Theses B.ed

Faculty of Education
Allama Iqbal Open University, Islamabad
APPROVAL FORM
The research project attached here to, titled “ Study of relationship between Head
Teachers’ Leadership practices and Teachers self-efficacy”. Proposed and submitted
by Usama Hanif Roll No. BM606189 in partial fulfillment of the requirements for the
degree of B.Ed. (1.5 year) Teacher Education is hereby accepted.
Supervisor: _______________________ (Signature)
Evaluator: ________________________ (Signature)
Dated: _____________
Research Project Submission Approval Form
Research Project entitled “ Study of relationship between Head Teachers’ Leadership
practices and Teachers self-efficacy”. Submitted by Usama Hanif Roll No.
BM606189 Registration No. 17-PHD-00739 Programme B.Ed (1.5 year) has been
read by me and has been found to be satisfactory regarding its quality, content,
language, format, citations, bibliographic style, and consistency, and thus fulfills the
quality requirements of this study. It is ready for submission to Allama Iqbal Open
University for evaluation.
Dr. Riaz Hussain
Date:___________ _____________________
ABSTRACT
The attrition rate of teachers in an urban/rural school district in Hafizabad caused
schools to fail to attain annual yearly progress. To reverse this problem,
administrators must understand the importance of their leadership and teacher efficacy
and the need to nurture teachers to increase student performance. The purpose of this
sequential mixed-methods study was to determine whether a relationship existed
between leadership and efficacy. Total-population sampling was used to obtain 21
high school and elementary teachers who completed two surveys to examine the
relationship between Head Teachers’ behaviors (human relations, trust/decision
making, instructional leadership, control, and conflict) and teacher efficacy (student
engagement, instructional strategies, and classroom management). Survey data were
analyzed using Pearson’s product-moment correlations. In addition, face-to-face
interviews were conducted with 5 teachers who had 10 or fewer years of teaching
experience. These data were analyzed using thematic analysis. Quantitative findings
indicated significant relationships between instructional leadership with teacher
engagement and conflict with teacher engagement. Themes, based on the integrated
model of teacher efficacy, revealed connections with the principal and support,
guidance, and structure provided by the principal. Principals must focus on leadership
behaviors that may increase teacher efficacy. These endeavors may contribute to
positive social change when school leaders support teachers, who, in turn support
students in their educational challenges to increase academic performance.

Table of Contents
Chapter Page
1. Introduction to the Study 1
1.1 Statement of the Problem 2
1.2 Nature of the Study 3
1.3 Research Questions 4
1.4 Research Hypotheses 7
1.5 Purpose of the Study 12
1.6 Theoretical Framework 13
1.7 Teacher Efficacy 13
1.8 Leadership Behaviors 13
1.9 Definition of Terms 15
1.10 Assumptions, Limitations, Scope, and Delimitations 17
1.11 Significance of the Study 18
1.12 Implications for Social Change 19
1.13Summary 20
2. Review of the Literature 22
2.1 Introduction 22
2.2 Leadership Theories 24
2.3 Leadership Scales 25
2.4 Social-Learning Theory and the Concept of Efficacy 28
2.5 Teacher Efficacy 28
2.6 Teacher-Efficacy Scales 29
2.7 Related Studies on Teacher Efficacy 31
2.8 Relationship Between Leadership and Self-Efficacy 33
2.9 Synthesis 35
2.10 Literature Related to the Method 38
2.11 Literature Related to Differing Methodologies 39
2.12 Summary 39
3. Methodology 41
3.1 Introduction 41
3.2 Mixed-Method Design 42
3.3 Setting and Sample 42
3.4 Population and Sample for the Survey 43
3.5 Participant Selection for Interviews 43
3.6 Connection to the Research Questions 44
3.7 Quantitative Survey 47

3.8 Survey Instruments 48
3.9 Qualitative Interviews 51
3.10 Data Analysis 52
3.11 Survey Data Analysis 52
3.12 Interview Data Analysis 53
3.13 Evidence of Quality 53
3.14 Protection of Participants’ Rights 56
3.15 Survey Informed Consent and Confidentiality 57
3.16 Interview Informed Consent and Confidentiality 57
3.17 Researcher’s Role 58
3.18 Summary 59
4. Results 61
4.1 Introduction 61
4.2 Quantitative Data Analysis 62
4.3 Data Management and Scoring 62
4.4 Survey Participants 63
4.5 Quantitative Survey Analysis 64
4.6 Qualitative Data Analysis 75
4.7 Interview Transcription and Coding 75

4.8 Interview Participants 75
4.9 Qualitative Interview Analysis 77
4.10 Evidence of Quality 81
4.11 Summary 81
5. Interpretations, Recommendations, and Reflections 83
5.1 Overview 83
5.2 Interpretation of Findings 88
5.3 Implications for Social Change 90
5.4 Recommendations for Action 91
5.5 Instructional Leadership 91
5.6 Conflict 92
5.7 Recommendations for Further Study 93
5.8 Researcher Reflection 94
5.9 Conclusion 95
References 96
List of Tables
Table 1. Studies Using Teacher Efficacy Scales and Leadership Scales 36
Table 2. Education and Teaching Demographics of the Teachers (N= 19) 64
Table 3. Means and Standard Deviations for Five Domains of Principal Behaviors,
Three Measures of Self-Efficacy and Reliability Coefficients (Cronbach’s α) 65
Table 4. Interco relations for Five Domains of Principal Behaviors and Three
Measures of Self-Efficacy 74
Table 5. TSES Score , Rank, and Years of Teaching Experience of the Teachers
(N= 4) 76
Chapter 1: Introduction to the Study
According to the Commission on No Child Left Behind(2007), teacher quality
is the single most important aspect related to student achievement. The commission’s
report goes on to say that credentials alone will not raise the quality or the
effectiveness of teachers in the profession. Therefore, the commission recommended
changing the method of assessing the quality of teachers by requiring all teachers to
be highly qualified and effective teachers, not just highly qualified teachers.
Teacher efficacy has been a vital component of teacher effectiveness (Henson,
Kogan, & Vacha-Haase, 2001). Tschannen-Moran and Woolfolk Hoy (2002) noted
that, over the past 25 years, persuasive data have accrued regarding the positive
association teacher efficacy has on students’ motivation and achievement. Teacher
efficacy is defined as “teachers’ beliefs in their abilities to organize and execute
courses of action necessary to bring about desired results” (Tschannen-
Moran,Woolfolk-Hoy, & Hoy 1998, p. 22), and strong associations can be found
between teacher efficacy, a deeper desire to teach, and a greater likelihood that
teachers will not leave the profession. In this mixed-methods study, I examined the
relationship between principal behaviors and teacher efficacy, as well as gathered
information on teacher perceptions of this relationship.
Urban districts, such as the HAFIZABAD district in which the current study
was conducted, want to retain strong, efficacious teachers who are committed to the
profession to be successful in increasing achievement in their underperforming
schools. The urban/suburban district highlighted in this study is comprised of four
schools that encompass Grades K–5, one middle school, and one junior high school.
Two of the five have not made adequate yearly progress (AYP), based on the
standards established by the No Child Left Behind Act (NCLB, 2001). One of those
schools failed to meet the NCLB standards for 2 years (from 2010-2012) and the other
for 5 years (from 2007-2012). In this district, the attrition rate of teachers in the
failing schools is much higher than in those schools that attained AYP. The high-
attrition rate makes it difficult to maintain and develop effective teachers and improve
student-achievement levels.
McEwan (2002) theorized that the capacity to ascertain and cultivate effective
teachers is a prerequisite for the instructional leader. Therefore, the principals,
especially those in the failing schools of this urban/rural district in HAFIZABAD,
must be cognizant of which leadership behaviors help teachers believe that what they
do makes a difference. Teachers must also be aware of which leadership qualities
assist them in becoming the most effective teachers possible. How leadership
behaviors may influence teacher efficacy will be discussed in Section 2.
Statement of the Problem
Current researchers have suggested that student performance depends on the
effectiveness of the teacher (Kinsey, 2006). Kinsey(2006) noted “Administrative
preparation programs must begin investing more time in making their candidates
aware of the importance of teacher efficacy and the need to adequately nurture our
new teachers in the profession” (p. 159). A gap exists in the knowledge about which
leadership behaviors influence teacher efficacy.
According to the March 6, 2002 Federal Register, student success is
determined by the chance that a student will attend a school that has highly qualified
teachers and principals (p. 10166). This premise requires teachers to have a vision
about what really affects student success, and teachers must be told consistently that
the work they do is vital (Di Giulio, 2004).Efficacious teachers are those who believe
they will be a medium for student success (Di Giulio, 2004; Hansen, 2006). A
teacher’s sense of self-efficacy is consistently related to student achievement. If
teacher self-efficacy is required to increase student achievement, researchers need to
discover the qualities or behaviors of principals that can improve teacher efficacy.
Those qualities and behaviors must be capitalized on and honed in principals in order
to encourage every teacher to develop the highest sense of teacher efficacy possible.
In this mixed-methods study, the quantitative portion examined the relationship
between principal behaviors (the independent variable) and teacher efficacy (the
dependent variable).
Rammer (2007) noted that NCLB placed performance requirements on
schools, delineating serious penalties for schools that could not meet the
requirements. Principals, as leaders of the school and the gatekeepers for performance
standards, need to understand teacher efficacy and the influence administrators have
on teacher efficacy. For these reasons, I chose to focus on principals’ leadership
behaviors in this study.
Nature of the Study
In this mixed-methods study, I attempted to discover whether a relationship
existed between leadership behaviors of principals and teachers’ self-efficacy in
engaging students, strategizing instructional practices, and managing their classrooms.
I chose to conduct a mixed-methods study to combine qualitative and quantitative
results and to establish validity and credibility (Bryman, 2006). Woolfolk Hoy (as
cited in Shaughnessy, 2004) stated that the concept of teachers’ sense of efficacy
“would benefit from more studies that use both qualitative and quantitative
methodologies” (p. 155). Through this study design, I attempted to address that need.
Originally, I asked all teachers in the four elementary/middle schools in a
single urban/suburban district in New Jersey who had fewer than 5 years of teaching
experience to participate in the study. Teachers who had taught fewer than 5 years are
still forming their sense of efficacy and can be influenced by principal leadership style
and are also likely to be most at risk for attrition (Woolfolk Hoy, 2000). To obtain
enough participants, I had to widen the criteria. The desired sample size was 30
teachers; however, the final sample size obtained was 19 teachers, half of whom had
more than 5 years’ experience.
Participants completed the Teacher Sense of Efficacy Scale (TSES) developed
by Tschannen-Moran and Woolfolk Hoy (2001, previously called the Ohio State
Teacher Efficacy Scale) and the Survey of Supervisory Behavior (SSB) developed by
Bulach, Boothe, and Pickett (1999). Once surveys were completed, I obtained a
purposeful sample of three teachers for interviews. I analyzed the surveys using
Pearson’s product-moment correlations and analyzed the interviews using thematic
analysis.
Research Questions
The essential question of this study was: What are the correlations between
leadership behaviors and teacher perceptions of self-efficacy? Specifically the
quantitative portion of the study included the following sub questions:

RQ1. Is there a correlation between teachers’ perceptions of principals’
behaviors in the human-relations domain and their sense of efficacy in
student engagement?
instructional strategies?
classroom management?
behaviors in the trust/decision-making domain and their sense of
efficacy in student engagement?
efficacy in instructional strategies?
efficacy in classroom management?
behaviors in the instructional-leadership domain and their sense of

behaviors in the control domain and their sense of efficacy in student
engagement?
behaviors in the control domain and their sense of efficacy in
behaviors in the conflict domain and their sense ofefficacy in student
engagement?
behaviors in the conflict domain and their sense ofefficacy in
Research Hypotheses
Bulach et al. (1999) designed an instrument to measure behaviors principals
practice and how those behaviors affect teachers. The instrument consists of five
factors -human relations, trust/decision making, instructional leadership, control, and
conflict -and I used these factors to develop the research hypotheses.
The following were the 15 hypotheses for this study:
H01: There is no statistically significant correlation between teachers’
perceptions of principals’ behaviors in the human-relations domain and
their sense of efficacy in student engagement.
Ha1: There is a statistically significant correlation between teachers’
their sense of efficacy in instructional strategies.
their sense of efficacy in classroom management.

perceptions of principals’ behaviors in the trust/decision-making
domain and their sense of efficacy in student engagement.
domain and their sense of efficacy in instructional strategies.
domain and their sense of efficacy in classroom management.

perceptions of principals’ behaviors in the instructional-leadership
perceptions of principals’ behaviors in the control and their sense of
efficacy in student engagement.

perceptions of principals’ behaviors in the control domain and their
sense of efficacy in student engagement.
perceptions of principals’ behaviors in the control domain and their
sense of efficacy in instructional strategies.
I used SPSS to analyze the results of the TSES. I used the Pearson product
moment correlation to measure relationships between leadership behaviors and
teacher-efficacy scores. The qualitative data analysis employed the open-coding
method to code for themes describing the relationship between eachleadership
behavioral domain (human relations, trust/decision making, instructional leadership,
control, and conflict) and teacher self-efficacy (engaging students, strategizing
instructional practices, and managing classrooms). This analysis provided the
foundation for the narrative and descriptive portion of the study. I also employed data
triangulation, member checking, and clarification of researcher bias to increase
confidence in the findings.
Purpose of the Study
The purpose of this mixed-method study was to ascertain the leadership
behaviors associated with the development of personal teacherefficacy. The predictor
variables in this study were the leadership behaviors displayed by principals: trust,
human relations, conflict, control, and instructional leadership. The outcome variables
in this study were the perceived effect those behaviors have on personal teacher
efficacy (engaging students, strategizing instructional practices, and managing
classrooms; Tschannen-Moran &Woolfolk-Hoy, 2001). Little is known about the

connection between principal’s behaviors and a teacher’s sense of efficacy (Elliot,
2000; Hipp, 1996; Shaughnessy, 2004; Staggs, 2002). I examined the behaviors of
principals that are associated with teachers’ self-efficacy beliefs.
Theoretical Framework
Teacher Efficacy
Student success depends on effective teachers (Ashton & Webb, 1986). Two
of the four schools from the urban/suburban district highlighted in the study have not
made AYP based on the standards established by NCLB. It is imperative to find ways
of improving teacher efficacy. Using the wealth of information about the construct of
teacher efficacy, Tschannen-Moran et al. (1998) recognized the need to make sense of
the work already published. I used the integrated modelof teacher efficacy developed
by Tschannen-Moran and Woolfolk-Hoy (2001) in this study. The Tschannen-Moran
model of teacher efficacy has three components: efficacy in student engagement,
efficacy in instructional strategies, and efficacy in classroom management.
Leadership Behaviors
Efficacy in student engagement starts with the belief that all children can learn
(Guskey, 1988). An efficacious teacher does not cease to engage the weak student
(Ashton & Webb, 1986). Efficacious use of instructional strategies includes using
different types of strategies, adapting lessons, facilitating learning, and attempting
new techniques (Gibson & Dembo, 1984; Guskey, 1988; Ross, 1992). The third and
final component of teacher efficacy is classroom management. Efficacious teachers
create an atmosphere that is beneficial to the needs of all learners by being prepared
and systematic, while adapting to the needs of the learner (Allinder, 1994).
The efficacious teacher is also able to use supervision techniques that
encourage students to be more self-sufficient (Ross, 1992). The ways principal leaders
influence the development of teacher efficacy are not well understood. Bulach,
Boothe, and Pickett (2006) observed that although many aspects of leadership have
been studied in relationship to teacher efficacy, few studies offered results that would
allow principals to develop their leadership behaviors. The authors also noted that
many studies were based on principal self-reports and thus may have resulted in an
overly positive response set. Bulach et al. (2006) constructed a model of principal
leadership based on teacher ratings of principals that targets five major areas of
principal behavior: human relations, trust/decision making, instructional leadership,
control, and conflict.
According to numerous researchers (Harrill, 1990; Harrison, 1993; Heller,
2002; Sass, 1989) behaviors in the human-relations area, such as caring and
interpersonal communication skills, are central to good leadership. Trust/decision
making is another factor that connects to leadership abilities. If teachers feel the
principal cannot make a good decision, then it follows that teachers will not trust them
(Bulach et al., 2006). Instructional leadership is another important influence on
teacher efficacy. According to Bulach, Berry, and Williams (2001), less than 50% of
teacher’s surveyed felt principals are aware of what was happening in the classroom
and able to provide them with positive feedback. Control is another aspect of principal
behavior that may have an effect on teachers. When principals talk about myschool
and my teachers, it sends a message that the principal owns the school. Teachers
resent this and feel the principal has an inflated sense of self. The final area included
in the model of principal behavior is conflict. Principals are quite likely to avoid
conflict. Glickman (2002) noted that if principals are going to make changes and
move forward, they must be prepared to address conflict. Furthermore, principals
need to value teachers, yet call them to higher standards when a behavior needs to be
changed (e.g., being consistently late to work).
Because the local school district in this study hasa high teacher-attrition rate
and two of the four schools have not met NCLB requirements, educators need to look
for ways to improve teacher retention and student outcomes. Teacher efficacy has
been linked to teacher retention (Brown & Wynn, 2009) and student performance
(Ashton & Webb, 1986). Therefore, I explored ways principal leadership influences
teacher efficacy in engaging students, strategizing instructional practices, and
managing classrooms.
Definition of Terms
The following definitions apply to this study:
Leadership behaviors fall into five major domains, denoted by the SSB
(Bulach et al., 1999).
1. Positive human relations include skills such as calling people by name,
using eye contact, having a caring attitude, interacting with staff, and
including staff in decision making (Bulach et al., 2006).
2. Trust/decision making is characterized by five factors: (a) character, (b)
ability, (c) truthfulness, (d) confidentiality, and (e) predictability of others
in the group (Bulach et al., 2006). Trust/decision-making skills include
listening to both sides of a story, not gossiping, and carefully thinking
through decisions (Bulach et al., 2006).

3. Instructional leadership skills involve vision, knowledge of curriculum,
accountability, and feedback (Bulach et al., 2006).
4. Control refers to behaviors such as principals sending a message that the
teachers and the building belong to them, assigninga duty during a
preparation period, assigning too much paperwork, and using the words I
and my too often (Bulach et al., 2006).
5. Conflict refers to behaviors such as being afraid to question superiors,
assigning responsibility elsewhere instead of dealing with an issue,
showing favoritism, and having double standards (Bulach et al., 2006).
Self-efficacy is a motivating feature—a specific assessment of the capacity
to successfully complete an assignment—and is shaped through mastery
experiences, explicit experiences, and public and verbal influences
(Bandura, 1977). Walker (2003) simplified self-efficacy to “the belief that
individuals have that they can do something like read a book, write a poem
or dance” (p. 174).
Teacher efficacy refers to teachers’ beliefs in their ability to bring about
necessary results (Tschannen-Moran et al., 1998). Tschannen-Moran et al. believed
that teacher efficacy could be partitioned into three subareas: engaging students,
strategizing instructional practices, and managing classrooms:
1. Engaging students refers to teacher behaviors that show how much the
teacher is willing to do to engage students, help them think critically, and
motivate them to show an interest in learning (Tschannen-Moran et al.,
1998).
2. Strategic instructional practices refers to behaviors that show how well a
teacher can respond to difficult questions, gauge student comprehension of

what is being taught, and craft good questions for their students
(Tschannen-Moran et al., 1998).
3. Managing classrooms refers to how well a teacher can control disruptive
behavior in the classroom, make clear the expectations for student
behavior, and establish routines so activities run smoothly (Tschannen-
Moran et al., 1998).
Assumptions, Limitations, Scope, and Delimitations
This study was built on a few assumptions. I assumed participants would respond
to the surveys and interviews as honestly and thoroughly as possible. I also assumed
that principal leaders could affect teacher efficacy and thereby how teachers practice
their craft. Finally, I assumed certain behaviors cross all leadership styles.
The study was limited to a sample from a single district in New Jersey. Therefore,
the results may not be generalizable to teachers and principals in other districts, and
other geographical regions. An additional limitation is that even though
confidentiality was assured, teachers in the schools may have been waryof responding
to questions regarding the principals at their school. All responses to the
questionnaires were coded in such a way as to ensure anonymity and confidentiality.
The study was also limited because it was delivered through the Web, allowing for
the possibility of misinterpretation of the directions and questions. The use of the Web
prohibited participants from seeking any clarification regarding the survey process
and mitigated the ability to control distractions or collegial influence.
The scope of the study was four elementary/middle schools in one urban New
Jersey school district. It was delimited in several ways. First, I only measured the
three types of teacher efficacy represented by the TSES. Second, leadership behaviors
were delimited to those measured by the SSB; leadership styles were outside the
scope of the research.
Significance of the Study
Researchers and practitioners may benefit from this study. This study defined a
new direction in the research on efficacy that could focus experts on principal
behaviors improve the way teachers teach. When teachers are ineffective, students are
less successful. Researchers need to know why teachers are ineffective and what can
be done to increase their efficacy.
Practitioners may profit from this study as well. I described new paradigms for in
service programs targeted at leadership behaviors that promote efficacy. Entry-level
teacher-mentoring/induction programs might include training on how to find
administrators or colleagues who will support/promote teacher efficacy.
Administrator-consciousness level may be raised about their role in teacher
efficacy and valid means to measure it. Principals may use the data for self-reflection
and improvement. I provided principals in the study with an overview of the efficacy
level of their newer faculty. This step may help the principal focus on certain areas of
professional development.
Finally, I provided the superintendent in the urban New Jersey school district
where I conducted the study a summary of the results. Because efficacy is a concern
for principals, especially in urban districts, it is important that results were shared
with principals to aid in deciding whether a strategic plan is needed to focus on how
principal-leadership behaviors influence teacher efficacy. The relationship can be

used to design individual principal-improvement plans and professional development
targeted at leadership behaviors that were shown to have a relationship with teacher
efficacy in the study.
The results can also be shared with teachers at the school level. Teachers in the
underperforming New Jersey urban district referenced in this study must be involved
in the conversation about their own teacher efficacy. Teachers need to be aware of
their own efficacy level and what they can do to increase efficacy, especially in
engaging students, strategizing instructional practices, and managing classrooms.
Findings from the current study may be used to address the local problem of teacher
attrition and student underachievement by suggesting ways to improve teacher
efficacy that, in turn, may improve teacher retention and student achievement.
Implications for Social Change
Collins (2009), educator and winner of the prestigious National Humanities Medal
in 2004, believed that when students were unsuccessful, the teacher was unsuccessful,
and that educators must fix the teacher first before fixing the student. Collins
encapsulated the reason teachers decide to dedicate their lives to teaching: they want
students to succeed. Although no teacher explicitly sets out to promote student failure,
lack of teacher efficacy can result in inappropriate attitudes and behaviors that are
likely to be conveyed to students. The result is that students believe the teacher thinks
they are not capable of learning.
If the level of teacher efficacy becomes one of the standards for teacher
accreditation, leaders will be encouraged to develop behaviors that support advanced
stages of teacher efficacy. Then, as Collins (2009) said, they will be fixing
themselves. This paradigm shift in the field of education will lead to a higher ability
to engage students, strategize instructional practices, and manage classrooms, and
these in turn will promote student success. In an educated society that is moving from
an industrialized economy to a knowledge economy, it is vital that today’s students
succeed. Students must possess the skills essential to ensure future competitiveness in
the world and maintain the quality of life of all Americans.
In 1991, Hilliard believed that teachers did not accept the challenge to educate all
children, working to build on students’ strengths so each could succeed. To ensure
that every child develops to their peak potential, the educational community must
focus on policies and procedures that increase teacher efficacy and leadership
behaviors proven to increase efficacy. The results of the study showed positive
relationships between the principal behavior measures and teacher self-efficacy.
Therefore; educators have additional evidence to increase teachers’ efficacy by
modifying specific principals’ behaviors. The focus on improving teacher efficacy
may lead to an increase in student learning, and will help prepare future citizens to
take on the challenges of a global knowledge society.
Summary
In this study, I examined the behaviors of administrative leaders who influence
teacher efficacy. In Section 1, the groundwork for this research study was established.
There is little understanding of which leadership behaviors influence teacher efficacy.
The purpose of the study was to explore the relationship between principals’
behaviors and teacher efficacy. Theories of self-efficacy (Bandura, 1977), teacher
efficacy (Tschannen-Moran & Woolfolk-Hoy, 2001), and a mode lof principal
leadership (Bulach et al., 2006) served as the conceptual framework for this study. I
used a mixed-method study to explore the correlations between leadership behaviors

and efficacy, and to interview teachers about those relationships. Section 2 will
include the theoretical framework and a review of the literature that lays the
groundwork for this study. In Chapter 3, I will describe the research methods and
procedures used for the completion of this study.

Chapter 2: Review of the Literature
Introduction
The purpose of this mixed-method study was to ascertain the leadership behaviors
that aid in the development of personal teacher efficacy. This section includes an
overview of leadership theories, social-learning theories, leadership scales, and the
concept of teacher efficacy. Then, I examine the relationship between leadership and
efficacy. The section concludes with a synthesis of the literature, noting strengths,
weaknesses, gaps in the literature, and the significance of the present study.
I conducted the literature review using various databases, peer-reviewed journals,
books, and professional journals. Searches in EBSCOhost, ProQuest, Sage, and ERIC
databases used the following keywords: leadership, principal leadership, educational
leadership, leadership behaviors, teacher efficacy, leadership scales, and teacher
efficacy scales. Questia Online Library, Google Scholar, and the Teacher Reference
Center Database resources were also used. The literature review spanned the years
1920 to 2011 but focused primarily on literature from the past 5years.
An article published in 1998 by Tschannen-Moran et al. firmly established the
constructs of teacher efficacy and that teacher efficacy was ready to stand on its own
merits. Since then, thousands of articles have been published on the subject of teacher
efficacy. However, when searching ERIC, Sage, and Teacher Reference Center using
the terms teacher efficacy and principal, only one was returned. The goal of that study
was to assess previous research at a new level by exploring the concept that teacher
efficacy, collective efficacy, and principal efficacy might be able to predict teacher
commitment (Ware & Kitsantas, 2007).

Forty-five results appeared when I entered the term teacher efficacy into the title
field in ERIC. Of those 45 articles, seven focused on efficacy beliefs (Charalambous,
Philippou, & Kyriakides, 2008; Chong, Klassen, & Huan, 2010; Gencer & Cakiroglu,
2007; Knoblauch & Woolfolk Hoy, 2008; Nunn & Jantz,2009; Rethlefsen & Park,
2011; Tsouloupas, Carson, & Matthews, 2010). Nine of the 45 articles focused on
teacher efficacy in relationship to mathematics and science (Angle & Moseley, 2009;
Bruce & Ross, 2008; Gresham, 2008, 2009; Marat, 2007; Puchner & Taylor, 2006;
Richardson & Liang, 2008; Swars, Daane, & Giesen, 2006; Yu-Liang, 2010). Of the
articles in the search, 8% studied efficacy across the globe (Cheung, 2006, 2008;
Chong et al., 2010; Dunham & Song’ony, 2008; Faleye, 2008; Kotaman, 2010). Two
journals published articles on teacher burnout (Pas, Bradshaw, & Hershfeldt, 2010;
Skaalvik & Skaalvik, 2007) and three researchers collaborated to explore the beliefs
of teachers in the Yukon (Klassen, Foster, & Rajani, 2009). Yet, no research emerged
on how the leadership behaviors of the principal affected teacher efficacy.
Between 1998 and 2009 the number of articles in peer-reviewed journals
increased significantly, along with the use of varied methodologies. In 2007, Ware
and Kitsantas used the 2005 School and Staffing Survey to study whether efficacy
beliefs predicted allegiance to the profession. They discovered that principal feedback
was important to teacher efficacy. They also found thatthe ability to garner the
principal’s backing, impact decision making, and skillfully manage the classroom
were related to teacher willingness to remain in the profession (Ware & Kitsantas,
2007).
In 2010, Klassen, Tze, Betts, and Gordon looked at 218 empirical studies
published between 1998 and 2009 to review the stateof affairs of research on teacher
efficacy. They found that only 14.7% of studies completed between 1998 and 2009
used mixed methods (Al-Fadhli & Singh, 2006; Gado, Ferguson, & Van t Hooft,
2006; Klassen et al., 2010; Marcos, 2008; Onafowora, 2005; Williams, 2009).
However, not one researcher focused on the principal as a source of efficacy. They
suggested that the sources for teacher efficacy are still in need of further investigation
to clarify how it develops. The Klassen et al. (2010) review also suggested that if
researchers understood the source of efficacy, it could lead to more awareness of how
to augment the self-efficacy of teachers. One area for investigation is the role of
principal leaders in promoting teacher efficacy.
Leadership Theories
Theorists put forth five models of leadership. The great-man theory (Chemers,
1997) and trait theory (Kohs & Irle, 1920) are based on the belief that leaders are
born, not made. Great-man theory states that only men can become leaders, in contrast
to trait theory, which suggests that as long as one was born with certain traits
(assertiveness, confidence, and willingness to assume responsibility) a person of
either gender could become a leader.
Transformational and transactional leadership are distinguished by the manner in
which a leader interacts with followers. Transformational leaders (Rost, 1993) work
with their followers, together pursuing a higher moral purpose. Transactional leaders
reward or punish their followers based on their performance. Situational-leadership
theory (Hersey & Blanchard, 1969) presumes that one type of leadership does not
work for every situation. All five theories establish that there is some intrinsic quality
that defines a leader. The theories differ in their definitions of extrinsic manifestations
of being a leader.
These major leadership theories apply to leaders in general, whereas the
leadership theory used in this study focuses on educational leadership. In the 1980s,
researchers showed that principals who focused on curriculum and instruction usually
led effective schools. These trends led to instructional leadership as a new area of
leadership. Glickman (1985) defined the five primary tasks of instructional leadership
as teacher support, group development, professional development, curriculum
development, and action research. Instructional-leadership theory provided the
foundation for this study.
Leadership Scales
A number of leadership scales preceded the scale developed by Bulach et al.
(1999) that I used in the current study. Blake and Mouton (1964) and Hersey and
Blanchard (1969) created leadership inventories that focused on people rather than
tasks. The Barrett-Lennard Relationship Inventory (1964), based on the work of
Rogers, was designed to measure the opinions of either member in a two-person
relationship and contained scales for empathy, congruence, regard, and un
conditionality of regard (King, 2001). Halpin and Winer (1952) developed the first
instrument focused solely on measuring instructional-leadership behaviors (Bulach et
al., 2006). Bass and Avolio (1992) developed a survey that rated the frequency and
behavior of the leader. Leithwood developed the Nature of Leadership Survey (1993),
and then collaborated with Jantzi in 1997 to create the Principal Leadership
Questionnaire. This questionnaire collected data about teachers’ perceptions of their
principals’ leadership behaviors. These two surveys were designed only for teacher
responses (Elliot, 2000; Ryan, 2007). Wirt and Krug (1998) designed a new scale
based on cognitive scales. The instrument was adapted so teachers could report how
they perceived the principal’s leadership actions (Bulach et al., 2006).
Bulach et al. (1999) saw a need to focus on behaviors that instructional leaders
could change; therefore, they set out to develop an instrument that highlights
leadership behaviors that could be improved. In this study, I used the instrument they
developed as a result of their work. Bulach et al. (1998) sampled teachers enrolled in
a master’s program to discover what mistakes their principals made. Bulach et al.
(1999) then used that list of mistakes to develop a survey instrument. Responding to
the survey were 208 educational-leadership graduate students; a factor analysis of the
survey yielded nine major areas accounting for 64% of the variance. Factors that
accounted for much smaller variances were combined and five categories were
created: human relations, trust/decision making, instructional leadership, control, and
conflict. Bulach et al. (2006) then delineated competencies in each of these five
categories.
Human relations are an area that fosters the advancement of self-assurance and
openness between the leader and the followers (Bulach et al., 2006). Human-relations
skills include calling people by name, using eye contact, having a caring attitude,
interacting with staff, and including staff in decisions. Research by Martin (1990),
Deluca, Rogus, Raisch, and Place (1997), Harrill (1990), Hutchison (1988), Jolly
(1995), and Rouss (1992) under laid the notion of Bulach etal.(2006) that human
relations and interpersonal skills are aptitudes needed by successful leaders.
Trust between people was characterized by five factors: an assured reliance or
confident dependence on the (a) character, (b) ability, (c) truthfulness, (d)
confidentiality, and (e) predictability of others in the group (Bulach, 1993).

Trust/decision-making skills include listening to both sides of a story, not gossiping,
and carefully thinking through decisions.
Instructional-leadership skills involve vision, knowledge of curriculum,
accountability, and feedback. The behaviors connected to instructional leadership
were involvement in the daily working of the classroom and providing meaningful
feedback to teachers. When principals were conducted staff development, held
instructional conferences, and developed teacher reflection, teachers’ feelings and
attitudes were dramatically impacted (Blasé & Blasé, 2004).
In the area of control, principals often used the words “my school” or “my staff.”
Principals had difficulty delegating assignments and including teachers in major
decisions. Teachers reported resenting this exclusive language and felt it gave the
impression to staff that principals felt they “owned” the school (Bulach, 1993). Blasé
and Blasé (2004) also noted teachers felt principals were being controlling when they
limited teacher participation in decision making, directed instructional areas, and
influenced classroom instruction.
The last area in human relations and interpersonal skills is conflict. The conflict
scale includes behaviors such as the ability to keep a confidence, fairness, and support
for teachers. Bulach (1993) found that principals often went too far in avoiding
conflict and missed an opportunity to use conflict as a positive force to improve the
supervisory climate of the school.
Social-Learning Theory and the Concept of Efficacy
Social-cognitive theory is the academic foundation of self-efficacy. Bandura
(1977) showed that many learning theories of the day were lacking something crucial:
the element of self-belief. Since that article was published, researchers have shown
that self-efficacy beliefs reach and influence every aspect of people’s lives. Graham
and Weiner (1996) believed self-efficacy had been proven to forecast behavioral
effects more than has any other motivational idea.
Teacher Efficacy
A 1976 study by the Rand organization, inspired by an article by Rotter (1966),
found that teacher efficacy was associated with differences in reading scores for
minority students (Tschannen-Moran et al., 1998). These results sparked an interest in
a new educational research area: teacher efficacy. Efficacy is comprised of three
components: general teacher efficacy, personal teacher efficacy, and teacher efficacy.
General teacher efficacy refers to the belief that the influence ofa student’s home life
outweighs any influence teachers can bring to bear in the classroom (Ashton, Webb,
& Doda, 1982).
Personal teacher efficacy refers to the confidence teachers have in their ability to
teach any student, irrespective of their heritage and place in society. Teacher efficacy
refers to teachers’ inherent desire to engage students and increase learning, even with
disaffected students (Armor et al., 1976). Ashton et al. (1982) combined
psychological aspects and educational aspects in a multidisciplinary study that found
efficacy attitudes are fluid and that efficacy is critical to equalizing opportunities for
all students. Rich, Lev, and Fischer’s (1996) findings supported earlier findings that
teacher efficacy (or general teacher efficacy) and personal teacher efficacy were
independent of each other. In 1993, Hoy and Woolfolk found that principal influence
predicted personal efficacy, and when teachers perceived the principal to be
influential with superiors, they felt more efficacious.

Teacher efficacy is a teacher’s belief in their ability to affect student learning
(Ashton et al., 1982). Teacher efficacy includes three distinct areas that contribute to
the overall teacher-efficacy construct: student engagement, instructional strategies,
and classroom management (Ross, 1992). A strong sense of efficacy in the area of
student engagement starts with the belief that all childrencan learn (Guskey, 1988). As
a result, an efficacious teacher does not cease to assist thevulnerable student (Ashton
& Webb, 1986).
The second area in which efficacious teachers tend to be strong is instructional
strategies. They create new strategies, adapt lessons, facilitate learning, and attempt
new techniques (Gibson & Dembo, 1984; Guskey, 1988; Ross, 1992). This type of
teacher ensures that clear and attainable expectations are mutually established and that
they teach all the necessary skills and strategies (Alderman, 1990). The third and final
component of teacher efficacy is classroom management. Efficacious teachers create
an atmosphere that is beneficial to the needs of all learners by beingprepared and
systematic, yet lithe enough to adapt to the changing demands of the learner (Allinder,
1994). The efficacious teacher is also able to use supervision techniques that
encourage students to be more self-sufficient (Ross, 1992).
Teacher-Efficacy Scales
There has been a persistent and progressive interweaving of teachers’ sense of
efficacy and its measurement (Heneman, Kimball, & Milanowski, 2006). This
development has involved two separate yet linked theoretical strands that have
contributed to an ambiguous notion of teacher efficacy (Tschannen-Moran et al.,
1998). One strand, based on Rotter’s (1966) notion of locus of control focused on the
concept of internal and external control. Teachers who believed control was external
felt the environment overwhelmed their ability to be effective. In contrast, educators
who believed they could affect student achievement had internal control. The second
strand, based on Bandura’s (1977) concept of self-efficacy,focused on the idea that
efficacy described teachers’ assessment of their ability to effect student change rather
than about control.
Early measures of efficacy were the Teacher Locus of Control and the
Responsibility for Student Achievement, designed byRose and Medway (1981). The
Webb Scale, focused on Rotter’s (1966) idea of external and internal control. Gibson
and Dembo developed a tool in 1984 that was based on Bandura’s conceptual
foundations and was more reliable for evaluating teacher efficacy. This scale was
grounded in the belief that self-efficacy was a teacher’s evaluation of their capacity to
produce student change. The Teacher Efficacy Scale (TES) consisted of 30 items with
a 6-point Likert scale, and introduced the terms personal and general efficacy. This
scale was the standard for measuring teacher efficacy until 1993 when Hoy and
Woolfolk questioned whether the TES actually measured an individual teacher’s
sense of efficacy. The Gibson and Dembo instrument was popular, however there
were problems conceptually and statistically (Tschannen-Moran & Woolfolk Hoy,
2001). Tschannen-Moran and Woolfolk Hoy (2001) saw the need to address the
conflict between the idea of content and subject specificity, and refined the instrument
further. Working with a group of graduate students and fusing Bandura’s (1977) scale
with Gibson and Dembo’s (1984) scale, the Ohio State
Teacher Efficacy Scale (now more commonly referred to as the TSES) was born.
Rather than measuring personal and general efficacy, this instrument measures
individual teacher efficacy in engaging students, strategizinginstructional practices,
and managing classrooms. The 24-item (long-form) and the 12-item(short-form)

instruments have been found to be valid and reliable. The TSES surpasses other tools
for measuring teacher efficacy: it is unified, consolidates previous research and
measures abilities teachers deem central to good teaching (Tschannen-Moran &
Woolfolk Hoy, 2001).
Related Studies on Teacher Efficacy
Fry (2009), Erdem and Demirel (2007), Yost (2006), Palmer (2006), and Utley,
Bryant, and Moseley (2005) investigated procedures to promote higher levels of
teacher efficacy in 1st-year teachers. Swackhamer, Koellner, Basile, and Kimbrough
(2009) speculated that courses meant to increase content knowledge (such as
mathematics) would increase the efficacy of teachers who were inthe classroom for
more than 1 year. Yeo, Ang, Chong, Huan, and Quek (2008) focused on teachers and
low-achieving adolescents. Yeo et al. investigated engaging students, strategizing
instructional practices, and managing classrooms and their link to teacher
characteristics and teacher–student relationships. The study, conducted in Singapore,
focused on teachers of low-achieving adolescents who are at risk of failing and dropp
in gout. The 55 teachers in the study ranged in age from 23 to 55. The professional
experience of the subjects spanned 39 years. The authors investigated whether the
demographic profiles (age, years of experience, gender, and number of levels taught)
of a teacher coincide with teachers’ efficacy beliefs, and to what degree teacher
variables and teacher–student interactions foretell teacher-efficacy beliefs (Yeo et al.,
2008).
Yeo et al. (2008) used the 24-item TSES (Tschannen-Moran & Woolfolk Hoy,
2001), The TSES provides scores for three discrete areas of the teacher–student
relationship: instrumental help, satisfaction, and conflict. Teachers rated their

individual relationships with students. Higher ratings in the areas of influential help
and contentment meant that teachers saw themselves as aresource for students (Yeo et
al., 2008). In contrast, high scores in conflict inferred that teachers saw their dealings
with students negatively. Ang (2005) used the Teacher–Student Relationship
Inventory to evaluate teachers’ acuity about the value of their rapport with students in
Grades 5–12.
Researchers found no notable differences when they compared instructional
strategies, classroom management, and student engagement to gender and number of
levels taught (Yeo et al., 2008). However, disparities occurred between beliefs for
neophyte teachers and veteran teachers in all threedimensions of self-reported
efficacy: engaging students, strategizing instructional practices, and managing
classrooms. The researchers found that more experience in the fieldled to higher
levels of teacher efficacy. The study by Yeo et al. is notable because not only did they
use the TSES, the instrument that I used in the present study, but also because the
findings regarding the role of experience on teacher efficacy implied thatthe influence
of leaders on teacher efficacy is most critical during the novice years of teaching. This
finding prompted me to focus on a limited number of years of teaching. Therefore for
the current study I initially attempted to survey teachers with fewer than 5 years of
experience.
Relationship between Leadership and Self-Efficacy
Chester and Beaudin (1996) noted the limited research that examined the factors
contributing to changes in efficacy beliefs of teachers: when supervisors attended to
instructional dimensions of teachers’ roles, efficacy improved. Blasé and Blasé (1999)
found few studies that studied teachers’ viewpoints on everyday instructional

leadership actions and the import of those actions on teacher behavior. Angelle
(2010); McGuigan and Hoy (2006); Day, Sammons, Hopkins, Leithwood, and
Kington (2008); McCollum, Kajs, and Minter (2005); Kelley, Thornton, and
Daugherty (2005); Korkmaz (2007); Crum, Sherman, and Myran (2010); and Nash
(2010) studied the effect the principal had on the school climate and student
achievement. Hipp(1996), Elliot (2000), King (2001), Staggs (2002), Ebmeier (2003),
Ross and Gray (2004), Nir and Kranot (2006), Ryan (2007), and Griffin (2009),
reviewed below, delved into the relationship between leadership and teacher efficacy.
Of nine studies that focused on the relationship between leadership and teacher
efficacy, six looked at general teacher efficacy, teacher self-efficacy, and specific
indicators of efficacy (Elliot, 2000; Hipp, 1996; King, 2001; Leithwood, 1993; Ryan,
2007; Staggs, 2002). Both Hipp (1996) and Elliot (2000) conducted mixed-methods
studies using the Nature of Leadership Scale (Leithwood, 1993) and early forms of
the TES (Gibson & Dembo, 1984 adapted by Hoy and Woolfolk, 1993). Hipp’s study
provided no statistical data showing a relationship between principal support and
general teacher efficacy, however there was some evidence in the interviews
suggesting a relationship. Elliot’s study showed one significant relationship between
teachers’ perceptions of leadership behavior and efficacy: that relationship was
between general teacher efficacy and individualized support from the principal. Elliot
believed that the study left some questions unanswered, such as why there is no
relationship between personal efficacy and principals’ leadership, and what leadership
behaviors affect teachers at all levels of teaching.
Six years after Elliot’s (2000) findings were published, Ryan (2007) completed a
very similar study with different results. Ryan used the TSES (Tschannen-Moran &
Woolfolk-Hoy, 2001) and the Principal Leadership Questionnaire (Jantzi &

Leithwood, 1996). Ryan concluded no relationship existed between teachers feeling
supported and their sense of efficacy. These results did not support Elliot’s findings
and Ryan recommended further investigation of this connection.
King (2001) and Staggs (2002) used versions of the TES to examine the
leadership–teacher efficacy connection. King used the TES and the Barrett-Lennard
Relationship Inventory to see if there was a connection between teacher–principal
interpersonal relationships and teacher efficacy. King found no significant
relationship between teacher–principal associations and general teacher efficacy, but
did notice a statistically significant association between teacher views of teacher–
principal associations and personal teacher efficacy. Staggs’s study asked whether
teachers’ views of their principal’s leadership behaviors had an effect on teacher
efficacy. Using the Ohio Health Inventory (Hoy & Tarter, 1997) and TES, Staggs
found that teacher views of principal leadership correlated with general teaching
efficacy but not with personal teaching efficacy.
Nir and Kranot (2006) used the TES (Gibson & Dembo,1984) and the Multifactor
Leadership Questionnaire (Avolio, Bass,& Dung, 1996) and found a relationship
between transformational leadership and personal teacher efficacy. Griffin (2009)
used the TSES (Tschannen-Moran & Woolfolk Hoy, 2001) and the Multifactor
Leadership Questionnaire (Bass & Avolio, 1992) to explore the connection between
self-efficacy and opinions of the school principal’s leadership style. Griffin
established that only transformational leaders affected classroom management. Both
studies looked at three types of leadership and both studies found that
transformational leadership did impact efficacy, yet the impact was in different areas.
Ebmeier (2003) used path modeling derived from Tschannen-Moran et al. (1998)
and showed that school principals play an importantrole in the development of teacher
efficacy. Ross and Gray (2004) studied transformational leadership and the impact it
had on teacher efficacy. They developed a new instrument that combined items from
six previous studies. The instrument consisted of 46 Likert-type statements based on a
6-point response scale. Their results showed that transformational leadership directly
affected collective teacher efficacy.
Synthesis
Nine studies in the literature focused on the relationship between school leaders
and teacher efficacy. The studies varied in sample size (34 to 3,074), type of school
district (urban and suburban), and geographic location (areas of the United States,
Canada, and Israel). Furthermore, participants in the studies varied in grade levels
taught, with teachers in elementary, middle, and high schools. Many studies used
versions of the TSES. However, none of the studies used the measure of leadership
behavior used in the present study. Two of the nine studies (Griffin, 2009; Ryan,
2007) used the TSES, the measure of teacher efficacy used in this study. Most studies
used the TES, the precursor to the TSES. All nine studies relate to the purpose of this
study in that they all examined the link between leadership behaviors and teacher
efficacy. However the results were inconsistent, due perhaps to differences in samples
and measures of leadership. All the studies used quantitative surveys and three of the
nine studies used mixed methods to answer the research questions. Because the results
have not been easily replicated, the current research used a mixed-methods design to
geta more complete picture of relationships and to use an established method in this
field of study (see Table 1).

Table 1
Studies Using Teacher Efficacy Scales and Leadership Scales

Ryan (2007) and Griffin (2009) looked at the three subscales of the TSES; yet
they used different leadership scales. Griffin useda causal-comparative design,
whereas Ryan used mixed methods. Ryan found no relationshipbetween feeling
supported and a teacher’s sense of efficacy; Griffin found that classroom management
was affected by leadership style. Griffin’s findings warranted further exploration of
the link between leadership style and teacher efficacy.
Although evidence exists, findings are inconsistent as to whether there is a
relationship between principals’ behaviors (negative or positive) and teacher’s sense
of efficacy; no prior research used the SSB, developed by Bulach et al. (1999). This
instrument seemed well suited to measure principal-leadership behaviors and answer
the research questions for this study.
Literature Related to the Method
Two of the nine studies discussed above used mixed methods to examine the
link between principal behavior and teacher efficacy. Hipp (1996) and Elliot (2000)
conducted mixed-methods studies using the Nature ofLeadership Scale (Leithwood,
1993) and early versions of the TES (Gibson & Dembo, 1984 adapted by Hoy &
Woolfolk, 1993). Hipp’s study provided no statistical data showing a relationship
between principal support and general teacher efficacy, however there was some
evidence for the relationship in interviews. Elliot’s study showed one significant link
between teachers’ perceptions of leadership behavior and efficacy; that relationship
was between general teacher efficacy and individualizedsupport from the principal.
Elliot believed that the study left some questions unanswered, such as why no
relationship exists between personal efficacy and principals’ leadership, and what
leadership behaviors affect teachers at all levels of teaching. It is conceivable that the
five domains of behavior on the SSB relate to teacher efficacy. The current study
extended the literature by looking at leadership behaviors that have not been
considered in connection with teacher efficacy.
Literature Related to Differing Methodologies
Of the remaining seven studies that considered leadership and efficacy, four
used the survey design (King, 2001; Leithwood, 1993; Ryan, 2007; Staggs, 2002).
Two studies used the pathway model (Ebmeier, 2003; Ross & Gray,2004)) to
visualize the connections between leadership and efficacy. Griffin (2009) used a
causal-comparative model. I found no studies that used a qualitative design. Because
the findings on the links between leadership and teacher efficacy have been
inconsistent and the leadership behaviors measured have varied greatly, it was
necessary to include a qualitative component in the current study.
Summary
This section presented an overview of leadership theories and social-learning
theories relating to efficacy, leadership scales, and the concept of teacher efficacy. A
number of studies addressed relationships between teacher efficacy and aspects of
principal leadership behaviors. I discussed the strengths and weaknesses of these
studies and noted gaps in the literature.
Since the Rand Corporation first identified general, personal, and teacher
efficacy in 1976, researchers have exerted time and effort “capturing an elusive
construct” (Tschannen-Moran & Woolfolk Hoy, 2001, p. 783). After reviewing
multiple studies that used qualitative, quantitative, and mixed methods, it seems
“elusive construct” is still a valid moniker for teacher efficacy. First, researchers
suggested that the concept of teacher efficacy is still in its early stages and needs more
research. Second, many researchers believed that the TSES is a valid and reliable
measure and an important step toward defining that elusive construct. Third, there is
no consensus on whether efficacy is the same at all levels of the educational process
(elementary, middle, and high school).
Finally, there is some discord about what leadership styles, if any, impact
teacher efficacy. In conclusion, research must continue, the TSES should be used to
define the elusive construct; teacher efficacy needs to be defined at all levels, and
leadership styles that make the most impact need to be encouraged.

Chapter 3: Methodology
Introduction
In this section, I describe the sequential explanatory mixed method I used in
this study. I chose this mixed-method approach because the results of prior
quantitative research have been insufficient and need to be elucidated by qualitative
explanations. Johnson, Onwuegbuzie, and Turner (2007) noted that mixed methods
provide the researcher the ability to combine quantitative and qualitative research to
allow for breadth and depth of understanding and verification. Creswell and Plano-
Clark (2011) noted that quantitative trends supported by qualitative narrative tell the
whole story.
The aim of this two-phase sequential explanatory study was to examine the
relationships between principals’ leadership behaviors and teachers’ sense of efficacy.
I took a comprehensive look at human relations, trust/decision making, instructional
leadership, control, and conflict, and how these relate to teachers’ sense of efficacy in
engaging students, strategizing instructional practices, and managing classrooms. In
Phase 1, I used two surveys: one to identify levels of teacher efficacy, and the other to
identify leadership behaviors associated with teacher efficacy. In Phase 2, interviews
with a small subsample of teachers provided further insight into leadership behaviors
and possible links to teacher efficacy. I analyzed the data separately and used
triangulation to explain the quantitative and qualitative findings.
This section will be divided into the following subsections: mixed-method
design, setting and sample, participants’ selection for interviews, quantitative survey,
survey instruments, qualitative interviews, data analysis,protection of participants’
rights, and a summary.
Mixed-Method Design
I employed a mixed-method approach using surveys and individual interviews
to address the research questions of the study. In this explanatory mixed-methods
design, I began by conducting a quantitative phase and followed up with a second
phase (Creswell & Plano-Clark, 2011, p. 1015). The quantitative and qualitative
results were integrated during the analysis phase of the study.
The rationale for using both quantitative and qualitative data was to balance
the limitations inherent in one method with the strengths of the other method. This
design allowed for the integration of quantitative and qualitative analysis during the
interpretive phase. The most important advantage of the concurrent design is that I
could achieve a broader viewpoint by using different methods to obtain data. The
disadvantage to using this approach was that there are no recommendations given on
how I should resolve inconsistencies that came about as a result of using two types of
data (Creswell, 2003).
Setting and Sample
For this study, I selected the district involved because two of the schools failed
to meet NCLB standards. I drew participants in the study from the population of
teachers with fewer than 5 years’ experience in four elementary/middle schools in an
underperforming suburban/urban district in New Jersey. Beginning teachers are
known to be developing their sense of teaching efficacy and may be influenced by
principal leadership. During the study the criteria was relaxed to allow for enough
participants. I used purposeful sampling to select participants for in-depth, follow-up
interviews, once I had collected and analyzed the surveys.
Population and Sample for the Survey
An urban/suburban district in northeastern New Jersey has a population of
2,759 students. The teaching staff has 153 members and administrators number about
20. Two elementary schools house Grades 1–4, one school houses Grades 5–6, and
one school houses Grades 7–8 in the district of interest in this study. The entire
teacher population of those four schools was about 130. Initially, I invited all teachers
in the population with 5 years or fewer of teaching experience to participate in the
survey. However, the criteria were widened to obtain a sufficient sample size.
Because I invited all teachers in the four schools to participate, total-
population sampling was used. The number of eligible teachers who voluntarily
responded to the survey determined the sample size. Approximately 84% of potential
participants were women. About 55 female teachers have 5 years or less experience;
about 9 male teachers have 5 years or less experience. I proposed that 30respondents
who have 5 years or less of teaching experience be surveyed, however, 19 participants
actually completed the survey and half of them had more than 5 years of teaching
experience.
Participant Selection for Interviews
For the interview phase of the study, I used a maximum-variation sampling
technique (Patton, 2002). I proposed to interview 10 participants. However, only four
participants volunteered for the interviews. One participant had the second highest
score on the TSES; one had the second lowest score. The third and fourth participants
had the middle score on the TSES. Three participants had less than 5 years of teaching
experience, and the fourth participant had more than 10 years of teaching experience
although it was noted that the fourth participant saw no relationship between
leadership behaviors and her sense of teacher efficacy, this participant was dropped
from the narrative because her data offered no insights intothe relationships that were
found in the quantitative analysis.
Morse (1995) thought saturation was the most critical part of excellent
qualitative work. Researchers disagree about what saturation means in qualitative
research. Guest, Bunce, and Johnson (2006) suggested that if a group is relatively
homogenous, 6 to 12 participants will likely be enough to reach saturation. Typically,
major themes occur after the sixth interview. Because I was unable to obtain more
than four interviews, I was unable to reach saturation, and the qualitative data is
merely suggestive of themes that might arise with a larger sample.
As I conducted the interviews, it was critical to bracket my passion for the
subject to maintain objectivity. I believe that my 31 years of experience as an
educator enabled me to establish a trusting researcher–participant relationship. At the
beginning of each interview I introduced myself and talked a bit about my teaching
experience. I then allowed interviewees to tell me a bit about their teaching
experience. This brief introduction set the tone for the interview.
Connection to the Research Questions
The essential question of this study was: What are the correlations between
leadership behaviors and teacher perceptions of self-efficacy? The sequential design
involved collecting the quantitative (survey), then the qualitative (interview) data and
combining them during the interpretation of the results. Specifically, the quantitative
portion of the study included the following sub questions:
student engagement?

behaviors in the control domain and their sense of efficacy in student
engagement?
behaviors in the conflict domain and their sense of efficacy in student
engagement?
behaviors in the conflict domain and their sense of efficacy in
behaviors in the conflict domain and their sense of efficacy in
Quantitative Survey
I used two instruments to collect the quantitative data. The TSES (Tschannen-
Moran & Woolfolk Hoy, 2001; see Appendix A) provided data about teacher efficacy.
The SSB (Bulach et al., 1999; see Appendix B) provided information on leadership
behaviors of principals from the teachers’ perspective.
Survey Instruments
I chose the TSES scale because of its validity, reliability, and prolific use in
the educational-research community. The TSES is a 24-item, 9-point Likert-type scale
(Tschannen-Moran et al., 1998). The 24 items are distributed in three subscales
(outcome variables in the study): engaging students, strategizing instructional
practices, and managing classrooms. Each subscale has six items and is scored by
computing the mean for those items. The scale is included in Appendix A.
“Reliabilities for the teacher efficacy subscales were 0.91 for instruction, 0.90 for
management, and 0.87 for engagement” (Tschannen-Moran & Woolfolk Hoy, 2001,
p. 799). The TSES has been widely referenced in the literature and should be the
benchmark for research that aims to quantify teachers’ sense of efficacy (Heneman et
al., 2006). I obtained permission from the authors to use the TSES in this study (see
Appendix C).
I chose the SSB because it allows teachers to rate the leadership behaviors of a
principal rather asking principals to rate their own behaviors. In this study, I
correlated teachers’ perceptions of the leader’s behaviors with their own self-efficacy
and therefore this instrument was the only available instrument that was appropriate.
The SSB (Bulach et al., 1999) consists of 49 behaviors that measure teachers’
perceptions of their principal’s leadership style. Teachers rated how often they see a
principal exhibit a behavior, using a 5-point Likert-type scale that ranges from never
(A) to always (E). The scale consists of five domains (predictor variables in the
study): human relations, trust/decision making, instructional leadership, control, and
conflict. Each participant received a mean subscale score for each subscale. The
instrument’s reliability was confirmed by a correlation coefficient of + .95. The
reliability of the factors ranged from+ .86 to + .81. The instrument met the standard
for construct validity in recognizing principal behaviors that educators like or dislike
(Bulach et al., 2006). I obtained permission to use the SSB from Bulach (see
AppendixD).
The TSES and SSB were administered online using Constant Contact, an
online survey service. Surveys delivered through the Internet raise concerns about
trust (confidentiality or anonymity may often come in to question) and social
desirability (Smyth, Dillman, & Christian, 2007). To establish trust, I assured
confidentiality by assigning an alphabetical code rather than using participants’
names. I addressed social desirability, the tendency of participants to respond in a way
that will be viewed positively by others, by stating in the instructions of the survey
that there is no right or wrong answers.
Question context, the effect of earlier questions/statements on answers to later
questions/statements, is also a concern for researchers (Smyth et al., 2007). I
maintained full control over the order in which the questions were read and processed,
thereby eliminating some contextual effects. The online surveys used the exact same
order as the pencil-and-paper surveys and respondents saw only one statement at a
time. The respondents then had an opportunity to review their answers and make any
changes before they submitted their responses.
How an online survey is viewed by the respondent is subject to the type of
computer on which it is being viewed (Smyth et al.,2007). The computer on which the
respondent views the survey must have the exact same hardware/software as the
computer on which the survey was designed. If the hardware/software is not the same,
there can be variations in the user interface for respondents. To minimize these
variations, respondents accessed the survey throughthe district network so that the
hardware/software and interface were standardized. Thus, the fact that respondents
completed the survey through an online survey service and the school district’s
network reduced the likelihood of having hardware and software issues.
Careful consideration must be given, when designing an online survey, to
minimize the importance of some options over others (Smyth et al., 2007). In other
words, no words or statements should be in bold type, italics, or in a different size or
shape than any other words in the survey. The online survey mimicked the paper
surveys as closely as possible.
The survey data-collection process involved disseminating information about
the purpose of the study, enrolling study volunteers, distributing links to the Constant
Contact website where the two surveys were administered online, and sending
reminders to no responders. I contacted principals of the participating schools to
secure teacher participants who met the criteria for the study. Then I sent an email to
all eligible teachers in the three elementary schools and one middle school in the
district. The introductory email contained my short biography, a description of the

study, an explanation of the importance of the research, information on how the study
could assist participants, step-by-step instructions for interested participants, and a
link to the survey.
The email contained a tracking code that permitted access to the survey. The
tracking code contained the school number and the participant’s number. I assigned
the participants’ numbers in numerical order. The tracking code allowed me to know
exactly how many surveys were completed from each school.
I gave participants 1 month to complete the survey on the Internet. I sent
participants weekly automated email reminders to increase the response rate, once the
online survey was made available (Sue & Ritter, 2007). I advised participants, in the
initial email and in the instructions for the online survey, that they must complete the
survey once they start for their responses to be retained. I advised participants that the
survey would take 10–15 minutes. When participants completed the online survey, I
provided them with a confirmation message thanking them for their participation and
allowing them to download a $5 coupon to a local coffee house. An automated
reminder was sent by email to provide one more opportunity for nonresponding
participants to complete the survey. The survey remained online fora month; then the
online site and all links to it were disabled and no longer could be viewed by
participants.
Survey responses were stored in Constant Contact, and I was the only person
with access to the data. Each participant’s response to the survey items, along with the
means for each subscale of the survey, will be stored in Constant Contact for a year
after acceptance of the dissertation. Because the instruments have been shown to be
valid and reliable, I conducted no specific procedures for validity and reliability.
Qualitative Interviews
I collected qualitative data in the second phase of the study. This phase of the
data collection did not take place until the TSES and SSB scales had been completed
and analyzed. I used interviews to gather more in-depth information about the
associations or lack of associations between leadership behaviors and teachers’ beliefs
about their self-efficacy. The interviews lasted between 15 and 45 minutes. The
interview questions are provided in Appendix E. These questions were field tested
during dissertation-proposal development with a panel of three content experts who
hold doctorates in education. Each was given a copy of the proposed interview
questions and asked to provide written feedback on suggested changes.
I emailed participants, based on their scores on the TSES, to ask if they would
be willing to participate in a follow-up interview at a specific time and location. For
those who agreed to participate, I conducted the interviews in a private room
designated by an administrator in each of the participating school buildings. The
interviews began with some background about me; then I asked the interviewee to
share some background about them. This sharing was meant to put the interviewee at
ease and to establish rapport.
Once the interviews were transcribed and the results analyzed, the qualitative
results were triangulated with the quantitative results during the interpretation phase. I
discuss the data-analysis procedures in the next section.
Data Analysis
Survey Data Analysis

Two surveys were the sources of quantitative data. I planned to use
correlational analysis to discover the relationships between leadership behaviors and
teacher efficacy.
Gravetter and Wallnau (2005) defined correlation asa numerical procedure
used to explain an association between two variables. Pearson product-moment
correlation (Pearson r) was used measure relationships between the five areas of
leadership behaviors (human-relations, trust/decision-making, instructional
leadership, control, and classroom management) and three subscales of teacher
efficacy (engaging students, strategizing instructional practices, and managing
classrooms). I used Effect sizes, which according to Cohen (1992) are the measures of
the strength of the relationship between the variables and the study population, are
used to interpret the Pearson r.An rof .1 is considered a small effect; an rof .3 is
considered a medium effect; and an rof .5 considered a large effect. I reported
descriptive statistics (means and standard deviations) for each scale of the SSB and
the TSES.
Interview Data Analysis
I prepared the qualitative data for analysis by transcribing the interviews. I
read all data to get an overview of the information. After the transcripts were read, the
coding process began. I identified preliminary themes, highlighted relevant
quotations, and began to group the material into categories. Because of the small
number of interviews, although saturation was not reached, there was enough
consistency in the responses to indicate a number of emerging themes.
Evidence of Quality
Lincoln and Guba (1985) put forward four standards to establish the
trustworthiness of qualitative research and suggested they be used as an alternative to
more traditional quantitatively oriented criteria. I preserved the trustworthiness of the
research by implementing the four standards put forth by Lincoln and Guba:
credibility, transferability, dependability, and conformability.
Credibility is the degree to which findings and interpretations are plausible
and there is confidence in the “truth” of the findings (Lincoln & Guba, 1985). The
credible qualitative researcher provides comprehensive descriptions of the setting,
subjects, procedures, and interviews, so that the limitations and restrictions of the
study are clearly laid out. Lincoln and Guba recommended a number of techniques to
ensure credibility. I used three of those techniques in this study: triangulation, peer
debriefing, and member checking.
Triangulation was the first technique employed to maintain credibility
throughout the study. This cross-checking technique can use multiple sources,
methods, investigators, or theories (Creswell & Plano-Clark,2011). I used two
methods of collecting the data (quantitative surveys and qualitative interviews) to
verify participants’ meanings and thereby ensure credibility. I triangulated the
quantitative and qualitative data when I interpreted the results and wrote the
conclusions of the study.
Peer debriefing is the process of allowing an impartial peer to analyze the
inquiry and ask questions of the researcher (Lincoln & Guba, 1985), and when paired
with other strategies such as prolonged engagement, member checks, and
triangulation can support the findings as reliable and credible (Spall, 1998). This
approach may reveal certain aspects of the inquiry that may spur a new perspective. I
selected a respected colleague with a PhD and specialty in research design to support
this phase of the study. Peer debriefing permitted me to vocalize thoughts and
hypotheses, and the dialogue with the debriefer encouraged me to delve deeper into
the preliminary analyses of the data (Spall, 1998). I met with the peer debriefer
weekly during data analysis and the peer debriefer reviewed the report of the results.
Finally, the peer debriefing provided a cathartic environment whereby I was free to
clear my mind of all sentiments that may have biased the interpretation of the results.
Member checking was the final technique used to ensure credibility. Informal
member checking allowed participants to clarify intentions, correct errors, volunteer
additional information, and provide an assessment of what has happened. Informal
member checking also allowed me to consider the suitability of the responses and
summarize them (Lincoln & Guba, 1985). Formal member’ checking involved
sending a summary of the findings to participants for subsequent modification and
clarification, prior to writing the discussion section of the dissertation. Formal
member checking has its concerns. I ensured I did not reconstruct an overly
generalized account of what occurred and was aware of any myths or attempts to
cover up what had occurred (Lincoln & Guba, 1985). Another area of concern was in
the transcription of the interviews. The way the interviews are transcribed can impact
member checking and the method of transcription (verbatim, condensed, or cleaned-
up versions) must be chosen carefully (Carlson, 2010).
I employed the aforementioned techniques—triangulation, peer debriefing,
and member checking—during the qualitative phase of the study to ensure credibility
and quality of the results. Three criteria furthered trustworthiness and quality once the
study was completed: transferability, dependability, and conformability are de facto
results of research that is well conducted.

In qualitative research, the ability to take the results from one study and apply
them to another study is referenced as transferability (Lincoln & Guba, 1985). The
researcher does not define the parameters for transferability; instead, the researcher
provides rich and detailed information. The research report is a rich description that
“closely approximates the reality it represents” (Strauss & Corbin, 1990, p. 55). The
reader then makes the determination as to whether the results can be applied to a new
situation or framework.
To establish dependability and conformability of the results, I conducted an
audit trail. The audit trail “provides evidence in the form of data and documents such
as excerpts from field notes, transcripts, and research diaries, etc. which the
researcher’s peers, the auditors of the research, can follow” (Holloway, 1997, p. 92).
To maintain an audit trail I maintained records of each step in the data-collection and
data-analysis process. I will continue to save all Word files, interview transcripts,
personal notes, and drafts for at least 5 years, so that other researchers will be able to
follow and replicate the process (Carlson, 2010).
Protection of Participants’ Rights
I sought permission to conduct this study from Walden University’s
Institutional Review Board and the Superintendent of Schools in the district where the
study was conducted. I followed procedures to ensure participants were protected. I
defined the purpose for conducting the study and balanced that purpose with the
privacy of the participants. Improprieties resulting from these cross purposes can be
avoided by ensuring that proper ethical procedures are in place and participant rights
are protected at all costs. The choice to take part in any research study must be an
informed decision and must be completely deliberate on the part of the participant. I
took all necessary precautions to ensure the rights of human participants. The initial
email to potential participants established how I intended to protect their rights and
who to contact if they had any questions.
Survey Informed Consent and Confidentiality
At the beginning of the online survey I explained the title and purpose of the
study, procedures for completing the surveys, potential risks and benefits, the
confidentiality policy, and the withdrawal policy. I included a statement at the
beginning of the online survey that explained informed consent and established that
completion of the study survey constituted informed consent. Participants were
required to check an “I understand” box on the first page of the survey indicating they
agreed to participate in the study. The first page of the survey also contained the
phone number of who participants could contact if they had any questions or
concerns. Once the doctoral study was completed, the superintendent received a
summary of the study results.
To protect participants’ identity, I assigned them code numbers. The first two
digits of the code identified the school; the last three digits of the code established a
participant number. The password-protected website, Constant Contact, hosting the
online surveys, was no longer accessible to participants after the last day of the survey
period.
The electronic results of the online surveys and all raw data were transferred
from Constant Contact to a password-protected computer in my home office. After 5
years, the electronic results will be erased.
Interview Informed Consent and Confidentiality

At the beginning of the interview, I apprised each participant of their rights to
informed consent, what informed consent means, why it is necessary, and their right
to choose not to participate in the study at any point. At the time of the interview, I
also told participants who to contact should they have any questions. A copy of the
informed-consent form appears in Appendix F.
To preserve confidentiality, I referenced interview participants by an
alphabetical code. For example, Mary Jones would have been coded as MJ. I
transcribed the digital recordings of the interviews and saved them as text files. I
removed all identifying information from the transcripts. I erased the digital
recordings and saved only the text files. I transferred the text files to a flash drive and
stored in my home in a locked file drawer where they will remain for 5 years. After 5
years the flash drive will be reformatted, erasing all the data it contains.
Researcher’s Role
In 31 years as an educator in a single New Jersey public school district, I
served as a special educator, a regular educator, a master teacher, a staff developer,
and a coach. I had the occasion to work with an exemplary leader who had vision,
inspired greatness, and had long-range goals. This experience left a lasting
impression. I was then transferred to a new teaching environment. I observed quite a
different sense of teaching self-efficacy in the new leadership situation. I was also an
instructional coach for 2 years in the large urban district I served for 31 years,
affording me the opportunity to interact with novice educators. In this capacity I
sought to ensure that educators stayed in the profession.
I did not choose to complete this study in a school district in which I have
never been employed, nor do I know anyone in the district. I also chose this district
because I have no knowledge of the leaders or their leadership style. For this study, I
was the sole person responsible for collecting quantitative and qualitative data. I
ranked the scores and decided which participants would be eligible for follow-up
interviews. I also conducted the interviews, took notes during the interviews, and
transcribed the interview data. I coded the qualitative data to derive themes.
In conducting the interviews with teachers, I believe that my experience
enabled me to establish a trusting researcher–participant relationship. The participants
did not know me nor I them. Again, it was critical to bracket my passion for the
subject and my personal experiences during the interviews. The interviews began with
some background about me; then I asked the interviewee to share some background
about them. This sharing was meant to put the interviewee at ease and to establish
rapport.
Summary
This section has provided the methodology I used to conduct the study. I
described the sequential explanatory mixed-method design and laid out the setting,
sample, and participant selection. I used two survey instruments, the SSB and the
TSES, in the quantitative phase to correlate leadership-behavior variables with
teacher-efficacy variables. I gathered the data from the two surveys through the
Internet from a secure site to which only I had access. I used the Pearson product-
moment correlation to measure correlations between leadership behaviors and
teacher-efficacy scores for the two surveys.
The qualitative phase began after all surveys had been scored. Based on the
scores of the TSES, I used a purposeful-sampling strategy to select participants for
interviews. I recorded, transcribed, coded, and analyzed the interviews for themes
pertaining to the relationships of interest in the study. I used credibility,
transferability, dependability, and conformability techniques to maintain the
trustworthiness of the study results and procedures for informed consent, and
maintained confidentiality. I combined the quantitative and qualitative results during
the interpretation of results.

Chapter 4: Results
Introduction
The purpose of this sequential mixed-method study was to examine the
leadership behaviors associated with the development of personal teacher efficacy.
The predictor variables were leadership behaviors displayed by principals as
perceived by the teachers: trust, human relations, conflict, control, and instructional
leadership. The criterion variables were the effect the behaviors have on personal
teacher efficacy (engaging students, strategizing instructional practices, and managing
classrooms; Tschannen-Moran &Woolfolk-Hoy, 2001). I employed the sequential
mixed-method approach using a survey and individual interviews to address the
research questions of the study. In this explanatory mixed-methods design, I began by
conducting a quantitative phase and followed with a second phase (Creswell & Plano-
Clark, 2011, p. 1015). The quantitative and qualitative results were integrated during
the analysis phase of the study.
The rationale for using quantitative and qualitative data was to balance the
limitations inherent in one method with the strengths of the other method. This design
allowed for the integration of quantitative and qualitative analysis during the
interpretive phase. Here, I will explain the convergence of findings and explain lack
of convergence.
The most important advantage of the concurrent design is that I could achieve
a broader viewpoint by using different methods to obtain data. The disadvantage to
using this approach was that I found no recommendations about how I should resolve
inconsistencies that came about as a result of using two types of data, which may have
influenced the interpretation of the results (Creswell, 2003).
The study incurred two unexpected limitations: the sample size and the
number of volunteers who completed the interviews. Originally, I planned for 30
participants to respond to the survey, however only 19 responded. Ialso planned to
interview 10 participants, but only four were willing to be interviewed. Therefore, the
data analysis was limited by the data that were actually obtained.
I made every effort to obtain the needed responses for both the qualitative and
quantitative data. Incentives were given; email contact was repeatedly made.
Principals’ assistance was enlisted to contact and secure the number of respondents.
As a result of the low response rate, I made an adjustment to the quantitative analysis,
and used descriptive statistics were used. The low response rate and adjustments in
the qualitative analysis led to the study being more of a descriptive case study than a
mixed-method study.
Quantitative Data Analysis
Data Management and Scoring
I downloaded the data into Excel from Constant Contact, an online survey
service. The data consisted of the item-by-item responses byeach teacher for each of
the instruments in the survey. I then converted the Excel file to SPSS for data
screening, scoring, and subsequent statistical analysis.
As described in Section 3, the instruments consisted of a short demographics
form, the TSES, and the SSB. The TSES, a Likert-type scale (Tschannen-Moran et
al., 1998) distributed into three subscales (variables in the study): engaging students,
strategizing instructional practices, and managing classrooms. The SSB (Bulach et al.,
1999) consists of 49 behaviors that measure teachers’ perceptions of their principal’s
leadership style. Although there were data for 24 teachers, five of them did not
respond to any of the TSES items and were removed from the study leaving an N of
19. The TSES and the SSB were then scored according to the scoring instructions for
each instrument.
The scoring resulted in five subscale (domain) scores and three efficacy scores
for each of the 19 teachers. I then screened the scores for outliers, extreme high or low
individual scores that may have unduly influenced the statistics based on the total
group. To screen for outliers the raw scores for each teacher on each of the
instruments were transformed to standardized z-scores with a mean of zero and
standard deviation of 1 (Gravetter & Wallnau, 2013, p. 523). The criterion for
identifying an outlier was set at a z-score of +/-3.29, which indicates the score was
more than three standard deviations from the group mean and is considered extreme.
No outliers were identified that approached this criterion.
Survey Participants
Table 2 shows the teacher characteristics on three demographics. Because the
N of 19 is small, the numbers (n) in the table are perhaps more meaningful than the
percent column. The majority of teachers held bachelor’s degrees. Of the 19
responders, only three had less than 1 year of experience. More than half of the
participants had 6–10 years of teaching experience and more than half were teaching
in a participating school for 2–5 years. I did not determine how long they had been
assigned to their current school.

Table 2
Education and Teaching Demographics of the Teachers (N = 19)
Variables N %
Highest degree
Bachelor’s 12 63.2
Master’s 7 36.8
Years teaching in participating school
Less than 1 year 3 15.8
1 year but less than 2 years 6 31.6
2–5 years 10 52.6
Total number of years teaching
This is my first year 2 10.5
2–5 years 5 26.3
6–10 years 12 63.2
Quantitative Survey Analysis
I provide the descriptive statistics on each of the eight scales in Table 3. Scale
reliabilities are shown in the last column (α). Most previous researchers using the SSB
and TSES reported that both instruments showed adequate reliability. However,
although an instrument may be reliable based on the sample used in another study,
there is no guarantee that it will be reliable when a different study with a different
sample is conducted. Thus, an initial step in this analysis was to obtain the reliability
(Cronbach’s alpha) for each of the eight scales. Conventionally, a reliability
coefficient of approximately .70 or greater is considered to be adequate when

conducting group statistical analysis on an instrument (Gravetter & Wallnau, 2013, p.
221). Observation of the reliabilities, shown in Table 3, indicated that all but one were
above .70. The reliability for instructional strategies (α = .63) was less but was near
enough to .70 to be retained in the study.
Table 3
Means and Standard Deviations for Five Domains of Principal Behaviors, Three
Measures of Self-Efficacy and Reliability Coefficients (Cronbach’s α)
Variables M SD α
Behaviors
Human relations 4.07 .56 .86
Trust/decision making 4.21 .40 .77
Instructional leadership 4.29 .61 .89
Control 3.96 .62 .71
Conflict 4.21 .62 .79
Self-efficacy
Engagement 7.39 1.08 .86
Instructional strategies 7.70 .77 .63
Management 7.51 .97 .88
The mean of the five principal behaviors and the three self-efficacy scales indicated
that, overall, teachers perceived their principals favorably and considered themselves
confident with respect to self-efficacy about student engagement, instructional
strategies, and student management. The SSB (see Appendix A) shows the ratings of
principals, ranging from 1 (never), 2 (seldom), 3 (sometimes), 4 (often), to 5 (always)
in exhibiting a particular behavior. The higher the rating, the more favorably the
principal is viewed. The means shown in Table 3 indicated that, as a group, teachers
rated principals near and above 4 on each of the five domains, indicating they often or
always perceived the principals to be exhibiting desirable behaviors.
The teachers rated the TSES (see Appendix B) items with respect to their self-
efficacy when addressing student engagement, instructional strategies, and
management. The prompts asked teachers, “How much can you do?” followed by 24
items representing the three areas. The teachers responded to each item on a 9-point
scale ranging from 1 (nothing) to 9 (a great deal). The higher the rating, the greater
the self-efficacy; ratings of 6 or greater show high self-efficacy. The three self-
efficacy means shown in Table 3 are more than 7, indicating self-efficacy was high
for the group of teachers.
The quantitative portion of the study included research questions about the
teachers’ perceptions of their sense of efficacy in respect to student engagement,
instructional strategies, and classroom management as associated with the principals’
domains of human relations, trust/decision making, instructional leadership, control,
and conflict. Thus, there were 9 research questions. The questions (RQs) and
associated null (H0) and alternative (Ha) hypotheses are repeated from earlier sections
as follows:
student engagement?



I used bivariate correlation to test the null hypotheses. The .05 level of
probability was the criterion used for rejection. Statistical significance is the
probability of observing results as extreme as those observed if the null hypothesis is
true. However, statistical significance provides no evidence about the magnitude of
the size of a correlation regardless of the probability level achieved. For this reason it
is recommended that effect sizes be reported and interpreted in addition to statistical
significance results (American Psychological Association, 2009, p. 34). Thus, the
statistical results are reported first followed by the effect size results.
For the statistical tests I correlated each of the three measures of self-efficacy
with each of the measures of principal behaviors, resulting in 15 correlations in all.
Table 4 provides the results of the hypothesis tests. The actual probabilities (p) are
shown associated with each correlation (r).For a correlation to be statistically
significant, the p value had to be .05 or less. It may be seen in the table that there were
two statistically significant correlations, shown in bold type. The correlations between
the self-efficacy measure of engagement and instructional leadership (r= .53, p= .02)
and conflict (r= .51, p= .03) were both statistically significant. Thus, hypotheses
H07and H013were rejected for engagement, showing support for the positive
relationship between engagement and instructional leadership and conflict. I found no
other statistically significant correlations between the measures of self-efficacy and
principal behaviors.
Table 4
Interco relations for Five Domains of Principal Behaviors and Three Measures of
Self-Efficacy
Self-efficacy measures
Engagement Strategies Management
Domain r p r p R p
Human relations .19 .44 .28 .25 .07 .79
Trust/decision making .43 .07 .10 .67 .12 .64
Instructional leadership .53 .02 .16 .51 .23 .34
Control .33 .17 .30 .22 .23 .34
Conflict .51 .03 .37 .12 .29 .22

The correlation coefficients shown in Table 4 may also be used as indicators
of effect sizes and especially useful for these data because of the small sample size.
An effect size is independent of sample size and provides information about the
strength or magnitude of a correlation regardless of whether the correlation is
statistically significant (Newton & Rudestam, 2013). An effect size is considered to
be a measure of the magnitude of the correlation between two variables and covers the
entire range of a relationship from no relationship at all (r= 0) to a perfect relationship
(r= 1, or r= -1) (Gravetter & Wallnau, 2013).
Commonly used criteria (Newton & Rudestam, 2013) to interpret a correlation
coefficient as an effect size is as follows:
Weak effect size: .10
Moderate effect size .30
Strong effect size .50
Observation of the correlations in Table 4 between the self-efficacy measure
of student engagement as it related to the principals ‘measures on Instructional
Leadership (r= .53) and Conflict (r= .51) can be considered strong effect sizes using
the criterion of .50. The Trust/Decision Making relationship with Engagement (r =
.43) showed a moderate-to-strong effect, whereas the Control relationship with
Engagement (r = .33) shows a moderate effect size using the criterion of.30.
In summary, the quantitative analyses resulted in two statistically significant
correlations (p< .05) that can be considered to show positive relationships between the
self-efficacy and principal-behavior measures. Whereas statistical significance is the
probability of observing correlations as extreme as those observed if the null

hypothesis is true, it provides no information about the importance of the correlations.
By including effect sizes, the magnitude or importance of relationships is emphasized.
Qualitative Data Analysis
Interview Transcription and Coding
I taped the interviews using a digital recorder and transcribed them word for
word, typed into a Word document. I then created an Excel Workbook with a
spreadsheet for each of the four participants’ interview responses. I copied and pasted
the responses from the Word document into the Excel spreadsheet in the appropriate
cell for each of the 15 relationships studied. I then highlighted the commonalities and
discrepancies found in the responses and began to group the responses intocategories
or themes.
Interview Participants
Of the 19 teachers who completed the survey, four agreed to participate in the
follow-up interview. As shown in Table 5, three of the four were new teachers with 5
or fewer years of experience. Although overall TSES scores were high in this sample,
interview participants ranked near the top, the bottom, and the middle range of scores.
Table 5
Teacher Sense of Efficacy Scale Score, Rank, and Years of Teaching Experience of
the Teachers (N = 4)
Pseudonym TSES Score Rank 1–19 Years of teaching experience
Hillary 199 2 3-5
Lori 143 18 2
Miriam 173 10 3-5
Nola 178 07 More than 10
Note. TSES = Teacher Sense of Efficacy Scale.
Hillary (high TSES score, pseudonym) ranked second on the TSES Scale
(199) out of the 19 people who responded to the survey. Hillary has been teaching
between 3 and 5 years in a small neighborhood school with a team teacher in the room
at all times. Hillary said the principal is very supportive. “She’s behind me 100%. …
It’s always WE, it’s not going to be just I.”
Lori (low TSES score) ranked 18th of 19 on the TSES Scale (143). She has the
least amount of teaching experience of the 19 in the sample and works in a larger
middle school environment. She transitioned from corporate America and feels she is
in a coaching relationship with her principal. She compares her principal to a mentor
who “is big on communicating with us.”
Miriam (middle TSES score) ranked in the mid-range of the TSES Scale
(173). Miriam also has 3–5 years of teaching experience. Miriam also feels the she
has 100% the principal’s support. Miriam felt she has a good relationship and when it
comes to trust she said, “I do trust my principal.”
Nola also fell in the mid-range with a TSES score of 178. Nola has more than
years of teaching experience. Her responses indicated, “If you are a seasoned teacher,
you do what you need to do for your student’s regardless of your principal.” Nola
replied that principal behaviors do not impact her engagement or strategies of

management efficacy. Her data are incongruent with the quantitative results and the
results from the three novice teachers. Because her responses consistently indicated
she saw no relationship between principal behavior and her self-efficacy, I did not
include her data in the thematic descriptions below.

Theses B.ed

Uploaded by

Copyright:

Available Formats

Theses B.ed

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Theses B.ed

Uploaded by

Copyright:

Available Formats

Faculty of Education

Allama Iqbal Open University, Islamabad

Teachers’ Leadership practices and Teachers self-efficacy”. Proposed and submitted

degree of B.Ed. (1.5 year) Teacher Education is hereby accepted.

Supervisor: _______________________ (Signature)

Evaluator: ________________________ (Signature)

Research Project entitled “ Study of relationship between Head Teachers’ Leadership

practices and Teachers self-efficacy”. Submitted by Usama Hanif Roll No.

University for evaluation.

Dr. Riaz Hussain

The attrition rate of teachers in an urban/rural school district in Hafizabad caused

schools to fail to attain annual yearly progress. To reverse this problem,

sequential mixed-methods study was to determine whether a relationship existed

between leadership and efficacy. Total-population sampling was used to obtain 21

relationship between Head Teachers’ behaviors (human relations, trust/decision

engagement, instructional strategies, and classroom management). Survey data were

analyzed using Pearson’s product-moment correlations. In addition, face-to-face

indicated significant relationships between instructional leadership with teacher

students in their educational challenges to increase academic performance.

1. Introduction to the Study 1

1.1 Statement of the Problem 2

1.2 Nature of the Study 3

1.3 Research Questions 4

1.4 Research Hypotheses 7

1.5 Purpose of the Study 12

1.6 Theoretical Framework 13

1.7 Teacher Efficacy 13

1.8 Leadership Behaviors 13

1.9 Definition of Terms 15

1.10 Assumptions, Limitations, Scope, and Delimitations 17

1.11 Significance of the Study 18

1.12 Implications for Social Change 19

2. Review of the Literature 22

2.3 Leadership Scales 25

2.4 Social-Learning Theory and the Concept of Efficacy 28

2.5 Teacher Efficacy 28

2.6 Teacher-Efficacy Scales 29

2.7 Related Studies on Teacher Efficacy 31

2.8 Relationship Between Leadership and Self-Efficacy 33

2.10 Literature Related to the Method 38

2.11 Literature Related to Differing Methodologies 39

3.2 Mixed-Method Design 42

3.3 Setting and Sample 42

3.4 Population and Sample for the Survey 43

3.5 Participant Selection for Interviews 43

3.6 Connection to the Research Questions 44

3.7 Quantitative Survey 47

3.9 Qualitative Interviews 51

3.10 Data Analysis 52

3.11 Survey Data Analysis 52

3.12 Interview Data Analysis 53

3.13 Evidence of Quality 53

3.14 Protection of Participants’ Rights 56

3.15 Survey Informed Consent and Confidentiality 57

3.16 Interview Informed Consent and Confidentiality 57

3.17 Researcher’s Role 58

4.2 Quantitative Data Analysis 62

4.3 Data Management and Scoring 62

4.4 Survey Participants 63