Professional Documents
Culture Documents
A Dissertation
Presented to
W
University of Denver
IE
In Partial Fulfillment
EV
Of the Requirements for the Degree
Doctor of Philosophy
PR
Presented by
Janet E. Filbin
August, 2008
W
IE
EV
PR
3320581
2008
©Copyright by Janet E. Filbin 2008
All Rights Reserved
W
IE
EV
PR
Author: Janet E. Filbin
Title: THE EFFECTS OF DATA-DRIVEN TEACHING AND LEADING ON
COLLECTIVE EFFICACY
Advisor: Dr. Elinor Katz
Degree Date: August, 2008
Abstract
The sanctions are mounting for those schools who are failing to meet federal,
state, and local accountability requirements. States and districts continue to spotlight
schools that are considered underperforming for scrutiny. As a result, staff within these
W
efficacy, the belief that an individual has about their own and their colleague’s
year to year is challenging. Research in this field is emerging and little is known about
PR
the factors that directly influence the collective efficacy of a school staff.
This research examines the relationships that teacher pedagogical precision, the
use and application of data to inform teaching for learning, and the principal’s use of data
with staff has on collective efficacy. The mediating effects of changes in student
achievement on the relationship between teacher and principal pedagogical precision and
collective efficacy were also investigated. Educators from 50 elementary, middle, and
ii
Survey and achievement results were collected three times throughout the school year
The findings of this study indicate that there is a direct and significant impact on
collective efficacy when teachers report higher uses of data to change instruction with an
effect size of .908 (p<.01) by spring. Teachers’ perception of the principal’s use of data
had a positive effect size ranging from .795 in fall (p<.001) to .409 in spring (p<.001).
Results show that, as a school risk factor, socio-economic status had a significant,
W
year (p<001) but a small, significant, and positive impact on achievement in winter and
spring (p<.001). IE
EV
PR
iii
Acknowledgements
I would like to extend my deepest appreciation to Dr. Elinor Katz for her
unwavering support and gentle, calm spirit in stewarding me through this journey. Ellie,
you are an inspiration to us all. I would also like to thank Dr. Martin Tombari for his
encouraging push toward the methodology used in this study. Marty, not only are the
results more robust, but my own sense of efficacy has been greatly enhanced because you
thought I had the abilities to undertake such a complicated approach. For my other
W
committee members, Dr. Brian McNulty and Dr. Susan Korach, your participation and
expertise meant a great deal to me. Brian, you will forever been my role model of a
transformational leader.
IE
EV
I must also give credit to my colleague, Norman Alerta, for his patient guidance
Normie, it goes without saying, I could not have done this without you! To my BFF,
PR
Toni Mote-Ponce, for letting me process excessively about this work and for lending me
your editing expertise. To my mother and siblings, thanks for your understanding and for
letting me make this a priority for a while. Finally and most importantly, to my greatest
contribution to this world, Michael and Megan, for it was only when I looked into your
eyes that I believed I was worthy of pursuing this goal. I will never catch up with you
iv
TABLE OF CONTENTS
Chapter 1 INTRODUCTION............................................................................................. 1
W
Definition of Terms....................................................................................................... 13
Organization of the Dissertation ................................................................................... 14
Summary ....................................................................................................................... 15
IE
Chapter 2 REVIEW OF THE LITERATURE................................................................. 16
EV
Introduction....................................................................................................................... 16
An Overview of the Theoretical Underpinnings of Collective Efficacy .................. 16
The Role of Social Cognitive Theory in Efficacy Development............................ 18
PR
Mastery Experiences............................................................................................. 19
Increasing the Likelihood of Mastery Experience through Goal Setting ............. 20
Evaluating the Precision of Practice .................................................................... 25
Data-Driven Decision-Making and Pedagogical Precision................................. 26
Teacher Perception, Teacher Efficacy, and Collective Efficacy .............................. 27
Impact of Teacher Perception on Student Achievement ....................................... 27
Teacher Efficacy and Student Achievement.......................................................... 30
The Link between Individual Efficacy, Collective Efficacy, and Achievement ..... 31
Measuring Collective Efficacy.............................................................................. 33
Pedagogical Precision and Collective Efficacy ........................................................ 35
The Influence of Pedagogical Precision on Efficacy............................................ 35
Creating a Culture of Collective Efficacy through Pedagogical Precision ......... 41
The Role of Leadership in the Development of Pedagogical Precision ................... 42
The Principal’s Contribution to Pedagogical Precision ...................................... 42
Defining the Vision, Mission, and Goals of the School ............................................ 44
Managing Curriculum, Instruction, and Assessment................................................ 45
Supervising and Supporting Teaching .................................................................. 47
v
Monitoring Student Progress................................................................................ 48
Promoting an Instructional Climate ..................................................................... 49
Leadership in Action: Cultivating Collective Efficacy through Pedagogical Precision
................................................................................................................................... 50
Measuring Pedagogical Precision of Teachers and Data Use by Principals ...... 51
The Impact of Data-Driven Changes in Teaching and Leading on Collective
Efficacy ..................................................................................................................... 52
Summary ....................................................................................................................... 54
Chapter 3 METHODOLOGY.......................................................................................... 56
Introduction................................................................................................................... 56
Approach....................................................................................................................... 57
W
Research Questions....................................................................................................... 60
Setting of the Study....................................................................................................... 60
Population ..................................................................................................................... 61
IE
Survey Sample....................................................................................................... 65
Instrumentation ............................................................................................................. 67
Collective Efficacy Survey .................................................................................... 67
EV
Data Use Survey ................................................................................................... 70
Validity Considerations ............................................................................................ 72
Content Validity .................................................................................................... 72
Construct Validity ................................................................................................. 73
Operational Survey ............................................................................................... 77
PR
Chapter 4 RESULTS........................................................................................................ 93
Introduction................................................................................................................... 93
Description of the Sample......................................................................................... 94
Correlations .......................................................................................................... 99
Missing Data....................................................................................................... 100
Validation of the Measurement Model ................................................................... 101
Factor Analysis ................................................................................................... 101
Verification of the Structural Model....................................................................... 107
Preparation of the Data ...................................................................................... 107
vi
Modeling ............................................................................................................. 112
Variables in the Model............................................................................................ 112
Latent Variables.................................................................................................. 112
Observed Variables............................................................................................. 113
Null Hypothesis ...................................................................................................... 114
Verification of the Measurement Model................................................................. 115
Analysis of the variables ..................................................................................... 115
The Structural Model .............................................................................................. 118
Parsimonious Models.......................................................................................... 137
Analysis at the School Level................................................................................... 142
Criterion Validity Check..................................................................................... 143
Summary ..................................................................................................................... 145
W
RECOMMENDATIONS................................................................................................ 148
Recommendations....................................................................................................... 175
District Leadership.................................................................................................. 175
Building Principals.................................................................................................. 177
Teacher Leaders ...................................................................................................... 179
Considerations for Future Research............................................................................ 181
Conclusions................................................................................................................. 183
REFLECTIONS.............................................................................................................. 184
vii
Appendix A -Collective Efficacy Scale and Instructional Data Decision-Making Survey
......................................................................................................................................... 193
Mean and Standard Deviations for Survey Responses across the School Year ............. 202
Appendix C -Spring Correlations between Latent Variables and Demographics .......... 207
W
Appendix F –Spring Factor Analysis Loadings............................................................. 214
IE
EV
PR
viii
LIST OF TABLES
TABLE 3. FACTOR MATRIX FOR THE 12 ITEM COLLECTIVE EFFICACY SCALE REPORTED BY
GODDARD (2002)....................................................................................................... 69
TABLE 5. PILOT TEACHER DATA USE AND TEACHER PERCEPTION OF PRINCIPAL DATA
USE TOTAL SCORE ..................................................................................................... 73
W
TABLE 8. SURVEY QUESTIONS FACTORED TOGETHER ...................................................... 75
IE
TABLE 9. SUBSCALE LABELS ............................................................................................ 75
TABLE 12. COACH SURVEY ITEMS USED TO VALIDATE TEACHER DATA USE .................. 81
PR
TABLE 15. SURVEY RESPONSE RATE BY LEVEL FOR EACH COLLECTION PERIOD ............ 95
TABLE 16. YEARS TEACHING IN THE SCHOOL BY DATA COLLECTION PERIOD ................ 96
TABLE 19. DESCRIPTIVE FOR THE FALL TEACHER DATA USE, TEACHER PERCEPTION OF
PRINCIPAL DATA USE, AND TEACHER COLLECTIVE EFFICACY BELIEFS .................... 97
ix
TABLE 20. DESCRIPTIVE FOR THE WINTER TEACHER DATA USE, TEACHER PERCEPTION OF
PRINCIPAL DATA USE, AND TEACHER COLLECTIVE EFFICACY BELIEFS .................... 98
TABLE 21. DESCRIPTIVE FOR THE SPRING TEACHER DATA USE, TEACHER PERCEPTION OF
PRINCIPAL DATA USE, AND TEACHER COLLECTIVE EFFICACY BELIEFS .................... 99
TABLE 22. FALL SURVEY EIGENVALUES AFTER VARIMAX ROTATION ............................ 102
TABLE 24. SPRING SURVEY EIGENVALUES AFTER VARIMAX ROTATION ......................... 102
TABLE 26. SECONDARY EFA EIGENVALUES FOR THE FALL SURVEY ............................. 104
W
TABLE 27. SECONDARY EFA EIGENVALUES FOR THE WINTER SURVEY ........................ 104
TABLE 28. SECONDARY EFA EIGENVALUES FOR THE SPRING SURVEY .......................... 105
IE
TABLE 29. CHRONBACH’S ALPHA COEFFICIENTS ESTIMATIONS FOR FACTORS ............... 106
EV
TABLE 30 EFA LOADINGS OF OBSERVED VARIABLES ON LATENT VARIABLES ............ 107
TABLE 31. NUMBER OF CASES BY SCHOOL AFTER DELETION OF RESPONSE SETS WITH
MISSING ITEMS ........................................................................................................ 109
PR
TABLE 32. PARAMETER ESTIMATES AND FIT STATISTICS FOR MODEL 1 TDU TO CE .. 121
TABLE 33. PARAMETER ESTIMATES AND FIT STATISTICS FOR MODEL 2 TPPDU TO CE
................................................................................................................................. 124
TABLE 34. PARAMETER ESTIMATES AND FIT STATISTICS FOR MODEL 3 TPPDU TO TDU
................................................................................................................................. 127
TABLE 35. PARAMETER ESTIMATES AND FIT STATISTICS FOR MODEL 4 TDU TO AC ... 130
TABLE 38. PARSIMONIOUS MODEL FOR FALL, WINTER, AND SPRING ........................... 141
x
TABLE 39. MEAN DIFFERENCES FOR CE, TDU, AND TPPDU ACROSS FALL, WINTER, AND
SPRING ..................................................................................................................... 142
TABLE 40. MEAN SCORES FOR TEACHERS BY HIGH OR LOW DATA USE ....................... 144
TABLE 41. MEAN SCORES FOR TEACHERS’ PERCEPTION OF PRINCIPALS BY HIGH OR LOW
DATA USE ................................................................................................................ 145
W
IE
EV
PR
xi
LIST OF FIGURES
W
FIGURE 8. MODEL 5: TPPDU TO AC ............................................................................. 131
IE
FIGURE 9. MODEL 6 AC TO CE ...................................................................................... 134
FIGURE 10. PARSIMONIOUS MODEL FOR THE FALL DATA ............................................. 138
EV
FIGURE 11. PARSIMONIOUS MODEL FOR THE WINTER DATA ......................................... 139
FIGURE 12. PARSIMONIOUS MODEL FOR THE SPRING DATA ........................................... 140
PR
FIGURE 13. CHANGES IN CE TOTAL MEAN SCORES ACROSS THE SCHOOL YEAR........... 154
FIGURE 14. TOTAL MEAN OF SUMMED COLLECTIVE EFFICACY SCORES BY SCHOOL SES
STATUS .................................................................................................................... 160
FIGURE 15. TOTAL MEAN OF SUMMED CE, TDU, AND TPPDU SCORES BY ACHIEVEMENT
GAINS AND SES STATUS .......................................................................................... 167
xii
Chapter 1
INTRODUCTION
“We pass through this world but once. Few tragedies can be more
extensive than the stunting of life, few injustices deeper than the denial of
an opportunity to strive or even to hope, by a limit imposed from without,
but falsely identified as lying within.”
Gould, 1996 (p. 50).
W
The Birth of the Urban Myth
Twenty-seven middle school faculty members from a large urban district huddle
IE
together around two long tables in the school library for a mandatory in-service. Theirs is
a school identified as “in need of improvement”; the achievement scores of the 7th and 8th
EV
grade students were far below the expected level of proficiency for their grade. It is
February and just three weeks away from the next administration of the state assessment.
PR
benchmark assessment results to identify the language arts concepts and skills with which
their students may be struggling. “This won’t make a difference. Our kids can’t read,
they’re low in everything. They can’t do this test. This doesn’t give me any information I
don’t already know,” states one teacher. “This isn’t fair to test them on grade-level. I
can’t even teach the stuff on the test because my kids just aren’t ready for it,” disputes
another. “No one really understands what these kids are dealing with. With their home
life, we are just happy they come to school,” adds yet another educator. “Okay, okay I
know. We all have to realize we have a much harder population to deal with and we have
1
a long way to go. I agree it’s not fair that we have to have the same results as other
schools in the better neighborhoods. I know you all are trying your hardest,” sighs the
school principal. “We just have to face the fact that ours is the urban challenge.”
Since the No Child Left Behind (NCLB) legislation took root eight years ago, a
substantial amount of attention and energy has been directed toward increasing student
W
Although this joint legislative commitment has provided the national, state, and local
impetus to seriously wrestle with the inequities inherent to the public education system,
IE
the achievement gaps continue to persist into the 21st century. Despite some promising
practices that were launched in the 1960’s, such as Head Start and the Elementary and
EV
Secondary Education Act, progress in closing these gaps stopped around 1990 (Ferguson
& Mehta, 2004). Hence, there is a push for a potent accountability system designed to
PR
hold schools, districts, and states responsible for ensuring that every learner achieves
NCLB requires that all students be included in state assessments and their results
be reported based on their grade, language, ethnic, gender, and economic identifiers as
well as specialized learning needs. Therefore, the “not so hidden” agenda that the NCLB
mandate brings to the forefront is accountability at a school level for every learner sitting
within a classroom, regardless of their preparation to access the test. As expected, the
pressure to include each student in the state assessment system and rapidly raise scores
2
continues to mount. Yet, despite the significant sanctions attached to federal and state
children who still do not read or apply mathematical reasoning skills at grade level.
Students of color, lower socio-economic status, those with language differences, and
students with special education support needs continue to be the majority of youth in
today’s classrooms that are not achieving at a proficient level as defined by the state
W
continue to problem-solve ways to effectively impact the learning of students who
achievement has remained static or steadily decreased over time. In fact, many of these
educators arrive at the school doorsteps less than prepared to undertake such a task.
PR
Indeed, teacher retention has been found to be problematic in these schools; and this in
turn translates to a less than qualified staff for students who have the greatest need for
highly-trained teachers (Darling-Hammond & Sykes, 2003). These teachers are often not
adequately prepared to support youth who require someone with extensive content
Research has shown that teacher satisfaction is related to how much influence
they feel they can make on the student population with whom they are working
3
(Coladarci, 1992). As a result, efficacious teachers are able to produce environments
where learning occurs. There is also evidence that teacher efficacy translated to the
school level provides a picture of how staff perceives their ability as a group to impact
changes. (Bandura, 1997; Goddard, Hoy, & Hoy, 2000; Goddard, 2001). Group pressure,
Therefore, if they believe as a group that they can accomplish the task at hand then
individual teachers are more likely to believe that they can do so as well.
Problem Statement
W
The notion of collective efficacy, the belief of a group that they have the ability
perform and succeed at responsibilities that may be challenging, has been the subject of
IE
educational research in the past decade. Higher levels of collective efficacy in school
EV
staff have been associated with increased student achievement (Bandura, 1997; Goddard,
2001; Tschannen-Moran, 2004). Goddard, Hoy, and Hoy (2000) and Goddard (2001)
found that when controlling for demographic factors that have been highly associated
PR
with lower performing schools, collective efficacy had a substantial positive effect on
student achievement. Collective efficacy holds great promise as a factor associated with
schools that are making differences for students who otherwise might be at risk.
However, the body of research available on collective efficacy is fairly recent. Although
the findings of Bandura (1997), Goddard, Hoy, and Hoy (2000), and Goddard (2001)
have been replicated, exploration of the direct and indirect influences on collective
decision-making, and prior student achievement are variables that have been found to be
4
correlated with collective efficacy within schools (Bandura, 1993; Goddard, 2001; Ross,
a school system. When provided with evidence that their students are making gains,
turn, manifest in higher levels of student achievement. As this cycle becomes more
integrated into the organization, teachers become acculturated in this belief system.
However, there are schools plagued with repeated poor performance on state and national
W
assessments.
IE
The low levels of efficacy at the group level may influence teachers’ perceptions
that their students are incapable of learning skills defined as proficient for their grade
EV
level regardless of what they do in their classroom. As a result, this problem may become
educational staff. Tournaki and Podell (2005) found that teachers with lower levels of
efficacy predicted less chance of success for students who were reading two grade levels
below while teachers with stronger efficacy were more positive about student outcomes.
Although these are the schools that require the most competent staff, teachers may
and Bandura (2002) suggest that “unless people believe they can produce desired
outcomes and forestall undesired ones through their actions, they have little incentive to
5
act or to persevere in the face of difficulties” (p.108). Creating an atmosphere of
optimism is a difficult task under these conditions. Once the culture of low expectation
for student performance has been established and embedded, the effort required to shift
A key to turning around schools that struggle to support student learning may lie
in the ability of building and district administrators to cultivate the collective efficacy of
the staff. However, the theoretical underpinnings of collective efficacy have only recently
been established and related studies are sparse. Ross, Hogaboam-Gray, and Gray (2004)
W
note that currently there are no studies that have examined the interaction of the principal
in developing the collective efficacy of the school staff, however, Bandura (1993) and
IE
Goddard (2001) have provided evidence that a supportive principal can contribute to the
EV
development of both individual and group efficacy. Creating an organization of shared
inquiry and decision-making using student achievement data has been suggested as a
positive administrative action that can impact collective efficacy (Goddard, 2002; Ross,
PR
There have also been a number of studies that validate the link between strong
instructional leadership, data use, and student achievement. Marzano, Waters, and
McNulty (2005), Reeves (2002), and Youngs and King ( 2002) found that shared
decision-making in using data to identify and evaluate the progress of school goals,
understand and monitor curriculum and instruction, and to assess student progress with
making have also been found to be highly predictive of collective efficacy (Bandura,
6
1993; Goddard, 2001; Ross, Hogaboam-Gray, & Gray, 2004). Therefore, the use and
With the call for proficiency for all students by 2014, as mandated by NCLB,
schools are placing greater emphasis on achievement in math and reading. Schools
identified as “lower performing” are uniquely challenged to make the progress necessary
to meet the performance targets, a task that may seem ominous and unattainable to the
W
staff involved. However, there are a number of promising school characteristics that
have been found to be associated with schools that are making strides forward even with
IE
variables traditionally associated with the lower performing schools, such as greater
EV
ethnic diversity and students from lower socio-economic backgrounds. The
characteristics of these schools include collective efficacy of the staff, shared decision-
making, and the use of data for collaborative inquiry (Bandura, 1993; Goddard, 2001;
PR
Marzano, Waters, & McNulty, 2005; Reeves, 2002; Ross, Hogaboam-Gray, & Gray,
formative measures can provide powerful information for building leaders and teachers
(Robinson, Phillips, & Timperley, 2002; Shepard, 2002; Stiggins, 2005), the information
must be used to shift teacher pedagogy to affect student learning. Using assessment data
can help provide more precision to inform teachers about what students have learned as
7
well as the best way to adjust their pedagogical approach (Shepard, Hammerness,
The purpose of this study is to examine the effect that “pedagogical precision”, or
the analysis and application of data to adjust instructional approaches, has on student
achievement and on the collective efficacy beliefs of teachers. This research also
investigates the relationship between teachers’ perception of the principal’s use of data to
drive instructional conversations and the impact this has on their own use of pedagogical
precision, and their individual collective efficacy beliefs. Therefore, the unit of analysis
W
of this study is at the individual teacher. In addition, since collective efficacy is a group
as under-performing participated in this study. The population of interest was schools that
PR
initiative that included the administration and use of district benchmark assessments three
times throughout the year to inform school and classroom curriculum alignment and
responsive instruction during the 2007/2008 school year. Data from the entire population
of 31 elementary schools, 10 middle schools, and 9 high schools were used in the
analysis.
the initiative. Building leadership teams that included principals, building instructional
8
coaches, grade or content area teacher leaders, and technology specialists were identified
and charged with overseeing that all teachers in the building followed the improvement
practices. To carry out this charge, the building leadership team members were provided
pacing guides;
W
data, interpret results to inform the coverage and depth of curriculum
Principals were asked to share results of the benchmark assessments with staff
immediately after the assessment window and to provide time for staff to have
conversations about the results. Principals were also engaged in guided inquiry during
their monthly meetings to exchange ways in which they were supporting shared decision-
9
A comprehensive program evaluation was utilized to monitor the fidelity to
implementation of the curriculum and the use of the information provided through the
benchmark assessments to shift instructional practices. The district also gathered teacher
perception of the efficacy of the staff to impact changes. A set of surveys to examine
changes throughout the school year in teachers’ use of data-driven instructional decision-
making, their perception of the principal’s use of data for instructional conversations, and
their collective efficacy was administered three times from September 2007 through May
2008 to evaluate the initiative impact. In addition to the survey data, changes in student
W
achievement mediate the relationship between the pedagogical precision of teachers and
IE
their collective efficacy, and between teacher perception of principals’ use of data and the
Research Questions
PR
The following three research questions were used to structure the study and are
10