principals have a great d eal of responsibility for the curriculum and instruction of Floridas students. In addition to the every day management and administrative tasks associated with ru nning a large organization within a school, they al so have responsibility for the integrity of the instructional program a nd the resulting student achievement. Data-driven decision making represents an important tool for principals to use to be tter understand the needs of students, teachers, parents, community member s, and other stakeholders, make decisions that improve school effectiveness and efficiency, and communicate results. From a management perspective, principals should further support th e use data-driven decision making to determine allocation of resources and professional develo pment for teachers. They should increase their emphasis on the use of multiple sources of data when engaging in data-driven decision making. Culture is an Important Part of Data-Driven Decision Making The high ranking of beliefs regarding a data-dri v e culture indicates that principals believe that data-driven decision making must be anchored in the cu lture of the school. Principals who are implementing data-driven practic es must pay attention to cultu re and recognize that change

PAGE 103

103 takes time. Principals should model data-drive n decision making in their everyday activities, especially in their role as the instructional le ader who works closely with teachers. Principals should also ensure that the school has a climate of data safety wh ere student achievement data is only used to improve student achievement and not used to evaluate teacher performance. Data transparency is also important to developing a data driven culture where all stakeholders are kept informed about student achievement results. Supporting Data-Driven Decision Making for Teachers Research sh ows that the teacher is the single most important factor that affects student achievement (Marzano, 2003). Because the results of this study indicate that the principals beliefs regarding data-driven de cision making have at best an indirect effect on student achievement, it would make sense for principals to seek to influence the one person who has a strong effect, the classroom teacher. Principals s hould seek to help teacher s to incorporate datadriven decision making practices into their daily routines. They should develop data-driven strategies that assist with choices in instructional strategies, the class curriculum, and assessments. Because principals are less confident in their beliefs regarding teacher use of datadriven decision making to affect student achieve ment, they should monitor and promote its use in the classroom. In order to maximize their role as the instructional le ader of the school, the principal should seek to influen ce the continual use of data to evaluate current performance and then make changes to instruc tion and curriculum, set instruc tional goals and targets, and communicate results. Principals should al so provide opportunities for collaboration among teachers as this is critical for successful impl ementation of data-driven decision making within the professional learning community.

PAGE 104

104 Closing the Gap The strong negative relationship between the percentage of students on free and reduced lunch and student achievem ent should be a cause for concern for school le aders as it raises the question as to whether Florida schools are indeed closing the achievemen t gap. Principals should seek to use data-driven decision making to gain insight into the status of these students within their schools, and should seek to evaluate, a ugment, and monitor the effectiveness of the instructional programs that support these students. Results of this study indicate that reading is more strongly affected by socioeconomic status and principals should prioritize efforts to support this subject area. Recommendations for Future Research. Additional research on the topic m ay allo w better understanding of the relationship between data-driven decision making by principals and student achievement. 1. There is a need for additional research that reviews the use of data-driven decision making and student achievement results over a longer period of time, not just one year. This research could also be expanded to include ot her measures within standardized tests, not just the scale scores. 2. Further research utilizing different surveys to collect information regarding principal beliefs may provide more detailed information that allows for a better understanding of how beliefs regarding data-driven deci sion making affect student achievement. 3. It may be useful to replicate the study using a different measure of student achievement. This particular study used standardized test resu lts in an effort to be consistent with the Florida school grading system. Of particular interest would be an analysis using other forms of summative data. 4. Additional research that targets the princi pals actions regarding data-driven decision making rather than beliefs about the topic mi ght provide informati on that more closely reflects how data-driven decision making act ivities within schools affects student achievement. 5. There is a need to understand how the principal uses data-d riven decision making to make decisions that ultimately affect student achieve ment. Additional research utilizing methods other than a self-report by prin cipals might provide additional insight into the use of data-

PAGE 105

105 driven decision making by principals. These other methods of data collection could include artifacts or interviews and focus groups with other stakeholders. 6. Additional qualitative research may provide fu rther detail and depth re lating to the thinking processes that are used by principals with respect to data-driven decision making within their schools. 7. Qualitative research that includes other stak eholders in the data-driven decision making process may also enhance understanding of the impact of the principals use of data-driven decision on school operations and student achievement. Summary The role of the elem entary school principal has changed as a result of increased accountability requirements, and principals have embraced data-decision making in order to make more informed decisions regarding student achievement. The results of the study indicated that principals in Florida elementary schools be lieve in the use of data-driven decision making within their schools, and they believe that the quality of the decision making within their schools has improved through its use. The results of the factor analysis indicated that four key constructs were present in Florida schools; beliefs rega rding the use of data-driven decision making by teachers to affect student achievement, beliefs re garding data-driven cultures, beliefs regarding the systems that incorporate data-driven deci sion making, and beliefs regarding collaboration among teachers using data-driven decision making. Does the use of data-driven decision making by principals improve student achievement? The results of this study indica ted that although there is an indi rect relationship between datadriven decision making and stude nt achievement, the principals beliefs were not strong enough to act as a mediator for student achievement. More research is needed re garding the use of datadriven decision making by principals and its effectiveness as a leadership tool. The strong negative relationship between the nu mber of students on fr ee and reduced lunch and student achievement represents a cause for concern. The results of this study indicate that the

PAGE 106

106 principals beliefs regarding data-driven deci sion making were not strong enough to overcome the effect of this relationship. What does overcome the negative re lationship between the percentage of students on free and reduced lunc h and student achievemen t? Educational leaders have long searched for the answer to this question, and additional research is st ill needed. The true benefit of data-driven deci sion making is that through its use, school leaders have increased visibility into the many issues that face our child ren. Data-driven decision making acts as a lens through which to view the current situation, al lowing principals to make the best possible choices that connect the realities of t oday with the possibi lities of tomorrow.

PAGE 107

107 APPENDIX A PERMISSION TO USE THE DATA-DRI VEN DECISION MAKING R EADINESS PRINCIPAL SURVEY

PAGE 108

108 APPENDIX B REQUEST FOR PRINCIPAL EMAI L AND R ESIDENTIAL ADDRESSES

PAGE 109

109 APPENDIX C INSTITUATIONAL REVIEW BOARD APPROVAL

PAGE 110

110 APPENDIX D INITIAL CONTACT LETTER

PAGE 111

111 APPENDIX E INITIAL EMAIL LETTER

PAGE 112

112 APPENDIX F FOLLOW-UP EMAIL LETTER

PAGE 113

113 APPENDIX G FOLLOW-UP INFORMED CONSENT LETTER

PAGE 114

114 APPENDIX H THANK YOU LETTER

PAGE 115

115 APPENDIX I IRB FOLLOW-UP AP PROVAL LETTER

PAGE 116

116 APPENDIX J IRB APPROVAL TO SEND EMAIL TO VOL USIA COUNTY PRINCIPALS

PAGE 117

117 APPENDIX K DATA-DRIVEN DECISION MAKING SURVEY

PAGE 118

118

PAGE 119

119

PAGE 120

120

PAGE 121

121

PAGE 122

122

PAGE 123

123 LIST OF REFERENCES Am erican Association of School Administrators. (2002). Using data to improve schools: Whats working Arlington, VA: American Associa tion of School Administrators. Anderegg, C. (2007). Classrooms and schools analyzing student data: A study of educational practice. Ed.D. dissertation, Pepperdine Univer sity, United States California. Retrieved August 12, 2007, from ProQuest Digital Dissertations database (Publication No. AAT 3252719). Arnold, J. (2007). School capacity for data-drive n decision making and student achievement. Ph.D. dissertation, University of South Caro lina, United State-Sout h Carolina. Retrieved August 11, 2008, from ProQuest Digital Dissertations database (UMI Publication No. 3280295). Barker, J.A. (1992), Future Edge: Discovering th e New Paradigms of Success William Morrow and Company, Inc., New York, NY. Baron, R.M., & Kenny, D.A. (1986). The moderator-m ediator variable dis tinction in social psychological research: con ceptual, strategic, and st atistical considerations. Journal of Personality and Social Psychology 51, 1173-1182. Bennis, W. (1989). On becoming a leader. Cambridge, MA: Perseus Publishing. Bernhardt, V. (1998). The school portfolio: A comprehensive framework for school improvement Larchmont, NY: Eye on Education. Bernhardt, V. (2002). The school portfolio toolkit Larchmont, NY: Eye on Education. Bernhardt, V. (2004). Data analysis for continuous school improvement Larchmont, NY: Eye on Education. Bettesworth, L. R. ( 2006 ). Administrators' use of data to guide decision-making. Ph.D. dissertation, University of Oregon, United States Oregon. Retrieved August 12, 2007, from ProQuest Digital Dissertations database (Publication No. AAT 3224072)." Black, P. & Wiliam, D. (1998). Inside the bl ack box: Raising standards through classroom assessment. Phi Delta Kappan 80 (2), Retrieved July 6, 2008 from http://www.pdkintl.org/kappan/kbla9810.htm. Bolman, L. & Deal, T. (1997). Reframing organizations: Artistry, choice, and leadership (Second Edition) San Francisco, CA: Jossey-Bass. Creighton, T.B. (2001). Schools and data: The educators guide to using data to improve decision making Thousand Oaks, CA: Corwin Press. Daggett, W. R. (2000). Moving from standards to instructional practice National Association of secondary School Principals. NASSP Bulletin, 84, 66-72

PAGE 124

124 Dale, K. (2005). "Lead is an Action Verb : Leader Actions Make the Difference." Academic leadership 3, 1. Darling-Hammond, L. (2007) Evaluating No Child Left Behind. The Nation. Retrieved August 22, 2008, from http://www.thenation.com/doc/20070521/darling-hammond. Darling-Hammond, L., Snyder, J., Ancess, J., Ei nbender, L., Goodwin, A. L., & Macdonald, M. B. (1993, May). Creating learner-centered accountability. New York: Columbia University, National Center for Restruct uring Education, Sch ools, and Teaching. DuFour, R. (2003). Building a professional learning community. School Administrator, 60 (5), 13-18 DuFour, R. (2002). The learning centered principal. Educational Leadership, 59(8), 12-15 DuFour, R., et. al. (2006). Learning by doing Bloomington, IN: Solution Tree. DuFour, R. (2004). "What Is a "Professional Learning Community"?" Educational Leadership 61(8): 6-11. Elmore, R.F. (2000). Building a new structur e for school leadership New York: Albert Shanker Institute. Felix, K. (2005). "Data-driven decision making.(Useful TOOLS)(From vision to action: How school districts use data to improve pe rformance report)(B rief Article)." ( Useful TOOLS)( From Vision to Action: How School Distri cts Use Data to Improve Performance report )(Brief Article) 12(3): 21(1). Florida Department of Education. (2005). Florida principal leadership standards Retrieved July 2, 2008 from https://www.floridaschoolleaders.org/fpls.aspx. Fullan, M. & Striegelbauer, S. (1992). The new meaning of educational exchange, 2nd ed. New York: Teachers College Press, Columbia University. Fullan, M. (2001). "How to make a turnaround succeed." Journal of Staff Development 22(1): 80. Good, R. (2008). Analyzing the impact of data analysis proc ess to improve instruction using a collaborative model Ph.D. Dissertation, Texas, A&M University-Commerce. Retrieved August 11, 2008 from ProQuest Digital Dissertations database (Publication No. 3245230). Greenleaf. R. (1988). Servant retrospect & prospect Indianapolis, IN: The Robert K Greenleaf Center. Hallinger, P., Heck, R.H. (1998), "Exploring the pr incipals contribution to school effectiveness: 1980-1995", School Effectiveness and School Improvement 9 (2), 57-91.

PAGE 125

125 Hallinger, P., S, R.H (1996a), "Reassessing the prin cipals role in school e ffectiveness: a review of empirical research, 1980-1995", Educational Administration Quarterly (32),1, 5-44. Hallinger, P., Heck, R.H (1996b), "The principals ro le in school effectiven ess: an assessment of methodological progress, 1980-1995", in Leithwood K (Eds), International Handbook of Educational Leadership and Administration 723-83. Hutton, S. (2007). A study of how Virginia scho ol leaders use student achievement data in decision making. Ph. D. dissertation, College of William and Mary, United StatesVirginia. Retrieved August 11, 2008 from ProQuest Digital Dissertations database (Publication No. 3287896). Ingram, D., K. S. Louis, et al. (2004). "Accountability policies and teacher decision making: Barriers to the use of data to improve practice ." Teachers College Record 106(6), 12581287. Interstate School Leaders Licensure Consor tium. (1996). Standards for school leaders. In The Jossey Bass Reader on Educational Leadership (pp.97-113) California: Jossey-Bass. Kannapel, P. J., & Clements, S. K., with Taylor D., & Hibpshman, T. (2005). "Characteristics of high-performing, high-poverty schools. ". Retrieved August 8, 2008 from http://www.ascd.org/portal/site/ascd/template. Kitchens, J. (2005). "Real-time access to stud ent data leads to real school reform." eSchool News Online : 3. Retrieved August 5, 2008 from http://www.eschoolnews.com/news/pdf/sifjul06.pdf. Kotter, J. P. (1996). Leading change Boston, MA: Harvard Business School. Lambert, L. (2006). "Lasting Leadership: A St udy of High Leadership Capacity Schools." The Educational Forum 70,(3), 238-54. Love, N. (2004) "Taking data to new depths." National Staff Development Council, 25 (4) Lunenberg, F. & Ornstein, A. (1991). Educational Administrati on Concepts and Practices Belmont CA: Wadsworth Publishing Co. Marzano, R. (2003). What works in schools. Alexandria, VA: Associat ion for Supervision and Curriculum Development. Marzano, Robert, et al. (2005). Classroom management that works: research-based strategies for every teacher Alexandria, VA: Association fo r Supervision and Curriculum Development. Marzano, Robert, et al. (2005). School leadership that works: from research to results. Alexandria, VA: Association for Supe rvision and Curriculum Development.

PAGE 126

126 Mathews, L & Crow, G. (2003). Being and becoming a principal. Boston, MA, Pearson Education, Inc. McIntire, T. (2005). Data: Ma ximize your mining, part one. TechLearning. Retrieved July 5, 2007 from http://www.techlearning.com/s howArticle.php?articleID=160400818 McIntire, T. (2005). Data: Ma xim ize your mining, part two. TechLearning. Retrieved July 5,2007 from http://www.techlearning.com/shared/pri ntableArticle.jhtml?articleID=164300240 McLeod, S. (2005). Data-driven teachers UCEA Center for the Advanced Study of L eadership in Education. Retrieved June 2, 2007, from http://www.scottmcleod.net/process/Redire ct? url=http%3A%2F%2Fwww.scottmcleod.net % McLeod, S. (2005). Data-driven decision making. The Savvy Technologist. Retrieved October 4, 2007, from http://technosavvy.org/2005/09/16/scot t-mcleod-data-drive n-decision-making/. McLeod, S. (2005). DDDM competencies for teac hers and administrators (DRAFT) Developed in conjunction with the Chicago (IL) Public Schools. Retrieved October 4, 2007, from http://www.scottmcleod.net/storage/2005_DDDM_Co mpetencies_for_CPS_Educators_D RAFT_3.pdf McLeod, S., & Seashore, K. (2006). Data driven decision making readine ss survey: Principals. Minneapolis, MN: University of Minnesota Mndez-Morse, S. (1991). The prin cipal's role in the instructiona l process: Implications for atrisk students. Issues. .about Change, 1(3), Austin, TX: Southwest Educational Development Laboratory. Retr ieved September 6, 2007 from http://www.sedl.org/change/issues/issues13.html. National Association of Elem enta ry School Principals. (2004). Leading learning communities: standards for what principals should know and be able to do. Alexandria, VA: National Association of Elementa ry School Principals. National Commission of Excelle nce in Education (1983). A nation at risk: the imperative for educational reform Retrieved August 6, 2008, from http://www.ed.ove/pubs/NatAtRisk/index.html. National Educational Association. (2003). Balanced Assessment: The Key to Accountability and Improved Student Learning. Student Assessment Series. G. W. Cutlip. Washington DC: National Education Association 16. National Education Association Foundation for the Improvement of Education. (2003). The 90/90/90 schools: A case study. Washington DC: National Education Association Foundation for the Improvement of Education.

PAGE 127

127 North East Florida Educ ational Consortium (2005) Data analysis for in structional leaders Retrieved August 6, 2008, from http://www .nefec.org/dl/delta/dafil/lesson2-3.htm. Noyce, P., D. Perda, et al. ( 2000). "Creating Data-Driven Schools." Educational Leadership 57(5), 52. Panettieri, J. C. (2006). "Data driven: Savvy school administrators are using data to improve instruction district wide.(ISKM E Special Series: Part 1)." T. H. E. Journal 33(7), 25 Retrieved September 5, 2007, from http://www.thejournal.com/the/learni ngcenters/center/?id=17864&msid=16 Patten, T.D. (2006). Principals perceptions of their prof essional development implementation for sustained change. Ph. D. dissertation, University of Florida, United States. Patterson, J. L. (1993). Leadership for tomorrow's schools Alexandria, VA: Association for Supervision and Curriculum Development. Peterson, K. & Deal, T. (1999). Shaping school culture: th e school leaders role San Francisco, Jossey-Bass. Reeves, D. (2002). The leaders guide to standards: A bl ueprint for educational equity and excellence San Francisco, CA: Jossey-Bass. Reeves, D. (2004). Accountability for learning Alexandria, VA: Associa tion for Supervision and Curriculum Development. Reeves, D. B. (2006). The learning leader : how to focus sch ool improvement for better results Alexandria, VA: Association for Supe rvision and Curriculum Development. Safer, N. (2005). "Research Matters/ How Stude nt Progress Monitoring Improves Instruction." Educational Leadership, 62(5), 81-83. Stiggins, R. (1999). Assessment, st udent confidence, and success. Phi Delta Kappan, 81, (3), 191-198. Stiggins, R. (2004). New assessmen t beliefs for a new school mission, Phi Delta Kappan 86 (1), 22-27. Retrieved August 2, 2007, from http://www.powayusd.com/projects/literacy /SSTTL/AssessDocs/PDFs/NewBeliefs.pdf Schmoker, M. (1999). Results: The Key to Continuous School Improvement Alexandria, VA: Association for Supervision a nd Curriculum Development. Schmoker, M. (2003). First things first: Demystifying data analysis. Educational Leadership 60 (10) 48. Schmoker, M. (2005). "Tipping Point: From F eckless Reform to Subs tantive Instructional Improvement." Phi Delta Kappan, 85 (6), 424-432. Retrieved March 10, 2007 from http://www.pdkintl.org/kappan/k0402sch.htm#1a.

PAGE 128

128 Senge, P. M. (1990). The fifth discipline: th e art and practice of the learning organization New York: Doubleday/Currency. Smith, W. & Andrews, R. (1989). Instructional leadership: How principals make a difference Alexandria, VA: Association for Supervisi on and Curriculum. Retrieved March 7, 2007, from http://eric.ed.gov/ERIC Docs/data/ericdocs2sql/conten t_storage_01/0000019b/80/1f/cb/a9. pdf. Sulser, D. P. (2006). The relationship between the use of t echnology for data-driven decisionmaking and student achievement in high school mathematics Thesis (Ed. D.)--Montana State University--Bozeman, 2006. http://etd.lib.mont ana.edu/etd/2006/sulser/SulserD0506.pdf. Supovitz, J. A. and V. Klein (2003). Mapping a course for improved student learning: how innovative schools systematically use student performance data to guide improvement Consortium for Policy Research in Education University of Pennsylvania. U. S. Department of Education, Office of Elementary and Secondary Education (2002). No child left behind: A desktop reference Washington D.C. Van Fossen, C. (2002). Principals and data-driven decision making Dissertation Abstracts International 63, 41 (UMI#3038988)

PAGE 129

129 BIOGRAPHICAL SKETCH Vicki Lynn Conrad White was born in 1958 in Li bertyville, Illinois. The oldest of three children, she grew up mostly in Indiana and Illin ois, graduating from Jefferson High School in Lafayette, Indiana in 1976. She earned her B.A. in elementary education from Purdue University in 1979. After graduating, Vicki taught fifth grade in Roselawn, Indiana. She returned to school in fall 1980, and graduated with an M. Ed. from the University of Illinois in 1982. After receiving her M.Ed., Vicki worked for Control Data Corporation as an Education Analyst in Sunnyvale, CA. During the next fifteen years, she worked in a series of marketing, training, and management positions for Inte l Corporation, Conner Peripherals, Seagate Technology, and Archive Corporatio n. Mrs. White received an M.B.A. from the University of Phoenix in 1991. Vicki White returned to the education field in 1997 when she became a media specialist at Read-Pattillo Elementary School. She has worked for the school 11 years, most recently as a Teacher on Assignment in an administrative capacity. Upon completion of her Ph.D. program, Vicki Wh ite will continue in her role at ReadPattillo Elementary School. Vicki has been married to Douglas James White for twenty-one years. They have three children: Douglas Ja mes, age 20, Margaret Ellen, age 18, and Jennifer Lynn, age 16