ORGANIZATIONAL LEADERSHIP, TEACHER WORK CONDITIONS AND TEACHERS’
JOB SATISFACTION: A MULTI-COUNTRY ANALYSIS
By
Madhur Chandra

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
Educational Policy – Doctor of Philosophy
2014

ABSTRACT
ORGANIZATIONAL LEADERSHIP, TEACHER WORK CONDITIONS AND TEACHERS’
JOB SATISFACTION: A MULTI-COUNTRY ANALYSIS
By
Madhur Chandra
The notion that principals’ actions create distinct working environments in schools and
that these environments are associated with teacher satisfaction is well established in the United
States (Johnson, Kraft & Papay, 2011; Moore, 2012; Grissom, 2011). As part of the process of
creating a supportive work environment principals may also take steps to establish a culture of
collaboration among teachers. Notably however, multi-country analysis examining principal
leadership and teacher satisfaction is limited and a specific focus on Asia is missing from the
existing literature on the subject. Similarly, the relationships between teacher collaboration,
principal leadership and teacher satisfaction across varying social, cultural, and economic
contexts are understudied. This dissertation consisting of three interrelated papers is a systematic
effort to address this gap in the literature. Chapter 1 in this dissertation is an extensive analytic
review of teacher work conditions associated with teacher satisfaction within the United States
and internationally. This chapter is foundational because it identifies gaps in the literature on
teacher satisfaction. Based on this analytic review I focus on two relational and policy amenable
organizational factors – principal leadership and teacher collaboration in the two chapters that
follow.
Chapter 2 explores the association between principal leadership style and teachers’ job
satisfaction in six Asian countries – Hong Kong, Indonesia, Malaysia, Mongolia, Taiwan, and
Thailand, as well as the United States. I use a large-scale international dataset - the Trends in
International Math and Science (TIMSS) 2007, which are the only large-scale secondary data

publicly available that represent several countries in Asia and specifically pertain to education.
Additionally, TIMSS 2007 contain a comprehensive set of demographic and contextual variables
that enable a rich analysis of the relationships of interest. Results indicate that two aspects of
transformational leadership are significantly associated with principal reported teacher
satisfaction across all the countries analyzed. These are identified as the principal’s role in
building a shared vision and mission for the school, which Leithwood and Sun (2012) name
“setting directions” and the principal’s role in capacity building within schools which they name
“developing people.”
Chapter 3 focuses on teacher collaboration in five OECD countries – Denmark, Poland,
Brazil, Mexico and Hungary using the Teaching and Learning International Survey (TALIS
2008). A fractional response model is used to explore the association between factors measuring
transformational leadership in these countries and a summative index of teacher collaboration. I
also deconstruct the relationship between the leadership factors and twelve individual measures
of teacher collaboration. Results indicate a predominantly positive association between factors
measuring transformational leadership and individual measures of collaboration pertaining to
classroom teaching and learning. Additionally, I find that in all five countries there is a strong
positive association between the index of teacher collaboration and teachers’ job satisfaction.
Thus, cumulatively the two large-scale data sets used in this dissertation – (TIMSS 2007
and TALIS 2008) generate a more comprehensive understanding of how particular aspects of
teacher work conditions - principal leadership and teacher collaboration are associated with
variations in teacher satisfaction in a multi-country context.

Copyright by
MADHUR CHANDRA
2014

I dedicate this dissertation to my girls Anjini, Kavya, and Divya who have remained my loving
and enthusiastic supporters over the last five years and have kept me grounded.

v

ACKNOWLEDGEMENTS
This dissertation benefited from a great deal of advice, support and encouragement from
several people within the United States, as well as in India. First and foremost, I am deeply
grateful to my advisor Dr. Amita Chudgar for her guidance and support during the writing of this
dissertation. Working as Amita’s graduate research assistant I have had the opportunity to
observe and learn much from her these last three years. She has constructively challenged me to
expand my horizons and has often encouraged me to undertake work outside my comfort zone.
The UNICEF teacher project, as well as the Child Friendly School Initiative, which we jointly
worked on have greatly enriched my perspective on issues of marginalization and inequity in
schools.
My special thanks also go out to the members of my committee, Dr. Peter Youngs, Dr.
Susan Printy, and Dr. Spyros Konstantopoulos for their willingness to read earlier drafts of this
document and for their constructive comments during various iterations of this dissertation.
They have profoundly influenced my intellectual growth and development via the strength of
their own understanding of Educational Policy, Educational Administration and Statistics. I am
also deeply grateful to Dr. Peter Youngs and Dr. Spyros Konstantopoulos for including me as a
teaching assistant in their classes (TE 201 and CEP 933 respectively). The opportunity to work
closely with each of them and to observe different approaches to teaching has been invaluable.
Additionally, I would like to thank Dr. Robert Floden and Dr. Jeffrey Wooldridge for the
opportunity to participate in the IES Economics of Education specialization1 which has shaped

1

This work was supported in part by a Pre-Doctoral Training Grant from the Institute of Education Sciences, U.S.
Department of Education (Award # R305B090011) to Michigan State University. The opinions expressed here
are those of the author and do not represent the views of the Institute, the U.S. Department of Education, or
Michigan State University.

vi

my understanding of rigor in empirical educational research. It has been an honor to receive the
very generous and prestigious IES fellowship during the course of my doctoral study.
Words cannot fully express my gratitude to Professor Michael Sedlak whose belief in me
made this wonderful journey a reality. Thank you for your willingness to support me at different
stages of my professional development. In particular, I am deeply grateful for the Summer
Research Development Fellowship (SRDF’10) early on in my doctoral coursework, which gave
me the opportunity to write and publish my first sole-authored literature review. Additionally, I
wish to thank Dr. Tom Luschei for helpful suggestions regarding Mexico.
I would also like to acknowledge my peers and friends at Michigan State and those from
other institutions - Kaitlin Obenauf, Tracy Davis, Justina Judy, Andy Saultz, Dorothy HinesDatiri, Wei Li, Jeff Snyder, Ayesha Razzaque, Honey Ghods, Jim Pippin, Ben Creed, Alisha
Brown, Alyssa Morley, Mary Mason, Juanita Bautista-Guerra, Yisu Zhou, Radhika Iyengar,
Namrata Tognatta, Katyn Chmielewski and Min Sun for their camaraderie and willingness to
engage in policy conversations that have furthered my understanding of educational issues over
the past five years. Min and Mary, sharing an office with you during my first year was
inspirational to say the least.
I am eternally grateful to my mother, Mrs. Purnima Paul for her unconditional love and
support through the whole process. I thank you for being there for me no matter what.
Last, but by no means the least, I would like to thank my husband Siddharth, whose
cheerful optimism, steadfast faith in me, and enthusiastic words of encouragement have made
this dissertation possible. I am deeply appreciative of the fact that he gave me the space I needed
to engage fully in this intellectual endeavor.

vii

TABLE OF CONTENTS

LIST OF TABLES……………………………………………………………………………… xiii
LIST OF FIGURES…………………………………………………………………………….. xv
CHAPTER 1
TEACHERS’ WORK CONDITIONS AND THEIR JOB SATISFACTION: A REVIEW
OF THE LITERATURE ...............................................................................................................1
1.1 Introduction…………………………………………………………………………... 1
1.2 Theoretical Framework……………………………………………………………….. 3
1.3 Methodology…………………………………………………………………………. .5
1.4 A Review of Literature – U.S........................................................................................ 7
1.4.1 Demographic Factors……………………………………………………….. 8
1.4.2 School/Classroom Factors………………………………………………… 12
1.4.3 Organizational Factors…………………………………………………….. 14
1.5 A Review of Literature – International……………………………………………… 21
1.5.1 Demographic Factors……………………………………………………… 21
1.5.2 School/Classroom Factors………………………………………………… 23
1.5.3 Organizational Factors…………………………………………………….. 24
1.6 Policy Implications & Limitations of the Literature………………………………… 28
1.6.1 Lack of Multi-country Analysis…………………………………………… 28
1.6.2 Limited Focus on Asia…………………………………………………….. 28
1.6.3 Limited Attention to Important Correlates………………………………... 29
1.7 Conclusion…………………………………………………………………………... 29
APPENDIX………………………………………………………………………………………31
BIBLIOGRAPHY………………………………………………………………………………..43
CHAPTER 2
DOES LEADERSHIP STYLE MATTER? A MULTI-COUNTRY ANALYSIS OF
PRINCIPAL LEADERSHIP AND TEACHERS’ JOB SATISFACTION………………... 49
2.1 Introduction………………………………………………………………………….. 49
2.2 Literature Review…………………………………………………………………….50
2.2.1 Leadership & Satisfaction………………………………………………….50
2.2.2 Instructional Leadership……………………………………………………51
2.2.3Transformational Leadership……………………………………………… 52
2.2.4 International Literature on Leadership……………………………………..53
2.3 Conceptual & Theoretical Framework……………………………………………… 55
2.4 Country Context…………………………………………………………………....... 58
2.4.1 Hong Kong………………………………………………………………… 59
2.4.11 Location……………………………………………………………. 59
2.4.12 Historical Background……………………………………………... 59
2.4.13 Education System………………………………………………….. 59
2.4.2 Indonesia…………………………………………………………………... 60
viii

2.4.21 Location……………………………………………………………. 60
2.4.22 Historical Background……………………………………………... 60
2.4.23 Education System…………………………………………………...61
2.4.3 Malaysia…………………………………………………………………… 62
2.4.31 Location……………………………………………………………. 62
2.4.32 Historical Background……………………………………………... 62
2.4.33 Education System…………………………………………………...62
2.4.4 Mongolia…………………………………………………………………... 63
2.4.41 Location……………………………………………………………. 63
2.4.42 Historical Background……………………………………………... 63
2.4.43 Education System………………………………………………….. 64
2.4.5 Taiwan……………………………………………………………………...64
2.4.51 Location……………………………………………………………. 64
2.4.52 Historical Background……………………………………………... 65
2.4.53 Education System………………………………………………….. 65
2.4.6 Thailand…………………………………………………………………… 66
2.4.61 Location……………………………………………………………. 66
2.4.62 Historical Background……………………………………………... 66
2.4.63 Education System………………………………………………….. 66
2.4.7 U.S.A…………………………………………………………………….... 67
2.4.71 Location……………………………………………………………. 67
2.4.72 Historical Background……………………………………………... 67
2.4.73 Education System…………………………………………………...68
2.5 Research Questions ………………………………………………………………… 69
2.6 Data…………………………………………………………………………………..70
2.6.1 Sample Design & Weights………………………………………………… 70
2.6.2 Description of Dependent Variables………………………………………. 71
2.6.3 Description of Key Independent Variables………………………………... 72
2.6.4 Description of Controls……………………………………………………. 77
2.7 Method……………………………………………………………………………….78
2.7.1 Analysis I: Teacher Reported Teacher Satisfaction I (Teacher
Level)……………………………………………………………………… 80
2.7.2 Analysis II: Teacher Reported Teacher Satisfaction II (School
Level)……………………………………………………………………… 82
2.7.3 Analysis III: Principal Reported Teacher Satisfaction (School
Level)……………………………………………………………………… 82
2.8 Results………………………………………………………………………………..84
2.8.1 Analysis I: Research Question 1 (Teacher Level – Unweighted)………… 84
2.8.11 Hong Kong (SAR) Results………………………………………… 84
2.8.12 Indonesia Results…………………………………………………... 85
2.8.13 Malaysia Results…………………………………………………… 85
2.8.14 Taiwan Results…………………………………………………….. 85
2.8.15 Thailand Results…………………………………………………… 85
2.8.16 U.S.A. Results……………………………………………………… 86
2.8.2 Analysis II: Research Question 1 (School Level – Weighted)……………. 86
2.8.21 Hong Kong (SAR) Results………………………………………….86
ix

2.8.22 Indonesia Results………………………………………………….. 86
2.8.23 Malaysia Results…………………………………………………… 86
2.8.24 Mongolia Results…………………………………………………... 87
2.8.25 Taiwan Results……………………………………………………... 87
2.8.26 Thailand Results…………………………………………………… 88
2.8.27 U.S.A. Results……………………………………………………… 88
2.8.3 Analysis III: Research Question 2 (School Level – Weighted)………….... 88
2.8.31 Hong Kong (SAR) Results………………………………………… 88
2.8.32 Indonesia Results…………………………………………………... 89
2.8.33 Malaysia Results…………………………………………………... 89
2.8.34 Mongolia Results…………………………………………………... 89
2.8.35 Taiwan Results…………………………………………………...... 90
2.8.36 Thailand Results…………………………………………………... 90
2.8.37 U.S.A. Results……………………………………………………... 90
2.9 Discussion…………………………………………………………………………....91
2.9.1 Research Question 1………………………………………………………. 91
2.9.2 Research Question 2………………………………………………………. 93
2.10 Contextual Discussion of Results in Hong Kong………………………………….. 96
2.10.1 Country Background…………………………………………………….. 96
2.10.2 Overview of Educational System…………………………………………98
2.10.3 Profile of Teachers in Hong Kong……………………………………….. 99
2.10.4 School Leadership in Hong Kong………………………………………. 100
2.11 Conclusion and Limitations……………………………………………………..... 102
APPENDIX………………………………………………………………………….................. 115
BIBLIOGRAPHY……………………………………………………………………………… 126
CHAPTER 3
TEACHER COLLABORATION IN INTERNATIONAL CONTEXTS: DOES
TRANSFORMATIONAL LEADERSHIP PLAY A ROLE?................................................ 133
3.1 Introduction………………………………………………………………………… 134
3.2 Literature Review…………………………………………………………………...135
3.2.1 Collaboration…………………………………………………………..….135
3.2.2 Transformational Leadership & Collaboration…………………………... 138
3.2.3 Teacher Satisfaction & Collaboration……………………......................... 141
3.3 Theoretical Framework…………………………………………………………….. 142
3.4 Country Context…………………………………………………………………..... 144
3.4.1 Denmark………………………………………………………………..… 144
3.4.11 Background……………………………………………………….. 145
3.4.12 Teacher Characteristics…………………………………………… 145
3.4.2 Poland……………………………………………………………………. 145
3.4.21 Background………………………………………………………. 145
3.4.22 Teacher Characteristics…………………………………………… 146
3.4.3 Brazil…………………………………………………………………..…. 146
3.4.31 Background……………………………………………………….. 146
3.4.32 Teacher Characteristics…………………………………………… 147

x

3.4.4 Mexico………………………………………………………………….... 147
3.4.41 Background………………………………………………………. 147
3.4.42 Teacher Characteristics…………………………………………… 147
3.4.5 Hungary………………………………………………………………….. 148
3.4.51 Background………………………………………………………. 148
3.4.52 Teacher Characteristics…………………………………………… 148
3.5 Research Questions………………………………………………………………… 149
3.6 Data……………………………………………………………………………….... 151
3.6.1 Sample Design & Weights...……………………………………………... 151
3.6.2 Description of Dependent Variables……………………………………... 152
3.6.3 Description of Key Independent Variables……………………………….153
3.6.4 Description of Controls…………………………………………………... 158
3.7 Method……………………………………………………………………………... 164
3.7.1 Method – Research Question 1…………………………………………... 164
3.7.2 Method – Research Question 2…………………………………………... 165
3.7.3 Method – Research Question 3…………………………………………... 166
3.8 Results and Discussion…………………………………………………………….. 167
3.8.1 Results – Research Question 1…………………………………………… 168
3.8.11 Denmark………………………………………………………….. 168
3.8.12 Poland…………………………………………………………….. 168
3.8.13 Brazil……………………………………………………………… 169
3.8.14 Mexico……………………………………………………………. 169
3.8.15 Hungary……………………………………………………………170
3.8.16 Summary of Results – RQ1………………………………………..170
3.8.2 Results – Research Question 2…………………………………………… 172
3.8.21 Denmark………………………………………………………….. 172
3.8.22 Poland…………………………………………………………….. 173
3.8.23 Brazil……………………………………………………………… 174
3.8.24 Mexico……………………………………………………………. 175
3.8.25 Hungary……………………………………………………………175
3.8.26 Summary of Results – RQ2………………………………………..176
3.8.3 Results – Research Question 3…………………………………………… 182
3.8.31 Denmark………………………………………………………….. 182
3.8.32 Poland…………………………………………………………….. 182
3.8.33 Brazil……………………………………………………………… 183
3.8.34 Mexico……………………………………………………………. 183
3.8.35 Hungary……………………………………………………………183
3.8.36 Summary of Results – RQ1………………………………………..185
3.9 Contextual Discussion of Results in Mexico………………………………………. 185
3.9.1 The National Context……………………………………………………. 185
3.9.2 Overview of the Educational System……………………………………. 186
3.9.3 Profile of Teachers in Mexico…………………………………………… 187
3.9.4 Leadership in Mexico……………………………………………………. 190
3.10 Conclusion & Limitations………………………………………………………… 192
BIBLIOGRAPHY……………………………………………………………………………… 195

xi

CHAPTER 4
CUMULATIVE SYNTHESIS……………………………………………………………….. 202
4.1 Summary of Findings………………………………………………………………. 202
4.2 Study Contributions………………………………………………………………... 205
4.3 Comparative Insights………………………………………………………………. 206
4.4 Policy Implications & Limitations ………………………………………………… 207
4.5 Directions for Future Research…………………………………………………….. 209

xii

LIST OF TABLES

Table 1:

No. of Studies on Teacher Satisfaction U.S.A / International……………………… 6

Table 2:

No. of Studies on Teacher Satisfaction by Region / Study Type……………………...7

Table A1: List of Studies Reviewed with Key Findings……………………………………….. 32
Table 3:

Components of Leadership Indices………………………………………………….. 73

Table 4:

Factor Loadings for Transformational Leadership (Math & Science) -TRTS
(School)……………………………………………………………………………… 76

Table 5:

Cronbachs Alpha for Transformational Leadership (Math & Science) -TRTS
(School)……………………………………………………………………………… 77

Table 6:

Descriptive Statistics for Leadership Indices, Student Background, Teacher
Demographic and School Controls (TIMSS 07), School Level…………………….. 79

Table 7:

Table of Results for Teacher Perception of Teacher Satisfaction & Principal
Leadership Style (Analysis I - Teacher Level)…………………………………….. 106

Table 8:

Table of Results for Teacher Perception of Teacher Satisfaction & Principal
Leadership Style (Analysis II - School Level)……………………………………... 109

Table 9:

Table of Results for Principal Perception of Teacher Satisfaction & Principal
Leadership Style (Analysis III - School Level)……………………………………. 112

Table A2: Pearson Correlations for Principal and Teacher Perceptions of Teacher
Satisfaction…………………………………………………………………………. 116
Table A3: Pearson Correlations for Leadership Indices – Analysis I (Teacher Level)………...117
Table A4: Pearson Correlations for Leadership Indices – Analysis II (School Level)………...121
Table A5: Descriptive Statistics TRTSI – Teacher Level……………………………………...125
Table 10: Summary of Country Context – TALIS 2008………………………………………150
Table 11: Types of Teacher Collaboration…………………………………………………….154
Table 12: List of Variables Used in EFA for Transformational Leadership Indices – TALIS
2008…………………………………………………………………………………155

xiii

Table 13: Factor Loadings for Transformational Leadership Indices – TALIS 2008………... 156
Table 14: Scale Reliability Coefficient (Cronbachs Alpha) for Factor Scores of
Transformational Leadership – TALIS 2008………………………………………. 157
Table 15: Complete List of Variables Used – TALIS 2008………………………………….. 160
Table 16: Descriptive Statistics for Analysis on Teacher Collaboration – TALIS 2008……... 162
Table 17: Table of Results for Collaboration Index and Transformational Leadership……….171
Table 18: Results for Types of Collaboration & Transformational Leadership in Denmark.... 177
Table 19: Results for Types of Collaboration & Transformational Leadership in Poland....... 178
Table 20: Results for Types of Collaboration & Transformational Leadership in Brazil……. 179
Table 21: Results for Types of Collaboration & Transformational Leadership in Mexico…... 180
Table 22: Results for Types of Collaboration & Transformational Leadership in Hungary…. 181
Table 23: Table of Results for Teacher Satisfaction & Index of Collaboration………………. 184

xiv

LIST OF FIGURES

Figure 1: A Conceptual Diagram of Teacher Work Conditions and Teacher Satisfaction……... 4
Figure 2: A Conceptual Model of Principal Leadership and Teachers’ Job Satisfaction………57
Figure 3: Conceptual Interrelationships between Transformational Leadership, Teacher
Collaboration and Teachers’ Job Satisfaction………………………………………143

xv

CHAPTER 1
TEACHERS’ WORK CONDITIONS AND THEIR JOB SATISFACTION: A REVIEW
OF THE LITERATURE
1.1 Introduction
There is a long and venerable tradition of research on job satisfaction and behavioral
outcomes in industrial-organizational psychology (Judge, Parker, Colbert, Heller, & Illes, 2001;
Judge, Bono, Thoreson & Patton, 2001). Indeed, interest in the links between workplace
attitudes and behavioral outcomes in this field goes back several decades. Job satisfaction has
been explicated in this literature as an affective reaction to an individual’s work situation. Locke
(1976) for example defined job satisfaction as “….a pleasurable or positive emotional state
resulting from the appraisal of one’s job or job experiences” (p.1304). According to him, job
satisfaction is important for two reasons. First, from a humanitarian perspective it is indicative
of the emotional well-being of employees. Second, from a utilitarian perspective job satisfaction
can lead to behaviors that impact organizational functioning. It may be argued therefore, that
research on job satisfaction is important from the perspective of organizational effectiveness.
As defined in this paper, job satisfaction refers simply to the extent of an individual’s
overall satisfaction with their job. Despite the extensive literature on the importance of job
satisfaction in the industrial/organization literature, it has received comparatively less research
attention within the field of education. Nonetheless, there is mounting evidence in the education
literature as well in support of the finding that job satisfaction is associated with outcomes such
as teacher turnover and student achievement. For example, it is well documented within the
United States that job satisfaction is strongly associated with teacher turnover (Boyd, Grossman,
Ing, Lankford, Loeb, Wyckoff, 2011; Ingersoll, 2001; Ostroff, 1992; Stockard & Lehman, 2004).
1

It is equally well-established that there is a high level of teacher turnover in U.S. public schools2
(Boyd et al, 2011; Guarino, Santibanez & Daley, 2006; Ingersoll, 2001). Teacher turnover is
both costly and disruptive for schools in the U.S.3 Given the high rate of turnover and the fact
that satisfaction is a known antecedent of turnover, extending our understanding of the correlates
of teacher satisfaction appears to be both necessary and policy relevant.
Additionally, teacher satisfaction has direct implications not only for teacher turnover,
but also for teacher commitment, and student achievement. Previous literature on job
satisfaction and organizational commitment indicates that the two concepts are highly correlated
and are in turn significantly associated with work conditions. If a person is satisfied with their
job, they are also much more likely to be more committed to their work (Firestone & Pennell,
1993; Mowday, Steers & Porter, 1979). Concomitantly, more satisfied and committed workers
are far less likely to leave their jobs than those that are dissatisfied (Perie & Baker, 1997).
Research shows that teacher job satisfaction also has a crucial influence on the student
body. Highly satisfied teachers are likely to have more positive relationships with their students
and are therefore, much more likely to help them attain better academic outcomes (Michaelowa
& Wittmann, 2007). In addition, teacher satisfaction influences the stability and quality of
instruction provided to students (Perie & Baker, 1997). Ostroff (1992) argues that teachers who
are dissatisfied may not put forward their best effort in the classroom though they may continue
to stay in teaching. Dissatisfied teachers are unlikely to actively engage with students.
Systematic research on the correlates of teacher satisfaction can therefore, help to inform policy
2

According to Boyd et al (2011) across the U.S. approximately half a million teachers leave their schools every
year. Only 16% of this turnover can be attributed to retirement. The rest (84%) is attributed to the movement of
teachers between schools and to teachers leaving the profession entirely.

3

The National Commission on Teaching and America’s Future (NCTAF, 2007) estimated that teacher turnover cost
the nation more than $7 billion in the 2003-04 school year alone.

2

addressing teacher retention and commitment as well as policy aimed at improving student
achievement.
1.2 Theoretical Framework
The goal of this review is to extend our understanding of the factors that matter for
teacher satisfaction, particularly in the international context. I draw on a vast body of research
on school climate and schools as organizations as a theoretical framework within which to
explore the influence of teacher work conditions on teacher satisfaction across countries.
Although there is not a single agreed upon definition of school climate, one way to think about
school climate is in reference to the quality and character of school life (Cohen, Mccabe,
Michelli, & Pickeral, 2009). Early systematic studies of school climate were influenced by
organizational research and research on school effectiveness. Over the last three decades
researchers have come to realize that a complex range of internal and external influences shape
collective experiences of schooling. Briefly, school climate encompasses norms, values, and
goals pertaining to the school, interpersonal relationships, teaching and learning practices and
organizational structures.
It has been suggested that there are four main dimensions that shape school climate.
These dimensions are identified by Cohen et al. (2009) as follows.
1. Safety – both physical and socio-emotional.
2. Teaching and learning – encompassing professional development, effective instruction
and leadership support.
3. Positive relationships – among teachers, leaders, and staff, opportunities for collaboration
and the establishment of a school community, morale and connectedness.

3

4. Environmental / Structural – Availability of adequate instructional materials, school size
and clean facilities.
Figure 1 is a conceptual diagram of teacher work conditions and teacher satisfaction based on the
school climate framework. In recognition of the complexity of internal and external influences
that impact teacher outcomes, I include a comprehensive list of contextual variables –
demographic, school, and organizational. I conceptualize these contexts as shaping teachers’
work environments and hypothesize that specific aspects of these work conditions like principal
leadership and teacher collaboration in particular influence teacher satisfaction based on the
understanding that the school climate framework is to a large extent relational in nature. How
connected or not teachers and students feel within schools is therefore, an important aspect of
school climate. Although, teacher satisfaction is the outcome of interest in this review,
satisfaction itself is also conceptualized as an important antecedent of teacher outcomes such as
turnover and teacher commitment as well as student achievement as shown in Figure 1.

4

Based on this theoretical framework, in the next two sections I review the methodology
used to identify relevant literature on teacher satisfaction within the U.S. as well as
internationally.
1.3 Methodology
In this section I describe the process used to identify relevant literature and the data that
emerged as a result. A broad search of the literature on teacher satisfaction was conducted
initially using a variety of search strategies. The search parameters were constrained to include
all studies in the period between 1990 and 2013. First, I used the search engine Google Scholar
with terms such as “teacher satisfaction,” “work conditions,” “teacher turnover,” “job
satisfaction” in conjunction with the Boolean operator “AND.” Second, I conducted a separate
search within major education journals like American Educational Research Journal, Asia
Pacific Educational Review, Comparative Education Review, Educational Administration
Quarterly, Elementary School Journal, International Journal of Educational Development,
Review of Educational Research, School Effectiveness and School Improvement, and Teaching
and Teacher Education among others. I used the search term “teacher satisfaction” as well as
terms directly pertaining to aspects of teacher work conditions such as “teacher satisfaction AND
principal leadership” and “teacher satisfaction AND collaboration.” Finally, I reviewed the
reference sections of all the retrieved articles in an attempt to identify additional relevant
material that may have been missed by the earlier searches4. These methods together yielded
over 90 articles.
Next, I examined the abstracts of each of the articles thus identified to ascertain whether
they analyzed teacher satisfaction either directly or in the context of specific organizational
conditions like leadership, autonomy and collaboration. The rationale for focusing on these
4

Borman and Dowling (2008) follow a similar technique of citation searches.

5

organizational conditions is that an important dimension of school climate is relational.
Collaboration, leadership and autonomy all reflect this dimension. Further, these are all policy
amenable variables. Those studies that did not meet these criteria were excluded from the
database for review. At the end of this process I retained a database of seventy-seven articles.
A review of this database revealed an overemphasis on studies that were U.S. based.
Table 1 indicates that only twenty out of the seventy-seven articles were international in scope.
Further, the vast majority of these studies were single-country studies (75 out of a possible 77).
Only two of the studies identified classify as multi-country5 studies.

Table 1: No. of Studies on Teacher Satisfaction
U.S.A. / International

Single
Multi-Country
Total

U.S.A
57
0
57

International
18
2
20

No.
75
2
77

Interestingly, around three-quarters of the literature on job satisfaction was quantitative in nature.
Table 2 shows a summary of the studies reviewed by region as well as study type. This table
again indicates a paucity of international research on the subject. By far the largest number of
studies reviewed (58) pertain to North America.

5

I defined multi-country studies as those that analyzed two or more countries. One of these is a quantitative crosscountry analysis of student outcomes based in Africa (Michaelowa & Wittmann, 2007) and the second is a multiregion as well as multi-country case-study of Europe and North America (Ylimaki & Jacobson, 2013).

6

Table 2: No. of Studies on Teacher Satisfaction by
Region /Study Type
Region
North America
Africa
Europe
Asia
Middle East
Caribbean
Total

Quantitative
40
1
6
5
4
1
57

Qualitative
18
0
2
0
0
0
20

No.
58*
1
8
5
4
1
77

Note: *Category N. America has 58 studies instead of 57 because 1 Canadian
study is included here.

One way to inform this literature further is to analyze correlates of teacher satisfaction
across different social, cultural and organizational environments. However, such variation is
difficult to obtain from a single country study. This review attempts to address this gap in the
literature by examining teacher satisfaction and specific work conditions within the United
States, as well as internationally. The guiding questions for this review are:
1. Is there variation in contextual factors associated with teacher satisfaction in a multicountry setting?
2. Are there specific aspects of schools as organizations that matter for teacher
satisfaction?
1.4 A Review of Literature – U.S.
Research has shown that organizations with more satisfied employees tend to be more
effective than organizations with less satisfied employees (Ostroff, 1992). Extending this
understanding to schools, I hypothesize that more satisfied teachers are more likely to actively

7

promote student learning within their schools. In a post-NCLB6 era where schools, teachers and
students are under heavy pressure to exhibit yearly progress in student test scores, the issue of
teacher satisfaction thus becomes especially salient. Accordingly, in this paper I first examine
demographic factors such as teacher age, teacher experience, race and student socioeconomic
status. This is followed by a review of school and classroom factors associated with job
satisfaction in the U.S. such as school infrastructure and class size. Last, I review the literature
pertaining to organizational factors associated with teacher job satisfaction. In this review the
organizational factors I research are principal leadership, teacher collaboration and autonomy,
teacher salaries and student discipline issues because these are all factors that are policy
amenable. Identifying trends across countries in one or more of these factors therefore provides
policymakers good insight into the school factors they may manipulate to improve school and
teacher effectiveness going forward. Since job dissatisfaction is an important antecedent of
teacher turnover in the U.S (see Boyd, et al 2011; Ostroff, 1992; Liu & Ramsey, 2008) I also
include studies which focus on turnover as the outcome of interest (for the U.S. only7). Table A1
lists the main findings and data used for all studies reviewed including those on teacher turnover.
1.4.1

Demographic Factors

Prior literature on the effect of demographic variables indicates a statistically significant
relationship between teacher satisfaction and teacher age and experience though the direction of
this relationship is less clear. Ma and MacMillan (1999) for example, used teacher data from the
New Brunswick Elementary School Study to explore how teacher professional satisfaction is
6

No Child Left Behind (NCLB) Act 2001 in the U.S. which promotes standards based education reform and is
based on the premise that setting measurable goals and high standards can improve individual education
outcomes. NCLB is a reauthorization of the Elementary and Secondary Education Act.

7

The literature on turnover is reviewed for the U.S. specifically because there is a well-established relationship
between teachers’ job satisfaction and teacher turnover within the U.S. The same cannot be said of international
literature as international work on turnover and satisfaction is not as widespread.

8

related to background characteristics and workplace conditions and found that older teachers
were less satisfied with their professional roles as teachers. Additional support for the fact that
age is negatively correlated with satisfaction is provided by Perie and Baker (1997). In their
study, younger teachers were categorized as having higher levels of satisfaction as compared
with older teachers. Specifically, about “thirty-five percent of public secondary school teachers
below thirty years of age [in their study] had high satisfaction levels versus less than twenty-five
percent of teachers above forty years of age” (Perie et al, 1997, p.18). In the same study they
showed that public school teachers with less than three years of experience tended to be more
satisfied than teachers with between four and nine years of experience. Teachers in the middle
group for years of experience in turn had higher satisfaction levels than the most experienced
teachers with more than twenty years of experience. It should be noted however, that although
these findings were statistically significant the differences were not large. Liu and Ramsey
(2008) found that teachers’ job satisfaction improved rather than decreased with years of
experience.
The research with respect to age and turnover is similarly fairly consistent. Teachers
who are very young (in their first couple of years of teaching) and those nearing retirement age
are much more likely to leave teaching. Thus, the relationship between age and turnover tends to
follow a U-shaped curve that has been well documented in the literature (Allensworth et al,
2009; Guarino et al, 2006; Johnson, 2006; Ingersoll, 2001).
The role of gender as an explanatory factor for teacher satisfaction, however, is less clear
cut. Culver, Wolfle and Cross (1990) examined teacher satisfaction for Blacks and Whites
separately and found that for both groups demographic variables such as age and gender were
not very important as compared with school climate factors and this finding appears to be

9

consistent with more recent work on job satisfaction (Johnson, Kraft & Papay, 2011). However,
other studies indicate that female teachers in general appear to be more satisfied with their
professional role than male teachers. Perie and Baker (1997) for example, found that in their
sample that the teachers who were identified as being the most satisfied were predominantly
female. Levels of self-reported commitment are also thought to be higher for female teachers
(Guarino et al, 2006). Similarly, Ma and MacMillan (1998) found a significant gender
difference with respect to teacher satisfaction. Their study revealed that male teachers’ sense of
professional satisfaction appears to be more affected by the organizational culture of the school
than that of female teachers.
A few scholars claim that there is an association between job satisfaction and student
socioeconomic status (Shen, Leslie, Spybrook, & Ma, 2012; Grissom, 2011). They contend that
job satisfaction is lower in schools serving a high percentage of disadvantaged students. There is
a substantial literature on turnover in schools which also documents the phenomenon that
teachers tend to leave schools with high concentrations of poor students (Borman & Dowling,
2008; Boyd et al, 2011; Scafidi, Sjoquist, & Stinebrickner, 2005, Hanushek, Kane, & Rivkin,
2004; Johnson, 2003; Shen et al, 2012; Rosenholtz & Simpson, 1990). This finding has negative
implications for student achievement in high poverty schools (see Ostroff, 1992). High poverty,
low performing schools rarely close the achievement gap because they are constantly dealing
with staffing issues caused by the high rates of teacher turnover in these schools. “An inordinate
amount of their capital-both human and financial is consumed by the process of hiring and
replacing beginning teachers who leave before they have mastered the ability to create a
successful learning culture for their students” (Barnes, Crowe, & Schaefer, 2007, p.4). In
contrast to this body of work, Ingersoll and May (2011) find that after controlling for other

10

school demographics the effect of high poverty student enrollment on turnover is no longer
statistically significant.
The role of race as a determinant of teacher satisfaction as well as teacher retention also
appears to be important in this literature. Some scholars such as Renzulli et al (2011) and
Fairchild, Tobias, Corcoran, Djukic, Kovner and Noguera (2012) have researched the issue of
student-teacher racial mismatch and its impact on satisfaction. These scholars assert that teacher
satisfaction is higher when the teacher shares the same race as the student. They highlight that
race matters for teacher satisfaction and that white teachers are likely to be more satisfied
teaching predominantly white students. Thus, the ethnic composition of the student body is
considered by some scholars to be an important determinant of teacher satisfaction.
The literature on turnover also appears to support this finding. Hanushek et al (2004) for
example found that white teachers regardless of work experience are more likely to move to
schools with lower percentages of African American and Hispanic students. Similarly, Scafidi,
Sjoquist and Stinebrickner (2005) provide empirical evidence to support the claim that race
matters. They assert that “a one standard deviation increase in the proportion of black students
in a school increases the probability that a “median type” teacher will exit a particular school in a
particular year by more than twenty percent, whereas one standard deviation changes in student
test scores, poverty or teacher pay lead to only small changes in overall exit probability (Scafidi
et al. 2005, p.6).” Other scholars such as Ingersoll and May (2012), and Borman and Dowling
(2008) have also found that the percentage of minority students is significantly associated with
white teacher turnover after high poverty status of the students is held constant. Further, the
greater the percentage of minorities (Blacks and Hispanics) in a school, the lower is the average
teacher rating of work conditions (Boyd et al, 2011).

11

Regardless of whether student demographics are strongly related to teacher satisfaction,
these factors are difficult to change from the policy perspective. There is however, a large body
of scholarly work that examines specific components of organizational culture and work
conditions as well as other school factors as determinants of teacher satisfaction. Section 1.4.3
reviews this body of literature and identifies policy amenable factors associated with teacher
satisfaction. This more organization-centric view of teacher retention and satisfaction is
espoused by scholars like Horng (2009), Ladd (2011), Liu and Meyer (2005), Ingersoll & May
(2011), and Johnson, Kraft, and Papay (2011) among others. According to Horng, teachers
consider work characteristics such as administrative support, salaries, class sizes and facilities to
be more important than student characteristics with respect to turnover decisions. Horng
theorizes that teachers are choosing to move out of high minority/high poverty districts based on
the tougher work conditions they face in these districts rather than the student characteristics
within the district. Other scholars like Ladd (2011) and Johnson, Kraft and Papay (2011) have
also found that working conditions are significantly associated with teacher satisfaction and
movement independent of other school characteristics such as the race of the student. They
contend that the apparent relationship between student demographics and teacher turnover is
driven not by teachers’ responses to their students but by the work conditions teachers have to
cope with. According to these scholars, measures of the school environment account for much
of the relationship between student demographic characteristics and teacher satisfaction and
turnover.
1.4.2

School/Classroom Factors

Few scholars have specifically focused on the relationship between school location, class
size, and teacher satisfaction. These variables are typically included as correlates in most
models. With respect to school type, Perie and Baker (1997) found that private school teachers
12

tend to be more satisfied than public school teachers and elementary school teachers are more
satisfied than secondary school teachers. They did not however, find a significant association
between class size and teacher satisfaction.
A few studies in the U.S. have also specifically focused on the role of school facilities
with respect to teacher satisfaction and retention (Schneider, 2003; Buckley, Schneider & Shang,
2005). These scholars underscore the importance of good school facilities as central to the
process of teaching and learning and the health and safety of the teachers and students. Buckley
et al. (2005) argue that serious deficiencies with respect to school facility have been noted in
several urban school districts all over the United States. Issues with indoor air quality in these
buildings have been associated with increased student absenteeism and reduced student
performance as well as a decrease in teacher health. Not surprisingly, urban teachers are less
likely to report satisfaction in their jobs (Buckley et al, 2005; Hanushek & Rivkin, 2007).
Buckley et al. (2005) contend that the benefits of facility improvement for teacher retention are
greater than or at least equal to those from salary increases. Arguably, since facility
improvement is a one-time fixed cost this factor also appears to be more cost-effective in terms
of teacher retention than across the board salary increases for teachers.
The work of Schneider (2003) corroborates Buckley et al’s research with respect to the
role of school facilities. Schneider found that inadequate school facilities negatively impact the
quality of instruction delivered by teachers, which in turn may affect student achievement. The
teachers interviewed by Schneider (2003) strongly believed that the existing school conditions
were definitely affecting their future career decisions. In the Schneider study, teachers from
Chicago and Washington D.C. were asked to record their perception of working conditions
within their schools. On a graded scale (A-F) most teachers gave their school facilities a C with

13

respect to facilities and maintenance. Over forty percent of teachers had issues with class size.
About a third of the teachers reported having little or no teacher workspace. Most teachers were
also concerned about poor indoor air quality, poor lighting and dirty restrooms. Further, a
majority of teachers in Chicago as well as Washington D.C. were of the opinion that the high
noise level in their hallways was negatively affecting their ability to teach. As Johnson (2006)
aptly observed, poorly maintained schools send a negative signal about the state of public
education because they indicate that public education is low on the list of a community’s
priorities. Given the apparent influence of school facilities further research appears to be
necessary to get more clarity on the importance of school facilities vis-à-vis teacher satisfaction
and other organizational factors.
1.4.3

Organizational Factors

Ostroff (1992) has shown that satisfied employees tend to be more effective within
organizations. Accordingly, this section summarizes the research on teacher satisfaction and
turnover with respect to the organizational milieu within schools. I specifically focus on school
organizational factors such as principal leadership, teacher collaboration, teacher autonomy,
teacher salaries, student behaviors and student discipline and teachers’ job satisfaction.
Research on school organizations has found that work condition factors such as
administrative support and teacher control over classroom procedures are important for teacher
satisfaction (Culver et al., 1990; Perie & Baker, 1997; Taylor & Tashakkori, 1995). This finding
has been corroborated elsewhere in the literature as well (see Bogler, 2002; Guarino et al, 2006;
and Stockard & Lehman, 2004). In keeping with this view, other scholars have researched
various aspects of the work environment such as principal leadership, (Allensworth et al. 2009;
Boyd et al, 2011; Brown & Wynn, 2009) teacher autonomy, and teacher participation in school
processes (Ingersoll, 2001; Johnson & Birkeland, 2003).
14

In this review, a focus on leadership among the organizational factors within a school
context is motivated by the fact that the principal is considered to be a central figure in any
discussion of the school as a workplace. As a key figure in school administration, the principal
is responsible for making decisions in the day-to-day life of schools, which impact the social life
of the school and all those who work in it (Lortie, 1975). As the official head of the school, the
principal is also the “instructional leader of the school” (Lortie, 1975, p.197). Thus, it is the
principal who allocates the resources that alter teacher working conditions and to a great extent,
principals’ actions create distinct working environments within schools which are associated
with teacher satisfaction.
Given that the teacher-principal relationship is a complex one, several scholars have
focused their research specifically on how principal leadership affects teacher satisfaction and
retention. Rosenholtz (1990), and Lee et al. (1991) for example, found that “principal buffering”
is particularly helpful for novice teacher commitment. This is because in “loosely coupled”
public schools this buffering action allows teachers the autonomy they need to manage their
classrooms effectively. Similarly, Moore (2012) found that principal leadership decreased the
odds of teacher dissatisfaction while student problems increased teacher dissatisfaction.
Principal leadership also emerged as a strong predictor of teacher satisfaction in the work by
Taylor and Tashakkori (1995) and Stockard and Lehman (2004).
Hulpia, Devos, and Rosseel (2009) delved a little further into the concept of principal
leadership by centering on the notion of distributed leadership. Distributed leadership according
to Hulpia et al., is the formal distribution of supervisory and supportive leadership functions and
is indicative of interactions between members of a cohesive leadership team. Implicit in the
notion of distributed leadership is participative decision-making of the entire leadership team.

15

Hulpia et al’s research examined the relationship between distributed leadership, context
variables and teacher satisfaction and commitment. They found that the cohesion of the
leadership team and the amount of leadership support was indirectly related to job satisfaction.
The authors failed to find a significant relationship between participative decision-making and
job satisfaction (see Taylor & Tashakkori, 1995 for similar findings).
Previous literature on leadership has often centered on the principal’s supervisory role
within schools. Interestingly, Hulpia et al. (2009) found that supervision as a leadership function
played little role in predicting teacher job satisfaction. In fact, supervision was not significantly
related to either teacher job satisfaction or organizational commitment. This finding is important
because it has implications for teacher evaluation. Specifically, it implies that schools where
teachers are regularly evaluated and which have several formative or summative assessments do
not necessarily have more satisfied or committed teachers. This study also underscores the fact
that not all leadership functions should be distributed. In particular, supervision of teachers may
need to be concentrated within one leader.
Several studies have examined the link between principal leadership and retention as well
(Allensworth, 2009; Boyd et al., 2011; Culver et al., 1990; Ingersoll & May, 2012; Ladd, 2011,
Pogodzinski, Youngs, Frank & Belman, 2012). Overall, these studies have found that
administrative climate and leadership are associated with teachers’ career decisions.
A second organizational factor I focus on in this review is collaboration. Much has been
written about the isolated nature of teaching in previous years. Lortie (1975) noted this
phenomenon of teacher isolation, which he attributed to the “egg-crate” nature of schools. Thus
defined, school structure in the U.S. acts as a disincentive to teacher collaboration. The “eggcrate” structure of schools notwithstanding, the notion of teacher collaboration appears to be

16

capturing the attention of the scholarly community in recent years. Underscoring the importance
of teacher collaboration, Ladd (2011) asserts that the notion of collaboration is central to any
discussion on the workplace. Within the scholarly community there is an understanding with
respect to collaboration that cultures that promote collaboration are the ones most positively
associated with higher satisfaction and retention among teachers. Johnson and Birkland (2003)
for example, found that teacher retention was higher with “integrated professional cultures”
organized around collegial efforts. Support for teacher collaboration is also provided by Culver
et al. (1990), and Rosenholtz (1990). In particular, Rosenholtz (1990) found significant
differences with respect to progress made on reforms between schools where teachers
collaborated and those where they did not. She concluded that when teachers have shared goals
for student learning and achievement, students benefit. However, while these results look
promising in terms of teacher retention and satisfaction, further research is necessary to get a
more robust picture of the effects of teacher collaboration on teacher satisfaction.
Scholars have also examined teacher autonomy as an explanatory factor for teacher
satisfaction as well as teacher retention. Autonomy as a concept has been explicated as “a
worker’s freedom to schedule work and determine the procedures used to carry it out” (Firestone
& Pennell, 1993, p.498). Autonomy leads to successful instructional practice because teachers
can identify ways in which their own work contributes to student learning. There is a
consistently positive relationship between teacher autonomy and satisfaction. For example,
Moore (2012) found that teacher autonomy decreased the odds of teacher dissatisfaction. Perie
and Baker (1997) noted that teachers with greater autonomy showed higher levels of satisfaction
than teachers who felt they had less autonomy. Similarly, Renzulli et al (2011) and Lee et al
(1991) in their work on teacher satisfaction provided support for organizational models that

17

promote autonomy. Elsewhere in the literature the loss of autonomy has been notably associated
with teacher burnout.
The relationship between autonomy and turnover is also fairly unambiguous. Ingersoll
and May (2011) found that schools with higher average levels of classroom autonomy had lower
levels of turnover. The turnover of math teachers was especially strongly related to classroom
autonomy. Ingersoll et al reported that a one unit increase in autonomy was associated with a
seventy percent decline in the odds of a math teacher departing. For science teachers, however,
classroom autonomy was not significantly associated with turnover. Another surprising finding
with respect to autonomy is that it has the potential to ameliorate the negative impact of racial
mismatch on teacher attitude (Renzulli et al, 2011).
The literature is not definitive about the association between teacher salary and teacher
satisfaction and retention. Some studies have uncovered a weak relationship between teacher
compensation and teacher satisfaction and retention (Ingersoll & May, 2012; Lee, et al. 1991;
Perie & Baker, 1997). Others, such as Liu and Meyer (2005) have found low compensation to be
a leading cause of teacher dissatisfaction. Low salaries have also been associated with increased
organizational commitment (Firestone & Pennell, 1993). This finding is somewhat
counterintuitive if we consider the labor market theory of supply and demand (see Guarino et al.
2006). Firestone et al. (1993) however, explicate this finding by suggesting that workers with
low salaries develop alternative rationales for remaining at their job. Still others find that salary
is less important than work conditions vis-à-vis teacher satisfaction (Hanushek, Kane, & Rivkin,
2004; Hanushek & Rivkin, 2007; Horng, 2009). Horng (2009) for example finds that on average
the difference between $0 and $8000 in additional annual salary is not as important as

18

differences between a clean and safe work environment versus not having such an environment
or having administrator support versus not having administrator support.
As a counterpoint to the weak relationship mentioned above, much of the retention
literature provides extensive empirical evidence that teachers are more likely to transfer when
they work in districts paying lower wages especially relative to alternative wage opportunities
(Allensworth et al. 2009; Loeb, Darling-Hammond, & Luczak, 2005; Guarino et al. 2006;
Hanushek, 2004). Loeb et al. (2005) for example, contend that an increase in relative salaries
from one to two times the local wage decreases turnover by seventy-five percent. However,
others such as Johnson (2006) find that teachers are more likely to report intent to leave due to
poor working conditions than due to salary.
A number of studies have established a negative relationship between poor student
discipline and teacher satisfaction (Ingersoll, 2001; Liu & Meyer, 2005; Perie & Baker, 1997).
In fact, student discipline issues have been cited as a major reason for teacher dissatisfaction,
second only to compensation (Liu & Meyer, 2005). Kennedy (2005) in her book showed how
teachers are urgently driven by a need to maintain lesson momentum, avoid disruptions at all
costs, and move through their day in an orderly and stable way. Student discipline issues are
perceived by teachers as a disruption that takes time away from the process of teaching and
learning. Given the high importance teachers attach to maintaining classroom discipline at all
costs the finding that assignment to a safe and orderly school is one of the most significant
influencers of satisfaction (Stockard & Lehman, 2004) is hardly surprising.
Understanding teacher dissatisfaction with student discipline problems also has the
potential to provide some insight into teachers’ initial enthusiasm for the profession and their
subsequent wish to leave teaching. For many individuals student discipline problems may act as

19

a check to initial idealism about teaching because they are now forced to deal with students who
do not come to school motivated to learn. This may be the case in inner city schools that
typically experience higher student discipline problems (Ingersoll, 2001) as well as higher rates
of turnover. Within the range of teachers, Liu and Meyer (2005) found that minority teachers
expressed more disappointment with student discipline issues than non-minority teachers.
Based on these observations, policy interventions that use financial incentives to address
teacher turnover and dissatisfaction may fail to adequately address student discipline problems
because the correlation between teachers’ perceptions of student discipline issues and
compensation is fairly low. Liu and Meyer (2005) interpret this to imply that teachers’
satisfaction with the financial aspects of their profession has little bearing on their dissatisfaction
with student discipline problems. In other words, increasing teacher salaries does not
compensate for the disappointment teachers experience with respect to their students’ behavioral
issues. A more effective policy solution to address discipline issues and improve teacher
satisfaction may be to improve the quality of daily communication between parents and teachers
with respect to student problems. Parents have a good understanding of the behavioral issues of
their children. Enlisting parental support to address disciplinary issues before they become a
problem ensures that teachers are not left to single-handedly cope in the classroom.
The turnover literature too notes a significant relationship between student discipline and
the likelihood of teacher turnover. Turnover rates are lower in schools with lower levels of
discipline problems. Ingersoll (2001) for example, found that “a one unit difference in reported
discipline problems between two schools is associated with a forty-seven percent difference in
the odds of a teacher departing” (Ingersoll, 2001, p.518). Ingersoll and May (2011) have found a
similar association between poor behavioral climate in schools and turnover.

20

While the literature points to a clear association between student discipline and teacher
satisfaction, we do not as yet have a clear understanding of the causal mechanisms by which
student discipline problems make teachers discontented with their profession or policy solutions
that might help mitigate discipline problems.
In summary, recent work that is U.S. based appears to indicate that organizational factors
tend to explain more variation in teacher satisfaction than do demographic factors.
Notwithstanding this understanding, gaps remain with respect to the kinds of organizational
factors that really matter for teacher satisfaction as well as the processes by which these
organizational factors affect satisfaction. The next section briefly reviews the international
literature on satisfaction.
1.5 A Review of Literature – International
In contrast to the relatively well-documented body of work on teacher satisfaction and
retention in the U.S., there is a limited literature on teacher satisfaction in the international
context. Internationally, teacher satisfaction has been researched in a somewhat piecemeal
fashion with the research spanning a diverse selection of countries such as China, Norway,
Albania, Israel, Cyprus, the United Kingdom and Australia. The section below attempts to
briefly summarize salient findings from this collection of scholarly work and to highlight points
of divergence from the U.S. literature on teacher satisfaction.
1.5.1

Demographic Factors

Most studies conducted internationally have found that context is a powerful predictor of
overall teacher satisfaction. Zembylas and Papanastasiou (2004) for example, explored teacher
satisfaction in Cyprus and found that unlike teachers in developed countries like the U.S., U.K.,
Australia and New Zealand, Cypriot teachers choose to be in the teaching profession because of

21

salary, working hours and holidays. Their career choice appears to be guided more by pragmatic
considerations than professional idealism. Additionally, teacher age was significantly associated
with higher satisfaction in their study. Since the main criterion for promotions in Cyprus is
experience, the authors attributed the importance of the age variable to the significant increase in
salary and promotions that Cypriot teachers receive over time based on experience. Another
possible explanation is that as teachers improve their teaching practice their satisfaction levels
also go up.
In Jinzhou, younger and less experienced teachers were found to be more satisfied than
older teachers (Chen, 2010). However, in the Gansu province of China, Sargent and Hannum
(2005) found that the most dissatisfied teachers were the youngest teachers and the most senior
teachers (indicating a U-shaped curve for the age variable). Additionally, female teachers were
more satisfied than male teachers. These divergent results in different countries are perhaps
explicable by fact that contexts differ in different places.
Using a large-scale dataset (PASEC), Michaelowa and Wittmann (2007) researched
teacher satisfaction and its impact on education quality in Francophone Sub-Saharan Africa. In
addition to the positive correlation between teacher satisfaction and student achievement, they
found that if a teacher’s qualification increased beyond a certain level (Bachelor’s degree) it
resulted in a mismatch between professional realities and teachers expectations and actually
decreased teacher satisfaction in the net. Sargent and Hannum (2005) also found that better
qualified teachers tend to feel more dissatisfied than do less qualified teachers.
Internationally, there appears to be comparatively less focus on race vis-à-vis teacher
satisfaction. However, Sargent and Hannum (2005) did find that teachers who are more socially
similar to the local community are more satisfied than those that are not. With respect to

22

socioeconomic status of students Sargent and Hannum (2005) found that village income
significantly differentiates satisfied and dissatisfied teachers. Overall, more satisfied teachers
appear to teach in schools where the economic resources for the support of teaching are more
readily available.
1.5.2

School/Classroom Factors

With respect to school infrastructure, Sargent and Hannum’s (2005) findings in the
Gansu province of rural China corroborate those in the U.S. more closely than the study by Chen
(2010). Specifically, teachers in their analysis were found to be more satisfied in larger schools
which have more resources and in schools with better opportunities for advancement. However,
they also emphasize that a synergy between the teachers and the local community improved
teacher satisfaction suggesting the importance of ties to local areas. Their study indicates that
teachers in more remote areas are more likely to have higher satisfaction. This finding is
somewhat counterintuitive until we look at the context in Gansu. It is possible that in Gansu ties
with the local community are stronger than in most large urban cities thereby mitigating teacher
dissatisfaction in that province. The role of school facilities with respect to teacher satisfaction
was also found to be important in Albania and Jamaica (Kloep & Tarifa, 1994; Rogers-Jenkinson
& Chapman, 1990).
With respect to class size, Michaelowa and Wittman (2007) found that a large class-size
and double-shift classes were negatively correlated with both teacher satisfaction and student
achievement. Although this is one of the few studies internationally that examines the effect of
teacher satisfaction on student achievement, it does not include controls for correlates of
satisfaction such as teacher autonomy and salary.

23

1.5.3

Organizational Factors

Among the scholars who have focused on organizational leadership internationally,
Bogler (2001) found that the principal’s transformational leadership affected teachers’
satisfaction directly and indirectly through their occupational perceptions. In this study,
teachers’ occupational perceptions are understood as aspects of the teaching profession such as
autonomy, professional development, professional prestige, instructional freedom etc. Bogler
hypothesized that the more involved teachers are in decision-making, the more satisfied they are
with their jobs. Support for the relationship between a teacher’s participation in decision-making
and job satisfaction was also provided by Zembylas and Papanastasiou (2005) in Cyprus. As
expected, the authors found that teachers who are empowered to make instructional and
organizational decisions have higher levels of satisfaction.
Two studies, one by Abu-Saad and Hendrix (1995) and another by Bogler (2002)
examined teacher satisfaction in Israel. Both studies centered on principal leadership as the
dominant organizational climate factor. They found that leadership and autonomy were
significantly related to teacher satisfaction. The fact that principal leadership emerged as the
dominant school climate variable associated with satisfaction is attributed by the authors to the
power relationships within traditional Bedouin Arab communities in Israel in which power and
responsibility lies with the leader. In these communities, the leader is central to providing
direction, purpose and motivation and also key to determining the social relations of the group.
School leadership also emerged as a predictor of teacher satisfaction in the UK, Australia and
New Zealand (Dinham & Scott, 2000; Dinham, 2005). Interestingly, teacher satisfaction with
their current workload was the strongest predictor of overall satisfaction in these countries. This
finding points to the importance of context vis-a-vis teacher satisfaction in developed and

24

developing countries because teacher workload was negatively correlated with teacher
satisfaction in rural China (see Hannum et al. 2005).
Positive social interactions with colleagues have typically been associated in the literature
with reduced stress for teachers. In a follow-up to their 2004 study of Cypriot teachers,
Zembylas et al. (2006) found that a main source of teacher satisfaction was working
collaboratively with colleagues and having opportunities for personal and professional growth.
This finding emphasizes intrinsic rather than extrinsic factors as the most significant source of
teachers’ job satisfaction. It also appears to be more in keeping with findings from developed
countries and may potentially be common internationally regardless of country context.
Additional support for the satisfaction-collaboration relationship was provided by
Skaalvik and Skaalvik (2011) who found that satisfaction was positively related to relations with
colleagues, and relations with parents through teachers’ feelings of belonging. Both feelings of
belonging and emotional exhaustion were significantly related to job satisfaction.
Support for organizational structures that enhance teacher collaboration was also
provided by Sargent and Hannum (2005). These authors found that opportunities for
professional development also had a significant impact on teachers’ job satisfaction. Working
collaboratively with other teachers was found to be positively associated with teacher satisfaction
in Cyprus (Zembylas, 2006) and in the study by Dinham (2000). Powerful dis-satisfiers
identified in the Zembylas (2006) study included lack of adequate respect for teachers by the
community, lack of collegial relationships, and lack of teacher autonomy. The lack of autonomy
experienced by Cypriot teachers was attributed to country specific context as Cyprus has a
centralized educational system. Professional autonomy was also found to be important in the
Albanian context where it accounts for a considerable portion of the variance in job satisfaction

25

(Kloep & Tarifah, 1993). The work by Kloep et al. (1993) reveals that Albanian teachers in
general appear to have good relationships with their colleagues. The strength of these good
relations may help explain the high level of job satisfaction observed in that country.
Teachers’ relationships with other teachers were also significantly related to teacher
satisfaction in both public and private schools in Jamaica (Rogers-Jenkinson & Chapman,1990).
Approximately twenty percent of the variance was explained by school factors such as
interpersonal relationships with teachers, relationships with parents and organizational structure
(identified by the authors as a collection of variables pertaining to teacher participation in the
formation of school goals and policies as well as teacher autonomy).
Recent research by Chen (2010) in Jinzhou city in the north of China found that for
Chinese middle school teachers’ salary was significantly associated with teachers’ future career
plans and intent to move. Contrary to these findings, as well as, contrary to some research from
the U.S., Sargent and Hannum (2005) did not find salaries to be significantly related to teacher
satisfaction. Further evidence to support the notion that salary is not a key determinant of
satisfaction is provided by Abd-el-Fattah (2010) whose study investigated the longitudinal
effects of a pay-increase scheme in Egypt on teacher satisfaction. Their study showed that pay
increase did not have a significant impact on teacher satisfaction. Further, the impact of pay
increase was stronger for teachers with a high school diploma (low attainment) vs. those who had
a university degree (high attainment). Male teachers, however, were more satisfied than female
teachers with their profession indicating a significant gender effect associated with salary
increase.
With respect to student discipline issues, most of the international studies reviewed found
discipline issues are negatively associated with teacher satisfaction. The study by Skaalvik and

26

Skaalvik (2011) in Norway for example found that job satisfaction was negatively correlated
with time pressure and discipline problems through the emotional exhaustion of teachers and
motivation to leave the teaching profession. Similarly, Zembylas (2006) found that student
discipline issues in Cyprus were negatively associated with Cypriot teachers’ satisfaction levels.
Interestingly, in many of the international studies reviewed I found that major dissatisfiers were primarily found in the broader societal context of which each school is a part
rather than in the conditions of work within schools. In particular, the community’s poor opinion
of teachers and the perceived status of teachers were found to be powerful dis-satisfiers in
Australia and Cyprus (Dinham, 2000; Zembylas, 2006). In Jamaica, school prestige was a
significant predictor of satisfaction for public school teachers (Rogers-Jenkinson & Chapman,
1990). Specifically, the authors found that school prestige explained thirty-eight percent of the
variance in teachers’ satisfaction. The high level of importance to the school prestige variable is
attributed to the fact that typically teachers’ status in their community operates as a nonmonetary incentive to offset low wages. In many countries, this understanding is being eroded
by a perceived drop in teaching prestige thereby altering the incentive value of the job (RogersJenkinson & Chapman, 1990). In this respect, Jamaican elementary school teachers differ from
elementary school teachers in the United States.
In the U.S., teachers who are more satisfied assign a higher value to recognition by
administrators and supervisors as compared to teachers in Jamaica. The latter appear to assign
more importance to the reactions of the community. Another point of difference is the
importance attached to improving school facilities in the Jamaican context. Results of this study
suggest that one of the most direct ways to increase teacher satisfaction is to improve the
physical conditions in which teachers work. Poor physical working conditions are also relevant

27

in the Albanian context as shown by Kloep & Tarifa (1994) though this factor does not seem to
significantly alter teacher satisfaction there.
1.6 Policy Implications & Limitations of the Literature
This review shows that teacher satisfaction is considered to be an important antecedent of
teacher turnover as well as student achievement at least within the U.S. From a policy
perspective therefore, it is important to identify the correlates of teacher satisfaction. A fruitful
approach for identifying such correlates is to analyze teacher satisfaction across diverse social,
cultural and organizational situations. Yet, this review of the literature reveals certain limitations
which are enumerated as follows.
1.6.1

Lack of Multi-country Analysis

The review suggests that only a few scholars (Sargent & Hannum, 2005 and Michaelowa
& Wittman, 2007) have attempted to use large-scale survey data to understand and explore
teacher satisfaction. Large-scale survey data have the advantage of generalizability of inferences
within and across countries which is not necessarily true of self-collected or administrative data.
Further, systematic multi-country analysis examining the relationships between teacher
demographics, school demographics, organizational factors and teacher satisfaction has not been
attempted.
1.6.2

Limited Focus on Asia

A specific focus on Asia is interesting because of its rich cultural diversity. However, as
Table 1 shows, only five studies out of a total of seventy-seven studies reviewed are focused on
teacher satisfaction in Asia. Two of these papers focus on rural China, 1 study is on Pakistan, 1
focuses on South Korea and the last is an analysis on Singapore. Other countries in Asia are
seriously under represented. Research into the organizational factors that affect teacher
satisfaction levels in the Asian context would help to extend existing knowledge about
28

determinants of teacher satisfaction in a region outside of North America. Such research has the
potential to unearth links in the satisfaction-work condition relationship, which could help
inform the scholarly work in this area within the United States as well.
1.6.3

Limited Attention to Important Correlates

Finally, the literature is ambivalent with respect to the relationship between demographic
teacher attributes like teacher age, teacher experience and teacher satisfaction. Similarly,
research on the role of school infrastructure and teacher satisfaction in different contexts is
lacking. The analyses to follow will specifically examine the direction of the relationship
between such demographic and school factors as teacher experience and school infrastructure
with teacher satisfaction across multiple countries.
1.7 Conclusion
There is a dearth of systematic multi-country analysis examining the relationships
between a comprehensive set of student, teacher school, and organizational factors and teacher
satisfaction leaving considerable room for future analysis. Only a few scholars (Sargent &
Hannum, 2005 and Michaelowa & Wittman, 2007) have used large-scale data to understand and
explore the concept of teacher satisfaction. Further, despite a tentatively acknowledged
connection in the literature between teacher collaboration in schools and school improvement
there has been little systematic effort to understand the association between collaboration and
teacher satisfaction, to examine whether principal leadership is associated with teacher
collaboration in countries outside the U.S.A. or to parse out the kinds of collaborative
interactions associated with principal leadership in international contexts. From a policy
perspective, it is worth testing whether these relationships are generalizable internationally or
whether they are more U.S. context-specific.

29

This dissertation addresses these gaps in the literature by conducting a systematic
examination of teacher satisfaction with a focus on two key organizational factors - principal
leadership and teacher collaboration. I limit my focus to these two factors because these are
policy amenable variables that are also relational in nature. Specifically, in chapter 2, I analyze
the relationship between principal leadership style and teachers’ job satisfaction across six Asian
countries (Hong Kong, Indonesia, Malaysia, Mongolia, Thailand, and Taiwan) and the United
States to see if there is variation in the facets of principal leadership associated with teacher
satisfaction in Asia as compared with the United States. An empirical analysis of aspects of
leadership that affect teacher satisfaction levels in the Asian context will help to extend our
understanding of teacher satisfaction internationally. Teacher collaboration has been
conceptualized as an important factor contributing to teacher development. The beneficial
effects of teacher collaboration are widely acknowledged in the United States. Teachers are
encouraged to share and develop their expertise with each other and to make use of opportunities
for collaborative lesson planning, classroom observations, peer coaching and mentoring. There
is however ample opportunity for further empirical exploration of the relationship between
teacher collaboration and teacher satisfaction in other developed and developing countries
internationally. In chapter 3, I analyze five OECD countries – Denmark, Mexico, Brazil, Poland
and Hungary, to generate a richer and more comprehensive understanding of teacher
collaboration, school leadership and teacher satisfaction in a context outside of the United States.
Since the literature on teacher collaboration is not definitive, my dissertation will contribute to
the existing body of work by subjecting rich measures of teacher collaboration to rigorous
empirical investigation in a multi-country setting using nationally representative large scale
survey data – OECD’s Teaching and Learning International Survey (TALIS 2008).

30

APPENDIX

31

Table A1: List of Studies Reviewed with Key Findings
No.

Author(s)/Year

Key Findings

Data Source/Sample Size

1

Abd-El-Fattah, S.M.
(2010).

Pay increases did not have a significant effect on teachers' job satisfaction.
Pay increases appear to increase male teacher satisfaction more than female
teacher satisfaction.

Survey (N=155).

2

Abu-Saad, I., &
Hendrix, V. L. (1995).

Principal leadership and autonomy were significantly related to teachers’ job
satisfaction. The more supportive the leadership style, the higher the
satisfaction.

2 surveys administered - Wanous &
Lawler's Job Satisfaction Survey ;
Horowitz and Zak's organizational climate
survey (N=373 teachers)

3

Achinstein, B. (2002).

A focus on micro-political processes (border, politics, and ideology) in
teacher communities provides different opportunities for organizational
learning.

2 urban public middle schools in San
Francisco Bay Area (6th-8th grade)

4

Allensworth, E.,
Ponisciak, S., &
Mazzeo, C. (2009).

Student classroom behavior was the strongest predictor of mobility.
Teachers were more likely to stay in schools where they have supportive
leadership and opportunities to collaborate.

Teacher personnel records: Elementary
school survey data from 2002-3 to 2006-07
(N=16,000 ES teachers)

5

Barnes, G., Crowe, E., &
Schaefer, B. (2007).

Turnover costs were significant at the district as well as school level.
Teachers leave high minority/low performing schools at higher rates. Low
levels of teacher experience were associated with high turnover but other
teacher demographics not significantly associated.

Own data in 5 school districts. Database
linked to CCD

6

Bogler, R. (2001).

Principals’ transformational leadership affected teachers satisfaction both
directly and indirectly through their occupational perceptions.

MLQ (transformational behavior);
Friedman's questionnaire (decision-making
style),Yaniv's questionnaire (occupational
perceptions). N=930 (teachers) Not random
sample

7

Bogler, R. (2002).

They constructed profiles of teacher satisfaction using discriminant analysis.
Teacher perceptions of their occupations and principal leadership styles were
used as predictors

Quantitative questionnaire administered in
1997 to 930 teachers

8

Borman, G. D., &
Dowling, N. M. (2008).

Personal characteristics of teachers (backgrounds and qualifications) were
found to be important predictors of teacher turnover. Teacher attrition was
more strongly moderated by characteristics of teachers' work conditions than
previously noted in the literature.

Meta-analysis and review of the literature
(N=34 studies)

32

Table A1 (cont’d)
No.

Author(s)/Year

Key Findings

Data Source/Sample Size

9

Boyd, D., Grossman, P.,
Ing, M., Lankford, H.,
Loeb, S. & Wyckoff, J.
(2011).

Teachers’ perceptions of school administration have the greatest effect on
teacher retention. Dissatisfaction with work conditions was the number one
reason given by teachers for leaving. Lack of administrative support was the
most cited reason for dissatisfaction with work conditions.(40%)

Administrative data provided by the
NYCDOE (N= 4360 teachers)

10

Branch, G.F., Hanushek,
E.A., & Rivkin, S. G.
(2013).

Management of teacher quality is an important pathway through which
principals affect school quality.

Administrative data constructed as part of
the UTD Texas Schools Project (N=7420
principals)

11

Brown, K. M. & Wynn,
S. R. (2009).

Teacher retention is higher if principals are proactive in supporting new
teachers. Specifically, a commitment to professional growth/excellence,
supporting collegiality among teachers and fostering unofficial professional
learning communities are important.

Existing Literature. N=12 ES principals

12

Buckley, J., Schneider,
M., & Shang, Y. (2005).

Facility quality is an important predictor of teachers' decisions to leave their
current employment.

Self-collected survey data, Washington
D.C. (N=N/A)

13

Cha, Y.-K., & Ham, S.H. (2012).

Teacher collegiality is understood largely as teachers' collective effort to deal
with uncertainties that arise from their approach to teaching as a
constructivist endeavor.

OECD Teaching and Learning International
Survey, TALIS 2008 (N=149 Schools)

14

Chen, J. (2010).

Chinese middle school teachers were more satisfied with social
acknowledgement and interpersonal relationships that with leadership, salary,
physical work conditions and opportunities for advancement.

A convenience sample of 294 teachers in
Jinzhou City in China

15

Cosner, S. (2009).

Broad actions taken by the principal regarding cultivation of collegial trust
were important. These included an increase in interaction time within
department meetings, staff meetings, and site-based professional
development

Self-collected data on 11 nominated HS
principals

16

Culver, S.M., Wolfle,
L.M., & Cross, L.H.
(1990).

The study found that demographic controls were of little importance
compared to school climate variables.

Representative sample of public school
teachers in Virginia obtained from the
Virginia DOE (N=722 teachers)

33

Table A1 (cont’d)
No.

Author(s)/Year

Key Findings

Data Source/Sample Size

17

Dinham, S. & Scott, C.
(1998).

Teacher experience was positively related to satisfaction on the following
scales - school leadership; school infrastructure; school reputation; student
achievement and self-growth. It was significantly negatively related to the
merit promotion scale.

Self-collected (N=892 teachers)

18

Dinham, S. (2005).

School Leadership (inclusive of teacher leadership) was found to be a key
factor in the achievement of outstanding educational outcomes.

Self-collected (N=38 secondary schools in
New South Wales)

19

Fairchild, S., Tobias, R.,
Corcoran, S., Djukic,
M., Kovner, C., &
Noguera, P. (2012).

Components of relational demography directly affect teacher job satisfaction
over and above the effects of work attitudes.

2003-04 SASS (N=8,665 teachers in 1,992
schools)

20

Fallon, G. & Barnett, J.
(2009)

This study found that trust was the glue that held people together and kept
collaboration going. However, participants limited critical analysis of one
another's practices completely off the agenda thereby also limiting collegial
interaction in some senses.

Self-collected (N=18 teachers)

21

Firestone, W. A., &
Pennell, J.R. (1993).

The authors found that seven key workplace conditions - feedback,
participation, collaboration, autonomy, job design characteristics, resources,
and learning opportunities all contribute to teacher commitment.

Existing Literature

22

Grissom, J.A. (2011).

Principal effectiveness is associated with greater teacher job satisfaction and
a lower probability that the teacher leaves the school the same year. The
positive effects of principals are even greater for disadvantaged schools.

2003-04 SASS; 2004-05 Teacher Followup survey (N=30, 690 teachers in 6290
schools)

23

Grissom, J. A. & Loeb,
S. (2011).

The article shows that organization management effectiveness is particularly
important for school improvement. Their findings are not consistent with the
notion that instructional leadership is important. Teacher satisfaction is not
associated with Instructional Management.

Self-collected survey data on principals,
asst principals, teachers, students, and
parents combined with administrative data
(N=314 principals; N=482 teachers)

34

Table A1 (cont’d)
No.

Author(s)/Year

Key Findings

Data Source/Sample Size

24

Guarino, C.M.,
Santibanez, L., & Daley,
G.A. (2006).

Teachers teaching in urban and high poverty schools and those reporting
dissatisfaction with salary had higher attrition rates and a decreased
commitment to teaching. Schools with higher levels of collegiality and
administrative support had lower attrition.

Existing Literature (N=46 studies)

25

Hanushek, E. A., Kain,
J. F., & Rivkin, S.G.
(2004).

Teachers transfer or exit from the Texas school system more as a reaction to
the characteristics of their students than in response to better salaries in other
schools. This leaves disadvantaged students with relatively inexperienced
teachers.

Self-constructed database on Texas public
schools

26

Hanushek, E. A., &
Rivkin, S.G. (2007).

Difficult working conditions drive much of the difference in the turnover of
teachers and the transfer of teachers across schools. Overall salary increases
for teachers are expensive and ineffective.

Texas public school data

27

Horng, E.L. (20 09).

Teachers identify working conditions such as school facilities, administrative
support, class size and salary as more important than student characteristics
when selecting a school to work in.

Survey data (N=531 teachers)

28

Hulpia, H., Devos, D., &
Rosseel, Y. (2009).

Context and leadership variables explained 47% of the variance in the model
on organizational commitment. Context variables have a small and
significant association with job satisfaction. Leadership variables are the
most important for job satisfaction.

Self-collected survey (N=1522 teachers)

Ingersoll, R.M. and
May, H. (2011).

Organizational conditions were strongly related to minority teachers
departing. The degree of individual classroom autonomy and the level of
collective faculty decision making were the organizational factors most sig
associated with minority turnover. These factors were more significant than
salary, classroom resources or teacher professional development.

Uses all 6 cycles of SASS, Teacher Followup Survey

Ingersoll, R.M. (2001).

Teacher job dissatisfaction accounts for much of teacher turnover. The data
show that in particular inadequate support from school administration,
student discipline problems, limited faculty input into decision making, and
to a lesser extent low salaries are all associated with higher rates of turnover
after controlling for the characteristics of both teachers and schools.

1991-92 SASS/TFS (N=6733
ES/Secondary teachers)

29

30

35

Table A1 (cont’d)
No.

Author(s)/Year

Key Findings

Data Source/Sample Size

Johnson, S. M., Kraft,
M. A., & Papay, J.P.
(2011).

Teachers who are more satisfied tend to stay longer in schools with a more
positive work context independent of student demographics. The specific
elements of school work conditions that matter most to teachers are the social
conditions - principal leadership, school culture, and relationships among
colleagues. Providing a supportive context in which to work (for teachers)
appears to improve student achievement

MassTeLLS- Survey data (N=25,135
teachers in 1,142 schools)

32

Johnson, S. M. (2006).

Remarkably few low income schools provide the working conditions
(opportunities for collaboration and professional development, safe facilities,
flexible curriculum) that teachers need to do their job well. When teachers
are sustained and successful in their work they are more likely to stay
teaching.

Review of existing literature

33

Johnson, S. M., &
Birkeland, S. E. (2003).

For beginning teachers the respect and support of administrators was key to
their satisfaction.

Project on the Next Generation of Teachers
(N=50 teachers)

This is a conceptual review of job satisfaction. The chapter provides an
overview of major situational, dispositional theories of job satisfaction.

Review of existing literature

Their meta-analysis (N=312) reveals that the mean true correlation between
job satisfaction and job performance is .30

Review of existing literature

31

34

35

Judge, T. A., Parker, S.
K., Colbert, A.
E., Heller, D., & Ilies, R.
(2001).
Judge, T.A., Thoreson,
C. J., Bono, J. E., &
Patton, G.K. (2001)

36

Kelchtermans, G.
(2006).

A cultural and a micropolitical perspective are required to disentangle
collaboration, and autonomy and they should be understood as mediated by
other workplace conditions.

Review of existing literature

37

Kennedy, M. (2005).

The circumstances of daily teaching compel changes not the persuasiveness
of reform ideas. Professional development programs have the potential to
make powerful changes in teaching practice.

Self-collected data with purposive sampling
(N=45 teachers in 16 schools)

38

Kloep, M., & Tarifa, F.
(1994).

Teachers in general are quite satisfied with their job in Albania. Autonomy is
very important for teacher satisfaction in the Albanian context accounting for
27% of the variance in job satisfaction. Satisfaction is also predicted by job
security, respect from the community and cooperation with colleagues.

Large scale survey (N=349 teachers)

36

Table A1 (cont’d)
No.

Author(s)/Year

Key Findings

Data Source/Sample Size

39

Koh, W. L., Steers, R.
M., & Terborg, J. R.
(1995).

Transformational leadership has substantial add-on effects to transactional
leadership in the prediction of organizational commitment and satisfaction
with the leader in this sample. It had little direct effect on student academic
performance.

Self-collected data (N=100 randomly
selected secondary schools)

40

Ladd, H. (2011).

Working conditions are highly associated with teachers intended departure
from schools even after controlling for school demographics like race.
School leadership emerges as the most salient dimension of work conditions

Administrative data (N=22,941 ES
teachers, N=9,101 for MS; N=10,829 for
HS)

41

Lee, V.E., Dedrick, R.
F., & Smith, J.B. (1991).

42

Leonard, L., & Leonard,
P. (2003).

43

Liu, X. S., & Meyer, J.
P. (2005).

Student discipline problems were a major problem for teachers’
dissatisfaction with their jobs. Minority teachers are less satisfied than their
non-minority counterparts.

SASS/TFS (1994-95) N=6279 teachers

44

Liu, X. S., & Ramsey, J.
(2008).

Teacher job satisfaction varied with gender, years of teaching and career
status. Teacher compensation needs to be addressed separately from the issue
of teacher satisfaction with work conditions because teacher satisfaction with
compensation is not highly correlated with teacher satisfaction with work
conditions.

SASS/TFS (2000-01) N=4952 teachers

45

Loeb, S., DarlingHammond, L., &
Luczak, J. (2005).

Teachers' ratings of school conditions were the strongest predictors of teacher
turnover. Working conditions add substantive predictive power to the model
and the addition of work conditions substantially reduces the effect of student
demographics.

California teacher survey data linked to
district data on salaries (N=1071 teachers
in 1,018 schools)

Principal leadership, communal school organization, an orderly environment,
and average levels of control granted to teachers influence average efficacy.
Higher levels of efficacy in Catholic than in public schools are explained by
organizational differences in schools.
The most frequent forms of collaboration cited were department meetings,
faculty meetings, special education meetings, subject area meetings, team
teaching, lesson planning, and faculty workshops. The most frequent reason
cited for not collaborating was a lack of time and a large amount of
paperwork.

37

High School and Beyond (N=8,488
teachers)

Self-collected (N=56)

Table A1 (cont’d)
No.

Author(s)/Year

Key Findings

Data Source/Sample Size

46

Locke, E. A. (1976)

This paper stresses a conceptual approach to job satisfaction and finds that
mentally challenging work, working conditions compatible with an
individual’s needs, high self-esteem and the presence of agents in the
workplace who can help the employee are among the conditions most
conducive to satisfaction.

Review of existing literature

47

Lortie, D. (1975)

Lortie notes the phenomenon of teacher isolation and finds that teachers
reject principals who exercise close supervision. 38% of teachers prefer to
have supportive principals.

Content analysis of interviews of ES and
HS teachers. Selection of teachers was
random (N=94 teachers)

48

Ma, X., & MacMillan,
R.B. (1999).

Teachers' positive perceptions of their administrators played a significant role
in narrowing the teacher satisfaction gap among teachers with different
experience.

Teacher data from the New Brunswick
Elementary school study (N=2202 teachers)

49

Marks, H. M., & Printy,
S. M. (2003).

These findings provide significant implications for educational policy aimed
at retaining teachers in education.

School Restructuring Study (SRS); formal
interviews with principals of the schools.
N=24 schools; 144 core teachers

Michaelowa, K., &
Wittmann, E. (2007).

Measures ensuring control and incentive related working conditions for
teachers, significantly increase student achievement while reducing teacher
job satisfaction. In addition, teachers' academic qualification beyond the
"baccalaureate", while beneficial for students' learning, tends to lead to a
mismatch between teachers' expectations and professional realities and
thereby reduces teachers' job satisfaction. Class size has a considerable
negative relationship with job satisfaction.

Program of Analysis of Education Systems
(PASEC). N=100 schools in each of five
countries

51

Moore, C. M. (2012).

School environment played a statistically significant role in the
dissatisfaction of teachers. Specifically, teacher autonomy and principal
leadership decreased the odds of teacher dissatisfaction, while student and
community problems increased the odds of teacher dissatisfaction.

SASS (N=34,870 teachers and 6,800
schools)

52

Mowday, R. T., Steers,
R.M., & Porter, L.M.
(1979).

Relatively consistent relationships were found between the organizational
commitment construct and employee turnover, absenteeism and job
performance.

Self-collected (N=2563 employees in 9
divergent organizations)

53

Ostroff, C. (1992).

Results indicate that a relationship exists between satisfaction, performance,
and attitudes. Organizations with more satisfied employees tended to be more
effective than organizations with less satisfied employees.

NASSP surveys (N=364 schools in 36
states and 14,721 teachers)

50

38

Table A1 (cont’d)
No.

Author(s)/Year

Key Findings

Data Source/Sample Size

Perie, M., & Baker, D.
P. (1997).

Favorable working conditions were associated with higher satisfaction scores.
Demographic factors were not nearly as significant in explaining the different
levels of satisfaction as workplace condition factors such as administrative
support, parental involvement, and teacher control over classroom
procedures. Teacher satisfaction showed a weak relationship with salary and
benefits.

1993-94 SASS

55

Printy, S. M. (2008).

Results suggest that both principals and department chairpersons are
instrumental in shaping opportunities for teachers to learn in communities of
practice. Department chairpersons might serve to slow down the rate of
instructional change

NELS:88 (N=2,718 teachers)

56

Pogodzinski, B.,
Youngs, P., Frank, K.
A., Belman, D. (2012).

Perceptions regarding the relationships between the administrators and
teachers are important. If teacher-administrator relations are poor, novices
are less likely to stay in the school. Further, an individual teacher who
perceives that she has adequate resources would be more willing to remain
teaching.

Survey data (N=184 teachers across 99
schools)

57

Renzulli, L. A., Parrott,
H. M. & Beattie, I. R.
(2011).

Charter school teachers are more satisfied than public school teachers
because of greater autonomy. They also have higher rates of attrition..
Teachers in racially mismatched schools have lower levels of satisfaction.

1999-2000 SASS data (N=32,930 teachers
in 7,190 schools)

58

Rogers-Jenkinson, F., &
Chapman, D.W. (1990).

The quality of school working conditions and respondents' relationships with
other teachers were significantly related to teacher satisfaction for both public
and private school teachers. Leadership style mattered for private school
teachers.

Survey (N=190 public school teachers;
N=100 private school teachers).

59

Rosenholtz, S. J., &
Simpson, C. (1990).

Novice teachers’ commitment is influenced more by organization supports
while experienced teachers require more support for the core instructional
tasks. Mid-career teachers have a lower commitment to their jobs and place a
greater emphasis on autonomy in their jobs.

Self-collected (N=1213 teachers in 78 ES)

60

Sargent, T., & Hannum,
E. (2005).

The least qualified teachers are consistently assigned to the poorest kids but
these are also the most satisfied teachers in Gansu province.

Gansu Survey of Children and Families
(GSCF) (N=128 principals and 1,003
teachers)

54

39

Table A1 (cont’d)
No.

Author(s)/Year

Key Findings

Data Source/Sample Size

61

Scafidi, B., Sjoquist, D.
L., & Stinebrickner, T.R.
(2005).

Teachers are more likely to leave schools with lower income and higher
proportions of minorities.

Administrative data on Georgia schools
(N=11070 elementary teachers)

Schneider, M. (2003).

Teachers reported dissatisfaction and daily problems with their school
facilities. Over forty percent considered their classrooms to be inadequate to
their needs. A third of teachers had no teacher workspace. Teachers believed
that school conditions affected their career decisions.

Survey data in Chicago and Washington
D.C. (N=N/A).

The authors argue that the nature of purpose and autonomy within a teacher
team can influence the social distribution of leadership.

Field notes /video recordings (N=2 HS
teacher teams)

The findings of this study suggest that teachers' commitment to their
profession could be enhanced by increasing collegiality among them.

Self-collected survey (N=17 schools and
364 teachers)

SASS 2003-04 (N=7,670 principals and
40,770 teachers)

62

63
64

Scribner, J. P., Sawyer,
R. K., Watson, S. T., &
Myers, V. L. (2007).
Shah, M., & Abualrob,
M.A. (2012).

65

Shen, J., Leslie, J. M.,
Spybrook, J. K., Ma, X.
(2012).

17% of the variance in teacher satisfaction is between schools implying that
schools can make a difference in teacher job satisfaction. The authors found
that school processes—particularly career and working conditions, staff
collegiality, administrative support, and to a lesser extent, positive student
behavior and teacher empowerment—are positively associated with teacher
job satisfaction

66

Skaalvik, E. M., &
Skaalvik, S. (2011).

Teachers’ feelings of belonging and emotional exhaustion were predictive of
job satisfaction, while emotional exhaustion and job satisfaction were
predictive of motivation to leave the teaching profession.

Self-collected data (N=2569 ES and MS
teachers)

Somech, A. (2005).

Findings illustrate the limitation of using a directive approach in facilitating
team innovation. However, directive leadership was related to school staff
team performance directly and also indirectly through organizational
commitment. On the other hand, participative leadership was linked to
innovation indirectly through empowerment

Self-collected (N=140 elementary schools
and 712 teachers)

67

40

Table A1 (cont’d)
No.

Author(s)/Year

Key Findings

Data Source/Sample Size

68

Spillane, J. P.,
Halverson, R., &
Diamond, J. B. (2004).

They argue that leadership activity is constituted in the interaction of multiple
leaders and followers around particular leadership tasks. Interdependency
among leaders, followers, and the situation of leadership activity is critical.

NA

69

Stockard, J., & Lehman,
M. B. (2004).

More satisfied teachers tended to be older, to work in more orderly schools,
to have higher salaries, to receive more support from colleagues and parents,
to have more control and influence over their work, and to perceive their
principal as more effective. None of the other background, demographic, or
assignment variables was significantly associated with teacher satisfaction.

1993-95 SASS; TFS; N=379 teachers
responding to both surveys

70

Taylor, D., &
Tashakkori, A. (1995).

Aspects of school climate emerged as stronger predictors of job satisfaction
than did the elements of decision participation. Strongest among these school
climate dimensions were the lack of obstacles to teaching and principal
leadership.

1990 follow-up NELS (N=1296 schools
and 9987 teachers)

71

Tschannen-Moran, M.
(2001).

This study stresses the importance of trust in nurturing collaborative
relationships. Schools where there was a high level of trust could be
predicted to be schools where there would be a high level of collaboration.

Self-collected (N=50 teachers)

72

Wahlstrom, K.L., &
Louis, K.S. (2008).

Teachers professional community (reflective dialogue, de-privatized practice,
shared norms) have a robust effect on contemporary practice and these effects
are only moderately changed with the introduction of individual
characteristics.

Teacher survey developed for the national
research project "Learning from
Leadership" Wallace Foundation (N=4165
teachers in 138 schools)

73

Yasumoto, J. Y.,
Uekawa, K., & Bidwell,
C. E. (2001).

When the math and science departments of the LSAY HS formed collegial
foci, the effects of instructional practice on student achievement growth were
intensified.

Data collected in collaboration with the
Longitudinal Study of American Youth
(LSAY). N=52 schools

74

Ylimaki, R., &
Jacobson, S. (2013).

Successful principals distributed leadership in ways that cultivated OL,
improved IL and supported CRP. They also believed in the importance of
social and professional support from cohort models and similar networks.

Interviews with principals, teachers, staff,
parents, students

75

Zembylas, M. &
Papanastasiou, E.
(2004).

People who did not have pressure from their family to follow this career, and
who always wanted to become teachers were more likely to be satisfied with
being teachers. Educators in higher positions (vice-principals or principals)
tend to have higher levels of satisfaction than teachers.

Adapted version of the questionnaire
developed by the “Teacher 2000 Project”
(N=461 teachers/administrators)

41

Table A1 (cont’d)
No.

Author(s)/Year

Key Findings

Data Source/Sample Size

76

Zembylas, M. &
Papanastasiou, E.
(2005).

Status, promotion, decision-making, and personal growth all significantly
affect the sense of empowerment felt by Cypriot teachers. Teachers who were
very satisfied with their opportunities for promotion were found to have
lower empowerment compared to the teachers who were not as satisfied with
their opportunities for promotion.

Adapted version of the questionnaire
developed by the “Teacher 2000 Project”
(N=449 teachers)

77

Zembylas, M. &
Papanastasiou, E.
(2006).

Teacher satisfaction was associated with student interaction and collegial
opportunity. Students’ lack of interest and bad behavior, the centralized
educational system, and the lack of professional autonomy were associated
with teacher dissatisfaction.

52 semi-structured interviews with
teachers/administrators

42

BIBLIOGRAPHY

43

BIBLIOGRAPHY
Abd-el-Fattah, S. M. (2010). Longitudinal effects of pay increase on teachers’ job satisfaction:
A motivational perspective. The Journal of International Social Research, 3(10), 11-21.
Abu-Saad, I., & Hendrix, V. L. (1995). Organizational climate and teachers’ job satisfaction in a
multi-cultural milieu: The case of the Bedouin Arab schools in Israel. International Journal
of Educational Development, 15(2), 141-153.
Allensworth, E., Ponisciak, S., & Mazzeo, C. (2009). The schools teachers leave: Teacher
mobility in Chicago public schools. Consortium on Chicago School Research Report,
University of Chicago.
Barnes, G., Crowe, E., & Schaefer, B. (2007). The cost of teacher turnover in five school
districts: A pilot study. National Commission on Teaching and America’s Future (NCTAF).
Bogler, R. (2001). The influence of leadership style on teacher job satisfaction. Education
Administration Quarterly, 37(5), 662-683.
Bogler, R. (2002). Two profiles of schoolteachers: a discriminant analysis of job satisfaction.
Teaching and Teacher Education, 18, 665–673.
Borman, G.D., & Dowling, N.M. (2008). Teacher attrition and retention: A meta-analytic and
narrative review of the research. Review of Educational Research 78(3), 367-409.
Boyd, D., Grossman, P., Ing, M., Lankford, H., Loeb, S. & Wyckoff, J. (2011). The influence of
school administrators on teacher retention decisions. American Education Research Journal,
48(2), 303-333. DOI: 10.3102/0002831210380788.
Buckley, J., Schneider, M., & Shang, Y. (2005). Fix it and they might stay: School facility
quality and teacher retention in Washington, DC. Teachers College Record, 107, 1107-1123.
Chen, J. (2010). Chinese middle school teacher job satisfaction and its relationships with teacher
moving. Asia Pacific Educational Review, 11, 263-272.
Cohen, J., Mccabe, E.M., Michelli, N.M., Pickeral, T. (2009). School climate: Research, policy,
practice, and teacher education. Teachers College Record, 111(1), 180-213.
Culver, S.M., Wolfle, L.M., & Cross, L.H. (1990). Testing a Model of Teacher Satisfaction for
Blacks and Whites. American Educational Research Journal, 27(2), 323-349.
Dinham, S. and Scott, C. (2002), “The international Teacher 2000 Project: an international study
of teacher and school executive satisfaction, motivation and health in Australia, England,
USA, Malta and New Zealand”, paper presented at the Challenging Futures Conference,
University of New England, Armidale.
44

Dinham, S., & Scott, C. (2000). Moving into the third, outer domain of teacher satisfaction.
Journal of Educational Administration, 38 (4), 379 – 396.
Fairchild, S., Tobias, R., Corcoran, S., Djukic, M.,Kovner, C., & Noguera, P. (2012). White and
Black teachers’ job satisfaction: Does relational demography matter? Urban Education,
47(1), 170-197. DOI: 10. 1177/0042085911429582.
Firestone, W. A., & Pennell, J.R. (1993). Teacher commitment, working conditions and
differential incentive policies. Review of Educational Research, 63(4), 489-525.
Grissmer, D. W., & Kirby, S. (1992). Patterns of attrition among Indiana teachers. Santa Monica,
CA: RAND.
Guarino, C.M., Santibanez, L., & Daley, G.A. (2006). Teacher recruitment and retention: A
review of the recent empirical literature. Review of Education Research, 76 (2), 173–208,
DOI: 10.3102/00346543076002173.
Grissom, J.A. (2011). Can good principals keep teachers in disadvantaged schools? Linking
principal effectiveness to teacher satisfaction and turnover in hard-to-staff environments.
Teachers College Record, 113(11), 2552-2585.
Hackman, J.R., & Oldham, G.R. (1976). Motivation through the design of work: Test of a
theory. Organizational Behavior and Human Performance, 16, 250-279.
Hanushek, E.A., Kain, J. F., & Rivkin, S. G. (2004). The revolving door. Education Next,
Winter 2004, 77-82.
Hanushek, E. A., & Rivkin, S. G. (2007). Pay, working conditions, and teacher quality. Future of
Children, 17(1), 69–86.
Horng, E.L. (2009). Teacher tradeoffs: Disentangling teachers’ preferences for working
conditions and student demographics. American Educational Research Journal, 46(3), 690717. DOI: 10.3102/0002831208329599.
Hulpia, H., Devos, D., & Rosseel, Y. (2009). The relationship between the perception of
distributed leadership in secondary schools and teachers' and teacher leaders' job satisfaction
and organizational commitment. School Effectiveness and School Improvement, 20(3), 291317.
Ingersoll, R.M., & May, H. (2012). The magnitude, destinations and determinants of math and
science teacher turnover. Educational Evaluation and Policy Analysis, published online 14
August, 2012. DOI: 10.3102/0162373712454326.
Ingersoll, R.M. and May, H. (2011). Recruitment, retention and the minority teacher shortage.
Consortium for Policy Research in Education. CPRE Research Report #RR-69. URL:
http://repository.upenn.edu/gse_pubs/226.
45

Ingersoll, R.M. (2001). Teacher turnover and teacher shortages: An organizational analysis.
Amercian Educational Research Journal, 38(3), 499-534.
Johnson, S. M., Kraft, M. A., & Papay, J.P. (2011). How context matters in high needs schools:
The effects of teachers’ working conditions on their professional satisfaction and their
students’ achievement. Project on the Next Generation of Teachers, Harvard Graduate
School of Education.
Johnson, S. M. (2006). The workplace matters: Teacher quality, retention and effectiveness.
Harvard Graduate School of Education, July 2006.
Johnson, S. M., & Birkeland, S. E. (2003). Pursuing a ‘sense of success’: New teachers explain
their career decisions. American Educational Research Journal, 40(3), 581-617.
Judge, T.A., Parkers, S., Colbert, A., Heller, D., & Ilies, R. (2001). Job satisfaction: A crosscultural review in Anderson, N. (ed.), Handbook of Industrial and Organizational
Psychology, SAGE Publications.
Kelly, S. (2004). An event-history analysis of teacher attrition: Salary, teacher tracking and
socially disadvantaged schools. The Journal of Experimental Education, 72(3), 195-220.
Kennedy, M. (2005). Inside teaching: How classroom life undermines reform. Harvard
University Press: Cambridge, Massachusetts.
Kloep, M., & Tarifa, F. (1994). Working conditions, work style, and job satisfaction among
Albanian teachers. International Review of Education, 40 (2), 159-172.
Ladd, H. (2011). Teachers’ perceptions of their work environments: How predictive of planned
and actual teacher movement? Educational Evaluation and Policy Analysis, 33(2), 235-261.
DOI: 10.3102/0162373711398128.
Lee, V.E., Dedrick, R. F., & Smith, J.B. (1991). The effect of the social organization of schools
on teachers' efficacy and satisfaction. Sociology of Education, 64(3), 190-208.
Liu, X. S., & Meyer, J. P. (2005). Teachers’ perceptions of their jobs: A multilevel analysis
of the teacher follow-up survey for 1994-95. Teachers College Record, 107, 985–1003.
Liu, X. S., & Ramsey, J. (2008). Teachers’ job satisfaction: Analysis of the teacher follow-up
survey in the United States for 2000-2001. Teaching & Teacher Education, 24, 1173-1184.
Locke, E. A. (1976). The nature and causes of job satisfaction. In M.D. Dunnette (Ed.)
Handbook of Industrial and Organizational Psychology, 1297-1349. New York: Wiley.
Loeb, S., Darling-Hammond, L., & Luczak, J. (2005). How teaching conditions predict teacher
turnover in California schools. Peabody Journal of Education, 80(3), 44-70.

46

Lortie, D. C. 1975. Schoolteacher. Chicago: University of Chicago Press.
Ma, X., & MacMillan, R.B. (1999). Influences of workplace conditions on teachers' job
satisfaction. The Journal of Educational Research, 93(1), 39-47.
Michaelowa, K., & Wittmann, E. (2007). The cost, satisfaction and achievement of primary
education – Evidence from Francophone Sub-Saharan Africa. The Journal of Developing
Areas, 41(1), 51-78.
Moore, C. M. (2012). The role of the school environment in teacher dissatisfaction among U.S.
public school teachers. SAGE Open, published online 24 February 2012.
DOI: 10.1177/2158244012438888.
Mowday, R. T., Steers, R.M., & Porter, L.M. (1979). The measurement of organizational
commitment. Journal of Vocational Behavior, 14, 224-247.
Ostroff, C. (1992). The relationship between satisfaction, attitudes, and performance: An
organizational level analysis. Journal of Applied Psychology, 77(6), 963-974.
Perie, M., & Baker, D. P. (1997). Job satisfaction among America’s teachers: Effects of
workplace conditions, background characteristics, and teacher compensation (NCES 97-471).
Washington, DC: U.S. Government Printing Office.
Pogodzinski, B., Youngs, P., Frank, K. A., Belman, D. (2012). Administrative climate and
novices’ intent to remain teaching. Elementary School Journal, 113(2), 252-275.
Renzulli, L.A., Parrott, H.M., & Beattie, I.R. (2011). Racial mismatch and school type: Teacher
satisfaction and retention in charter and traditional public schools. Sociology of Education,
84(1), 23-48. DOI: 10.1177/0038040710392720.
Rodgers-Jenkinson, F., & Chapman, D. W., (1990). Job Satisfaction of Jamaican elementary
school teachers. International Review of Education, 36(3), 299-313.
Rosenholtz, S. J., & Simpson, C. (1990). Workplace conditions and the rise and fall of teachers’
commitment. Sociology of Education, 63(4), 241-257.
Sargent, T., & Hannum, E. (2005). Keeping teachers happy: Job satisfaction among primary
school teachers in rural Northwest China. Comparative Education Review, 49(2), 173-204.
Scafidi, B, Sjoquist, D.L., & Stinebrickner, T.R. (2006). Race, poverty and teacher mobility.
Andrew Young School of Policy Studies, research paper series. Working paper 06-51,
August, 2005.
Schneider, M. (2003). Linking school facility conditions to teacher satisfaction and success.
National Clearinghouse for Educational Facilities: Washington DC.

47

Shann, M.H. (1998). Professional commitment and satisfaction among teachers in urban middle
schools. The Journal of Educational Research, 92(2), 67-73.
Shen, J., Leslie, J. M., Spybrook, J. K., Ma, X. (2012). Are principal background and school
processes related to teacher job satisfaction? A multi-level study using Schools and Staffing
Survey 2003-04. American Educational Research Journal, 49 (2), 200-230.
Skaalvik, E. M., & Skaalvik, S. (2011). Teacher job satisfaction and motivation to leave the
teaching profession: Relations with school context, feeling of belonging and emotional
exhaustion. Teaching and Teacher Education, 27, 1027-1038.
Stockard, J., & Lehman, M. B. (2004). Influences on the satisfaction and retention of 1st-year
teachers: The importance of effective school management. Educational Administration
Quarterly, 40(5), 742-771. DOI: 10.1177/0013161X04268844.
Taylor, D., & Tashakkori, A. (1995). Decision participation and school climate as predictors of
job satisfaction and teachers’ sense of efficacy. Journal of Experimental Education, 63, 217–
230.
Zembylas, M. & Papanastasiou, E. (2004),"Job satisfaction among school teachers in Cyprus”,
Journal of Educational Administration, 42(3), 357 – 374.
Zembylas, M. & Papanastasiou, E. (2005): Modeling teacher empowerment: The role of job
satisfaction. Educational Research and Evaluation, 11(5), 433-459.
Zembylas, M. & Papanastasiou, E. (2006): Sources of teacher job satisfaction and dissatisfaction
in Cyprus. Compare, 36(2), 229-247.

48

CHAPTER 2
DOES LEADERSHIP STYLE MATTER? A MULTI-COUNTRY ANALYSIS OF
PRINCIPAL LEADERSHIP AND TEACHERS’ JOB SATISFACTION
2.1 Introduction
This paper explores how particular aspects of the workplace, specifically principals’
leadership styles are associated with teacher satisfaction in a multi-country context inclusive of
the United States and six Asian countries. Additionally, the data I use (TIMSS 2007) offer the
unique opportunity to examine how teacher and principal perceptions of teacher satisfaction
differ by country. Thus, both objective (principal reported) and subjective (teacher reported)
measures of teacher satisfaction are available to me. Although some aspects of teacher policy
such as teacher quality and teacher evaluation have received global attention, teacher satisfaction
particularly in the Asian context has been much less discussed internationally.
The issue of teacher satisfaction merits further study because satisfaction in the literature
(U.S.A) is related to productivity in the workplace and has direct implications not only for
teacher turnover, but also for teacher commitment, and student achievement (Barnes, Crowe &
Schaefer, 2007; Boyd, Grossman, Ing, Lankford, Loeb, Wyckoff, 2011; Guarino, Santibanez &
Daley, 2006; Ingersoll, 2001; Ostroff, 1992; Stockard & Lehman, 2004; Perie & Baker).
Previous research also links job satisfaction with job commitment (Firestone & Pennell, 1993;
Mowday, Steers & Porter, 1979).
A focus on Asia in this analysis is motivated by the fact that there is a paucity of
empirical research on teacher satisfaction in a multi-country context more generally and in a
multi-country context within Asia specifically (See Table 2 in Chapter 1 of this dissertation).
Further, the studies that are Asia-focused are mainly single-country studies within Asia. This
49

paucity of research that is Asia-focused is also corroborated by Hallinger (2011) who points out
that less than nine percent of the papers published in the four top journals in Educational
Administration in the year 2008-09 were on Asia.
Given that Asian economies are among the fastest growing economies in the world and
that the importance of Asia as a region vis-à-vis the U.S. will most likely continue to grow over
the next decade, a closer examination of leadership in Asian work environments as compared to
the U.S. may help to inform school policies aimed at improving teacher-principal interactions
within the U.S.
2.2. Literature Review
2.2.1

Leadership & Satisfaction

This section reviews relevant literature on instructional and transformational leadership
with respect to teacher satisfaction. The principal is often considered a central figure in any
discussion of the school as a workplace. This may be because leadership behavior can create
distinct working environments within schools, which in turn impact teacher efficacy and sense of
well-being. In fact, scholars such as Stockard and Lehman (2004) and Grissom (2011) found
that leadership is one of the most salient dimensions of work conditions with respect to both
teacher satisfaction and teacher turnover. Several other studies support similar results within the
U.S. (Branch et al, 2013; Grissom & Loeb, 2011; Hulpia, Devos & Rosseel, 2009; Lee, Dedrick
& Smith, 1991; Moore, 2012; Pogodzinski et al, 2012; Printy, 2008; Rosenholtz, 1990; Scribner
et al, 2007; Taylor & Tashakkori, 1995; Wahlstrom & Louis, 2008). Further support for
leadership efficacy comes from Allensworth et al (2009) who found that teachers preferred
schools where they felt they had supportive principals who helped them do their job better.
Similarly, Brown and Wynn (2009) found that principals effective at improving teacher

50

satisfaction and reducing turnover provide the conditions and resources necessary to support
teachers.
Although a variety of leadership models have been employed in leadership research over
the past twenty-five years, two conceptual models – transformational leadership and instructional
leadership have come to dominate the field of educational leadership. These models focus
explicitly on the manner in which leadership exercised by school administrators and teachers
brings about improved educational outcomes (Hallinger, 2003).
2.2.2

Instructional Leadership

The term ‘instructional leadership’ originated with the ‘effective schools movement’ in
the 1970’s when researchers started to compare characteristics of schools that were “effective”
(i.e. successful at educating all students regardless of their family background) to schools that
were not so effective. Key among the list of characteristics that emerged from this movement
with respect to effective schools was the principal’s role as an instructional leader. The aims of
instructional leadership are typically tied to the core work of schools – teaching and learning.
Instructional leaders are viewed in the literature as culture builders who create an academic press
focusing on student academic outcomes.
The most frequently used conceptualization of instructional leadership was developed by
Hallinger (2000). This model proposes three key dimensions of the instructional leadership
construct. These include defining the school’s mission (framing and communicating goals),
managing the instructional program (supervising instruction, coordinating curriculum, and
monitoring student progress) and promoting a positive learning climate in schools (protecting
instructional time, professional development, a visible presence, promoting high expectations
and providing incentives for teachers and students).

51

A vast body of extant research on instructional leadership indicates that school principals
contribute to student achievement indirectly (Heck et al, 1990; Lee, Walker & Chui, 2012;
Leithwood & Day, 2008; Robinson, Lloyd & Rowe, 2008). Further, school context attributes
such as socioeconomic status and school size do have an effect on the type of instructional
leadership exercised by principals (Hallinger & Heck, 1996; Hallinger & Heck, 2002).
Relatively few studies find a direct relationship between the principal’s hands-on supervision of
classroom instruction, teacher effectiveness and student achievement (see Hallinger & Heck,
1996 for one such study). Recent research on instructional leadership has begun to focus on the
interaction between leaders, between leaders and their followers and the particular contexts they
work in (Neumerski, 2013).
2.2.3

Transformational Leadership

While the persisting influence of the instructional leadership role of principals must be
acknowledged, principals undeniably play multi-faceted roles within schools that extend beyond
direct supervision of curriculum and instruction. In the 1990’s in response to a broad
dissatisfaction with the instructional leadership model, which many believed focused too
narrowly on the principal as the center of expertise and authority a new leadership paradigm
emerged known as transformational leadership. Transformational leaders focus on fostering the
organization’s capacity to innovate and on supporting the development of changes to teaching
and learning practices. In this sense, transformational leaders seek to generate second-order
effects (Hallinger, 2003) or to create conditions under which others become committed to school
improvement without direction from above. Transformational leadership may also be considered
distributive to the extent that it prioritizes developing a shared vision and a shared commitment
to school change.

52

Transformational leadership was first elucidated as a theory in the general management
literature. In the early nineties, Bass and Avolio developed and popularized the
conceptualization of the 4 I’s – Idealized influence, Inspirational motivation, Intellectual
stimulation and Individualized consideration. Leithwood and his colleagues subsequently
substantively adapted Bass’s transformational leadership construct to the field of education.
Notably, the transformational leadership model does not assume that the principal alone will
provide the leadership that creates conditions suitable for school change. On the contrary, the
model allows for leadership to be shared with teachers. By prioritizing the needs of individual
staff members, the model displays a more bottom-up approach to influencing people within the
organization. Considerable research has been conducted over the last two decades using the
transformational leadership model (Bogler, 2001; Hallinger & Heck, 1996; Leithwood & Jantzi,
2000; Marks & Printy, 2003; Silins, 1994). Key understandings that emerged from this research
include the importance of developing human resources within the organization. Thus,
transformational leadership emphasizes promoting change in people rather than change in
specific instructional practices. This body of research points to the strong direct effect of
transformational leadership on school and classroom conditions. Some studies also put forth the
view that transformational leadership has an impact on teacher perceptions of school conditions
and their commitment to change (Bogler, 2001).
2.2.4

International Literature on Leadership

Internationally, there is a broad set of leadership studies in countries like Hong Kong,
China, and Taiwan (Cheng & Wong, 1996; Cheong, 2000; Lee, Pan & Chen, 2011; Lee, Walker
& Chui, 2012; Walker, Hu & Qian, 2012) within the geographic regions of East and South-East
Asia. Interest in these countries is increasingly driven by the need to gain a deeper

53

understanding of what type of leadership works there and also how it works. Most recently,
Walker, Hu and Qian (2012) reviewed the literature on Chinese principalship in both English and
Chinese between 1998 and 2008. In doing so, they offer a number of insights into the role of
Chinese principals. The core patterns they identified from their review include imported
frameworks, indigenous investigations and contextual influences. Their paper highlights that
Chinese principals display transformational practices very similar to those described in the
imported frameworks such as inspiration, charismatic leadership and intellectual stimulation.
Similarly, Pan and Chen’s study (2011) highlights how decentralization reform in Taiwan in the
1990’s has created the need for a more participatory education system that emphasizes
instructional, curriculum and transformative styles of leadership as qualities necessary for
principals to fulfil educational reform mandates. Lee, Walker and Chui’s (2012) study in Hong
Kong found that instructional leadership that focuses on managing instruction boosts the positive
effect of school attachment on student learning. However, the direct supervision of instruction
undermined student achievement through school attachment.
Additionally, recent work by Hallinger(2011) and Hallinger and Huber(2012) seeks to
expand the empirical knowledge base that currently exists in the East Asia region and also
explores policy trends for school leadership in East Asia. These studies show that much work
still remains to be done in the region as a broad consensus regarding the specifics of leadership
that matter based on empirical research is yet to emerge.
In spite of this vast body of literature on school leadership both within the U.S.A. and
internationally, leadership studies that specifically focus on teacher satisfaction in the
international context are concentrated in a handful of countries like Israel (Abu-Saad & Hendrix,
1995; Bogler, 2001), Singapore (Koh et al, 1995); Flanders (Hulpia et al, 2009) and Australia

54

(Dinham, 2005; Dinham 2002). Additionally, these studies mostly examine one particular
leadership style in a single country analysis. Missing from this body of work is a comprehensive
multi-country analysis of the leadership-satisfaction association that will further explicate and
clarify this relationship in a wider Asian context.
2.3 Conceptual & Theoretical Framework
In this study, I seek to understand how leadership style in schools - transactional,
transformational and instructional is associated with teacher satisfaction and whether the positive
relationship between transformational leadership and teacher satisfaction is observed in countries
in Asia as well as the U.S.A.
In order to more fully explore the relationship between leadership and satisfaction in a
multi-country context, I use the ‘full range of leadership’ model proposed by Bass and Avolio
(1994) as my theoretical framework in this paper. The full range of leadership model developed
by Bass and Avolio (1994) includes transactional leadership as well as transformational
leadership as two ends of a leadership continuum. Thus, transactional leadership is not
necessarily a leadership style that is considered distinct from transformational leadership.
Rather, it has been treated in the leadership literature as a form of leadership which is situated at
one extreme of transformational leadership spectrum. Transactional leadership is a term used by
Bass and Avolio to characterize leadership that emphasizes an exchange i.e. the leader specifies
what needs to be done and rewards colleagues when those requirements are adequately fulfilled.
They term this ‘contingent reward.’ A second aspect of transactional leadership is what Bass and
Avolio refer to as ‘management by exception.’
Transformational leaders however go beyond this simple transaction in the workplace to
inspire and motivate followers, act as role models, generate a vision or mission of the

55

organization and develop higher levels of ability and work ethic among colleagues.
Interestingly, Hallinger’s (2005) often used definition of instructional leadership which
encompasses three leadership dimensions - defining the school’s mission, managing the
instructional program, and promoting a positive learning climate overlaps to a degree with
transformational leadership. While acknowledging that instructional and transformational
leadership may overlap as do transactional and transformational leadership I nevertheless treat
each of these leadership styles as distinct in order to keep the empirical analysis clean and to
avoid potential issues of multicollinearity that would arise if the leadership styles were allowed
to overlap in the analytic models. To this end, instructional leadership in the analysis is
measured by a single variable - the percent of time that the principal spent on instructional
leadership in the previous year.
Figure 2 provides a conceptual road map for the analysis proposed. Teacher satisfaction8
is the binary outcome of interest. I use an extensive set of student background, teacher
demographic and school controls known to be associated with teacher satisfaction in the
literature. These controls include variables like teacher age, experience, gender, certification,
school size, class size, school infrastructure and proxies for socioeconomic status aggregated to
the school level. These controls represent the demographic context of schools within which
teachers teach. The key independent variables are in the box labeled ‘Principal Leadership.’
Figure 2 (see next page) indicates that teacher satisfaction is an antecedent of outcomes such as

8

For analysis I the dependent variable is a teacher level variable which represents teacher perception of the level of
teacher satisfaction. For analysis II, the dependent variable (teacher reported satisfaction) is aggregated at the
school level. For analysis III, the dependent variable is principal perception of teacher satisfaction.

56

DEMOGRAPHIC CONTEXT
Student
Background

Teacher
Background

School
Background

PRINCIPAL LEADERSHIP
TRANSACTIONAL LEADERSHIP

INSTRUCTIONAL LEADERSHIP
LELELELEADERSHIP
TRANSFORMATIONAL LEADERSHIP

TEACHER SATISFACTION

Student
Achievement

Teacher
Retention

Teacher
Commitment

Source: Author conceptualization of the interrelationship between satisfaction and principal leadership style.

Figure 2: A Conceptual Model of Principal Leadership and Teachers’ Job Satisfaction

57

teacher commitment, teacher turnover and student achievement. These outcomes are depicted in
gray font to emphasize that these relationships are outside the scope of this analysis. Instead, the
key emphasis is on leadership style.
The remainder of the paper is organized as follows. Section 2.4 of the paper provides
country background information to better contextualize the results of the within country
regressions on leadership and teacher satisfaction. Section 2.5 outlines the research questions of
interest. Section 2.6 of the paper describes the data. Section 2.7 outlines the models and method
used. Section 2.8 provides a summary of results. Section 2.9 provides a discussion of findings.
Section 2.10 provides a detailed contextual explanation of the results for a selected country.
Section 2.11 concludes with the limitations of this study.
2.4. Country Context
The following section provides a brief overview of the location, historical background,
and national education system within each of the countries analyzed. The purpose of this section
is to help the reader contextualize the results of the within country regressions in the sections to
follow.
The countries analyzed were selected to provide a broad representation of the different
regions within Asia in order to capitalize on the rich regional and cultural variation within Asia.
Accordingly, of the six countries selected, two are located in East Asia, three are located in
South-East Asia and one is in North Asia. There are interesting overlaps in terms of similarities
and differences between these countries. For example, almost all these countries are moving or
have officially made the move from highly centralized school systems to a more decentralized
mode of operation. On the other hand, there are distinct cultural differences among the selected
countries. Two of the countries (Hong Kong and Taiwan) are heavily influenced by

58

Confucianism, two (Indonesia, and Malaysia) are predominantly Islamic and in Mongolia a huge
percentage of the population (40 percent) does not practice any religion.
2.4.1

Hong Kong (SAR)

2.4.11 Location: East Asia
2.4.12 Historical Background: Hong Kong was designated as a Special Administrative
Region (SAR) of the People's Republic of China in 1997 based on an agreement between the UK
and China. The understanding reached with the creation of the SAR was that Hong Kong would
enjoy a high degree of autonomy in all matters with the exception of foreign and defense
affairs for the next fifty years. Ethnically, Hong Kong primarily consists of a Chinese
population. The ethnic breakup is as follows: Chinese 93.6%, Filipino 1.9%, Indonesian 1.9%,
other 2.6% (U.S. Central Intelligence Agency, 2011 census)9. The official language is
Cantonese, which is spoken by 89.5% of the population followed by English, which is spoken by
3.5% of the population (U.S. Central Intelligence Agency, 2011 census). The population of
Hong Kong is entirely urban.
2.4.13 Education System: Educational expenditure in Hong Kong constitutes 3.4% of
GDP (U.S. Central Intelligence Agency, 2013). Traditionally, decision-making in Hong Kong is
concentrated in the hands of senior management. The current system is, however, moving
towards devolution of authority and school-based management (SBM) (Hui & Cheung, 2006).
The majority of secondary schools are homogenous in terms of salary structure, professional
qualification of teachers and principals, promotion, student-teacher ratio, facilities, curriculum
and examination structure, and supervision by the Education Department (Cheong, 2000).

9

All historical background statistics for countries taken from https://www.cia.gov/library/publications/the-worldfactbook/geos/hk.html unless otherwise indicated.

59

The Hong Kong Institute of Education (HKIEd) under the aegis of the University Grants
Commission (UGC) is primarily responsible for teacher preparation. It offers a range of degree
and postgraduate programs for pre-service and in-service teachers. Principals need to have five
years of teaching experience before they can earn certification for a principalship. The school
system in Hong Kong is still primarily driven by public examinations. Consequently, school
principals are quite concerned with academic achievement and curricular completion (Cheong,
2000).
Two areas of difficulty for principals in Hong Kong are providing incentives for teachers
and encouraging decisional participation (Cheong, 2000). This is because the salary structure for
teachers in Hong Kong is fairly rigid and there are few opportunities available for promotion.
The principal therefore has little discretion to reward teachers.
2.4.2

Indonesia

2.4.21 Location: South-east Asia
2.4.22 Historical Background: Indonesia was originally a Dutch colony in the 17th
Century. Around 1949, Netherlands ceded control of Indonesia. Subsequently, after four
decades of authoritarianism, Indonesia became the world’s third most populous democracy and
the world’s largest Muslim majority nation. Ethnically, the Javanese constitute 40.6% of the
population, Sundanese 15%, Madurese 3.3%, Minangkabau 2.7%, Betawi 2.4%, Bugis 2.4%,
Banten 2%, Banjar 1.7%, other or unspecified 29.9% (U.S. Central Intelligence Agency, 2000
census). The official language in the country is Bahasa Indonesia (official, modified form of
Malay). English, Dutch, and local dialects are also spoken. Around 86.1% of the population is
Muslim. 50.7% of the total population is urban (U.S. Central Intelligence Agency, 2013). Major
urban areas include Jakarta, which is the capital, Surabaya, Bandung, Medan, and Semarang.

60

2.4.23 Education System:10Educational expenditure in Indonesia constitutes
approximately 3% of the GDP (U.S. Central Intelligence Agency, 2013). Prior to 1999, the
Indonesia education system was highly centralized. Most course content, authorization of
textbooks, teaching hours and other matters associated with public school governance were
centrally determined (Hariri, Monypenny & Prideaux, 2012). As such, teachers did not enjoy
much autonomy with respect to curriculum design and teaching methods. Since that time,
districts have become important units in education development because of the move to
empower local communities through the Education Council at the district level.
In 2001, school-based management (SBM) was introduced nationally. Since then the
managerial and financial authority for public schools has been delegated to the district level. In
2003, Indonesia began to experience a fundamental restructuring in education because of the
enactment of the National Education System Law Number 20. A significant implication of this
reform was the greater autonomy accorded to local governments and schools. Specifically,
school principals came to be regarded as change agents expected to exercise their authority to
build capacity for reform in their schools and to improve student performance. According to
Sofo, Fitzgerald and Jawas (2012), however, many Indonesian principals are somewhat uncertain
about how to use this new authority. Further, the protection and freedom necessary for teachers
to be creative and to develop their teaching competencies is still not well-entrenched although
under SBM the expectation is that the teachers will play a critical role in designing curricula that
best meets local needs and conditions. Thus, teacher empowerment as a concept is still nascent
in Indonesia.

10

Source for data for the education system in Indonesia: Hariri, Monypenny & Prideaux (2012).

61

2.4.3. Malaysia
2.4.31 Location: South-east Asia
2.4.32 Historical Background: Malaysia was a British colony in the 18th and 19th
centuries. It gained independence in 1957. Politically, the country is a parliamentary
democracy. The ethnic breakup in Malaysia is as follows: Malay 50.4%, Chinese 23.7%,
indigenous 11%, Indian 7.1%, others 7.8% (2004). The official language in the country is
Bahasa Malaysia. English, Chinese (Cantonese, Mandarin, Hokkien, Hakka, Hainan, Foochow),
Tamil, Telugu, Malayalam, Punjabi, and Thai are also spoken. Around 60.4% of the population
is Muslim. 72.8% of total population (U.S. Central Intelligence Agency, 2013) is urban. Major
urban areas include Kuala Lumpur (which is also the capital), Klang, and Johor Bahru.
2.4.33 Education System11: Education Expenditure in Malaysia is higher than in Hong
Kong and Indonesia. It constitutes 5.1% of GDP (U.S. Central Intelligence Agency, 2013) and
reflects the importance attached to education by the national government. The main objectives of
education in Malaysia are national integration and national development. There is a strong
perception that education is critical for human resource development and to meet the needs of the
growing economy. The role of education in instilling positive values is also considered very
important. The internationalization of Malaysian thinking about teaching and learning is evident
by the introduction of several innovations such as collaborative/team teaching, peer evaluation,
and self-reflection (Schwille et al, 2013).
Teacher education takes place at both public and private universities as well as teacher
education institutes. All future teachers are required to pass comprehensive examinations (oral
and written), the Malaysia Teacher Education and the Malaysian Educators Selection Inventory.
Continuous assessment is also required within courses (Schwille et al, 2013). The Teacher
11

Source for information on the Education system in Malaysia: Schwille et al (2013).

62

Education Department of the MOE is the body that sets curriculum for the teacher education
institutes. The MOE also determines the number of teaching posts available by subject-matter
for each year, based on the demand for teachers in those disciplines in schools nationwide.
Public school teachers are initially appointed as probationary education officers by the Education
Service Commission. After a probationary period of three years the appointment is renewed
based on the teacher’s annual performance report prepared by the principal. Once confirmed,
teachers are treated as permanent staff until they retire at the age of 58 (Schwille et al, 2013).
The teaching profession in Malaysia has come to be regarded more favorably of late for
two reasons. One is the economic slowdown and the second reason is that teaching in public
schools is considered a government job complete with all the perks that other government
servants enjoy, including job-security. In terms of social status, teachers are now perceived to be
a professional group, like doctors, who are highly respected by society (Schwille et al, 2013).
2.4.4. Mongolia
2.4.41 Location: North Asia
2.4.42 Historical Background: Mongolia was ruled by the Chinese in the late 17th
century. Mongolia won its independence in 1921 when a communist regime was instated. The
Mongolian People's Revolutionary Party (MPRP) won elections in 1990 and 1992, but was
defeated by the Democratic Union Coalition (DUC) in the 1996 parliamentary election. A
coalition of four political parties led by the Democratic Party gained control of the Parliament in
the 2012 election. The ethnic breakdown of the population is as follows: Mongol (mostly
Khalkha) 94.9%, Turkic (mostly Kazakh) 5%, other (including Chinese and Russian) 0.1%
(2000). The official language spoken by 90% of the population is Khalkha Mongol. In addition,
there is some Turkic and Russian. Around 50% of the population practices Buddhism, 6%

63

practice Christianity, 4% practice Islam and around 40% do not practice any religion. Around
68% of the total population (U.S. Central Intelligence Agency, 2013) is urban. Ulaanbatar,
which is also the capital, is a major urban area.
2.4.43 Education System: Education Expenditure in Mongolia is around 5.5% of GDP
(U.S. Central Intelligence Agency, 2013). Since the 1990’s the Ministry of Education in
Mongolia has periodically changed from centralization to decentralized policy and back to
recentralization (Steiner-Khamsi & Stolpe, 2004). According to Steiner-Khamsi and Stolpe
(2004), borrowing education reform models from elsewhere is widely prevalent in Mongolia
(Steiner-Khamsi & Stolpe, 2004).
The central education authority in Mongolia is the Ministry of Education, Culture, and
Science. The Ministry is responsible for supervising virtually all publicly funded education. It
provides guidance and financial assistance for the operation of local public and private
educational institutions. The Ministry is headed by a Minister who is part of the Prime
Minister’s Cabinet. Provincial governments are responsible for implementing the nation-wide
education policy at the provincial and district levels.
Teacher education is provided at the State Pedagogical University, higher schools and
teacher training colleges. Primary and Secondary education teachers are recruited by the
principals of the schools where they wish to teach. Teacher salaries are determined by the salary
scheme for governmental service servants and are approved by the government. In addition to
their base salary, teachers get bonuses based on performance. Elementary school teachers are
typically female.
2.4.5. Taiwan
2.4.51 Location: East Asia

64

2.4.52 Historical Background: In 1946, Taiwan was ruled by a government controlled
under a single-party, Kuomintang of China (translated as the Chinese Nationalist Party). This
government was called the KMT government. By the 1950s, the ruling authorities gradually
began to democratize and in the year 2000, Taiwan underwent its first peaceful transfer of power
from the Nationalist to the Democratic Progressive Party. Ethnically, 84% of the population is
Taiwanese (including Hakka), 14% is mainland Chinese and 2% is indigenous. The official
language of Taiwan is Mandarin Chinese. The bulk of the population practices a mixture of
Buddhism and Taoism. The remainder of the population (4.5%) is Christian (U.S. Central
Intelligence Agency, 2013).
2.4.53 Education System: The teacher education system in Taiwan is highly
competitive and driven by national policy. As a result, teacher candidates face rigorous
evaluation throughout the teacher preparation process. Teacher training requirements are fairly
rigorous and culminate in a national test which is part of the quality control of pre-service
teacher education. Until recently, the government exercised fairly tight control over institutions
of teacher education as well as the deployment of novice teachers (Schwille et al, 2013). In the
early 1990’s, rigid control over teacher education was relaxed. Teachers now have to compete
for specific vacancies. In short, Taiwan has now moved toward a “position-based as opposed to
career-based teacher employment.” (Schwille et al, 2013).
Teaching in Taiwan is a very well-respected profession “infused with dignity and
authority.” (Schwille et al, 2013). Teachers in Taiwan thus, typically enjoy favorable work
conditions and incentives. All remuneration for public school teachers is government funded and
ensured. The high status accorded to teachers in Taiwan may be attributed to the prestige and
status associated with teaching in traditional Chinese culture. The role of teachers in Taiwan

65

goes beyond merely imparting subject matter knowledge. Teachers are also expected to support
the personal and social development of students and to foster good behavior. Thus, teachers play
a role similar to that of parents in terms of providing personal guidance and acting as role models
for the children.
2.4.6

Thailand

2.4.61 Location: South-east Asia
2.4.62 Historical Background: Thailand is the only Southeast Asian country never to
have been colonized by a European power. In 1932, a revolution led to the establishment of a
constitutional monarchy in Thailand. Since 2005, Thailand has experienced severe political
turmoil including a military coup in 2006 that ousted then Prime Minister Thaksin Chinnawat. In
2011, Thaksin’s youngest sister Yinglak Chinnawat and the Puea Thai Party assumed control of
the government. Ethnically, 75% of the population is Thai, 14% is Chinese and 11% are other
ethnicities. The official language is Thai with English being the secondary language of the elite.
Buddhism is the official religion in Thailand with 94.6% of the population practicing it. In
addition, 4.6% practice Islam (U.S. Central Intelligence Agency, 2013). 34.1% of total
population is urban (U.S. Central Intelligence Agency, 2013) with Bangkok being both a major
urban area and the capital of Thailand.
2.4.63 Education System: Educational expenditure in Thailand constitutes 3.8% of the
GDP (U.S. Central Intelligence Agency, 2013). Basic education in Thailand is the responsibility
of the Ministry of Education. Thai education reform was precipitated by the Asian economic
crisis in 1997. This crisis led to a drafting of a new Thai constitution which called for education
reform and decentralization of power. The rationale for the 1999 reform stemmed from the
desire to improve the country’s competitiveness vis-a-vis other Asian countries such as China,

66

India, Vietnam, Malaysia, and Singapore. The Thai system of centralized governance and
management was deemed to be poor. Economically, an important reason for reform was the
high level of income inequality, with a Gini Coefficient of 5.21 (Schwille et al, 2013).
Teaching is not as attractive as other occupations in Thailand for highly educated persons.
This is reflected in the fact that academically qualified students with high achievement are not
interested in becoming teachers. Students prefer fields such as business, finance, engineering and
medical science on account of the higher income and job availability associated with these fields.
In addition, teacher bonuses and incentives are not well-instituted. The majority of teachers
receive their customary one-tier salary increase, irrespective of their teaching performance.
The Teachers’ Council of Thailand is responsible for accrediting degrees and certificates.
Universities with a faculty of education are responsible for the preparation of future teachers for
both primary and secondary schools. In contrast to the U.S.A, there is no differentiation in the
preparation of teachers for the lower grades and secondary grades up to grade 12 (Schwille et al,
2013). Prospective teachers need to complete a five-year Bachelor’s degree in education
majoring in mathematics or complete four years of a Bachelor of Science with one year for a
graduate diploma in the teaching profession. After gaining these degrees, they can teach
mathematics in schools from Grades 1 to 12.
2.4.7. U.S.A.
2.4.71 Location: North America
2.4.72 Historical Background: In 1776, Britain's American colonies were recognized as
the new nation of the United States of America. During the 19th and 20th centuries, thirty-seven
new states were added to the original thirteen as the nation expanded. Since the end of World
War II, the economy has achieved relatively steady growth, low unemployment and inflation,

67

and rapid advances in technology. Ethnically, the USA is 79.96% white, 12.85% black, 4.43%
Asian, 0.97% Amerindian and Alaska native, and 0.18% native Hawaiian and other Pacific
islander(U.S. Central Intelligence Agency, 2013). The USA has no official national language
though the bulk of the population speaks English (82.1%) with 10.7% speaking Spanish (U.S.
Central Intelligence Agency, 2013). Major religions include: Protestant 51.3%, Roman Catholic
23.9%, Mormon 1.7%, other Christian 1.6%, Jewish 1.7%, Buddhist 0.7%, Muslim 0.6%, other
or unspecified 2.5%, unaffiliated 12.1%, none 4%. 82% of total population as of 2010 was urban
(U.S. Central Intelligence Agency, 2013). Major urban areas include New York-Newark, Los
Angeles, Miami, and Washington, D.C.
2.4.73 Education System: Education expenditure constitutes approximately 5.4% of the
GDP in the USA (U.S. Central Intelligence Agency, 2013). Teaching is regarded as a “positionbased occupation” or one in which school districts compete for teachers. Teacher recruitment and
hiring practices vary widely in the USA. However, most school districts have regulated career
progression. Common criteria that are met by teaching applicants across a majority of public
school districts in the U. S. include full standard state licensure in the field (77%), graduation
from a state-approved teacher preparation program ( 66% of districts) and a passing score on a
state test of basic skills (64% of districts) (Schwille et al, 2013).
With respect to teacher training, the U.S.A has gradually shifted from local control
toward a centralization of teacher certification policy at the state level. There is still, however,
considerable variation within and across states in licensure and program accreditation
requirements for primary school and lower secondary mathematics teaching. In the USA, more
than 1,300 public and private colleges and universities as well as school districts, state agencies,
and private organizations offer teacher education for future primary and secondary teachers

68

(Schwille et al, 2013). State approval is mandatory for teacher education programs though
approval standards vary across states.
Virtually all public school districts in the U. S. (93%) have a teacher salary schedule. The
average yearly base salary for a beginning public school teacher with a bachelor’s degree is not
comparable with salaries for graduates with jobs in engineering, computer science, health, or
business. Moreover, teachers’ working conditions vary greatly. Some teachers, particularly those
in high-poverty and/or low-resource schools, face issues such as large classes, lack of
instructional materials, limited access to technology, long work hours, long commutes, and
remote locations. Previous research has shown that poor working conditions are associated with
high levels of teacher turnover.
The country discussion shows that while the countries chosen for analysis have made the
move from more centralized to less centralized systems, there are distinct cultural and
geographic differences between them which allow us to capitalize on the rich cultural and
regional variation within Asia. For instance, teaching as a profession in Taiwan is highly
respected as compared to Thailand where it is not the first choice of many graduating students.
Similarly, teacher education is highly competitive in Taiwan (where performance is closely
linked to teacher evaluations) compared to Thailand where teachers get their customary salary
increases regardless of their performance.
The following sections - Section 2.5 and 2.6 address the guiding questions for this
analysis and a discussion of the data used respectively.
2.5 Research Questions
The main research questions of interest in this study are as follows:

69

1. What aspects of a principal’s leadership style (transactional, transformational and
instructional) are associated with teacher satisfaction in a multi-country context?
i) Is the direction of the association between these leadership indices (transactional,
transformational and instructional) and teacher satisfaction similar across the six countries
in Asia and the U.S.?
ii) Are the leadership dimensions significantly associated with teacher satisfaction consistent
across all countries?
2. Does the observed relationship between principal leadership style and teachers’ job
satisfaction vary when the outcome variable (teacher satisfaction) is principal reported rather
than teacher reported?
2.6 Data
2.6.1

Sample Design & Weights

I use the Trends in International Math and Science Study (TIMSS) 2007 data for this
analysis. TIMSS 2007 are cross-sectional data collected by the International Association for the
Evaluation of Educational Achievement (IEA). I analyze teacher satisfaction for grade 8 math
and science teachers12 in six Asian countries participating in TIMSS 2007, as well as the United
States. The countries represented in the analysis are Hong Kong (SAR), Indonesia, Malaysia,
Mongolia, Taipei (Taiwan) & Thailand. For this subset of countries data on approximately 120
schools and 300 teachers are available in each country13.
The TIMSS 2007 data are particularly well-suited to cross-national analysis first, because
they collect a rich array of student, school and teacher variables across different educational,
organizational and curricular systems (Foy & Olson, 2009). Second, the IEA takes care to

12
13

Math and Science teacher satisfaction is analyzed because of data availability for these subjects.
The exact number of students, teachers, and schools varies by country.

70

ensure a high level of comparability/reliability of the instruments across countries. TIMSS uses
a two-stage sampling procedure where a probability proportional-to-size sample of schools is
selected at the first sampling stage and one or two intact classrooms (8th grade) are sampled per
school with an equal probability of selection at the second stage (Joncas, 2008a). The TIMSS
schools are thus a nationally representative sample of schools within each country though the
teachers are not nationally representative. They are instead the teachers of a nationally
representative sample of students. Consequently, the teacher weight variable specified in TIMSS
is not suitable for teacher level analysis. It is instead designed for using teacher background data
in student level analyses. Analysis I (a detailed description of all analyses is provided in section
2.7), which is at the teacher level is therefore presented without weighting the results. Several
other sampling and weighting variables are included in the TIMSS data files to address the
complex sampling structure of the data (Joncas, 2008b). I use schwgt for Analysis 2 and 3,
which is the recommended probability weight for school level analysis using TIMSS data. In
this study, the data are not combined into a single cross-national dataset. Instead, I run separate
within country regressions. This has the advantage of obviating the need to control adequately
for country context and culture within the regression (which would be necessary in a pooled
cross-national comparison) as it is based on the assumption that culture will not differ
substantively between schools within the same country.
2.6.2. Description of Dependent Variables
A summary of descriptive statistics on all the variables used in the analysis is provided in
Table 1. The dependent variable for the first research question (Analysis I – teacher level) is a
measure of teacher perception of teacher satisfaction at the teacher level. It is a binary response
variable constructed from a categorical variable such that 1=very high/high satisfaction and

71

0=medium/ low satisfaction. For the same research question I conduct another analysis at the
school level because schools in the TIMSS data are nationally representative while teachers are
not. The dependent variable for Analysis II – (School level) is a measure of teacher satisfaction
aggregated at the school level. The dependent variable for the second research question
(Analysis III – School level) is a measure of principal perception of teacher satisfaction. This is
constructed in a similar manner to the teacher reported measure of satisfaction.
2.6.3. Description of Key Independent Variables
The primary independent variables of interest in all analyses are the leadership variables
measuring transactional and instructional leadership and the factor scores measuring
transformational leadership. Leithwood and Sun (2012) in their research indicate that two
leadership practices – contingent reward and management by exception may be classified as
traditional approaches to leadership in their own right. Bass and Avolio (1993) refer to these
leadership practices as “transactional leadership.” In this analysis, a binary variable that asks the
principal whether incentives are used to recruit or retain teachers in their school (tch_incentives)
was used as a measure of the leadership practice of contingent reward. Under management by
exception, the leader monitors the performance of staff members and steps in when their
behavior deviates from expectation. In this analysis, a binary variable that measures whether
observations by the principal are used to evaluate 8th grade teachers (prinobs) reflects the
leadership practice of management by exception.
I conducted an exploratory factor analysis (with varimax rotation) to develop an index of
transformational leadership. Factors with an eigenvalue of 1 or higher were retained based on

72

Table 3: Components of Leadership Indices 14
No.

Leadership Type

Variable Name

Variable description

Transactional Leadership
1.

Management by Exception

prinobs

2.

Contingent Reward

tch_incent

1.

Setting Directions
Inspirational Motivation

Are observations by the principal/senior staff used to
evaluate 8th grade math teacher practice?
Are incentives in the form of a bonus, housing, pay, or
small classes used to retain 8th grade math teachers?

Transformational Leadership

2.

Hold High Performance
Expectations

hightch_goals
enrichment
high_tchexp
tch_currsuccess

3.

Developing People
Provide Individual
Support/Intellectual
stimulation

high_pdschimp
pd_tchg
pd_tech

4.
5.

Re-designing the
Organization
School Discipline
Building Collaborative
Structures

high_stuattend
low_cldist
peerreview
coll_concept
obscoll

6.

Community Focus

fundraise
volunteer
hwcomplete
schcomm
events

14

How would you characterize teachers’ understanding of
school curricular goals?
Does school offer enrichment math for 8th grade math
students?
How would you characterize teachers’ expectations for
student performance?
How would you characterize Teachers degree of success
in implementing school’s curriculum
In past 2 years what percent of your math teachers have
been involved in PD for math targeted at supporting the
school’s own improvement goals?
PD targeted at improving teaching skills
PD targeted at using information technology for
educational purposes
Recoded index of good school attendance
How often does classroom disturbance occur for 8 th
grade students in your school?
Is teacher peer review used to evaluate practice of
teachers?
How often do you have discussions about how to teach a
particular concept with other teachers?
How often do you have informal observations of your
classroom by another teacher?
Does your school ask parents to raise funds for school?
Does your school ask parents to volunteer for school
projects and trips?
Does your school ask parents to ensure their child
completes his/her hw?
Does your school ask parents to serve on school
committees (e.g. Select school personnel, review school
finances)
Does school ask parents to attend special events

The variables listed under transformational leadership were used in an Exploratory Factor Analysis (EFA) to
create an index of transformational leadership. All variables for EFA were selected based on Leithwood and
Sun’s (2012) conceptualization of transformational leadership.

73

Table 3 (cont’d)
Instructional Leadership
1.

BC4GAPIL

By the end of this school year approximately what
percentage of time in your role as principal will you have
spent on instructional leadership?

the Kaiser criterion. In all countries the variables loaded on to at least two dimensions of
transformational leadership using this criterion. Additionally, because for almost all the
countries (except Mongolia) the variables loading on the community focus dimension of
leadership had factor loadings of .30 or higher they were included even if the eigenvalue for this
third factor was below 1. Table 4 thus reports the factor loadings above .30 for all three
leadership factors across all seven countries. Regression factor scores were then computed for
these transformational leadership factors using STATA’s ‘predict’ command. These leadership
factors were operationalized as ‘Setting Directions,’ ‘Developing People, and ‘Community
Focus’ in keeping with Leithwood’s model of transformational leadership based on the Nature of
School Leadership survey (NSL) (See Leithwood, Aitken & Jantzi, 2001 in Leithwood & Sun,
2012).
Table 3 provides a detailed list of the variables that were used in the exploratory factor
analysis to create the transformational leadership factor scores. The dimension ‘Setting
Directions’ references the leader’s role in developing a shared vision for the school that is
inspirational for the staff. This dimension also indicates the leader’s success in communicating
optimism about achieving specific school goals. In addition, leaders expect their teachers to hold
high performance expectations for students.

74

The dimension ‘Developing People’ refers to the leader’s role in acting as a mentor,
attending to individual needs and supporting the professional development of the staff. Leaders
also encourage staff creativity and help their staff to improve their practice.
The dimension ‘Redesigning the Organization’ references the leader’s role in creating a
cohesive school culture around a common set of values that promote trust among the staff
members. To this end, leaders establish working conditions that facilitate staff collaboration for
planning and professional development. Leaders also demonstrate sensitivity to parent
aspirations by actively encouraging parents and guardians to become involved in their children’s
education at home and in school (I operationalize this aspect of redesigning the organization as
‘Community Focus’). It is worth mentioning that only the variables measuring ‘Community
Focus’ loaded on to ‘Redesign the organization’ aspect of the transformational leadership index
in the countries analyzed. Specifically, ‘Community Focus’ in this paper refers to the principal’s
role in engaging parents in school matters such as serving on school committees pertaining to
school finance and personnel, volunteering in the classroom, raising funds for the school and
asking parents to attend special school events.
Bass and Avolio (1994) also included a laissez-faire leadership form in their
transformational leadership model. Laissez-faire leaders avoid their own supervisory
responsibilities. This is thus, essentially a “nonleadership” practice which was excluded from
the analysis following the work of Leithwood and Sun (2012) who found that it was not a
dimension of transformational leadership.
Finally, a continuous variable BC4GAPIL, which asks school principals the percent of
time they spent on instructional leadership over the previous year was used as a measure of
instructional leadership. The transactional and instructional leadership indices were originally

75

conceptualized and constructed as summative scales. A reliability test (Cronbach’s Alpha) was
then carried out for each of these indices to measure the consistency of each index thus created.
However, since Cronbach’s Alpha was very low (less than 0.30) for the transactional leadership
index and less than 0.38 for instructional leadership in all the countries analyzed except Hong

Table 4: Factor Loadings for Transformational Leadership (Math & Science ) - TRTS (School)
Factor Names/Variables
Set Directions
HKG
hightch_goals 0.580
tch_currsuccess 0.763
high_tchexp 0.514
low_cldist 0.344
enrichment
high_stuattend
Eigenvalue 1.910
Proportion Variance Explained
0.474
Develop People
pd_tchg 0.695
pd_tech 0.526
high_pdschimp 0.674
peerreview
Eigenvalue 1.477
Proportion Variance Explained 0.366
Community Focus
fundraise 0.352
volunteer 0.514
events 0.564
schcomm
hwcomplete
enrichment
high_pdschimp
high_stuattend
Eigenvalue 0.713
Proportion Variance Explained 0.177
Cumulative Variance Explained 1.016
N
205
** Reporting all factor loadings above .30

Factor Loadings
IDN
0.608
0.664
1.137
0.297

MYS
0.500
0.540
0.304
0.436
0.956
0.243

MNG
0.384
0.499
0.338
0.851
0.273

TWN
0.673
0.689
0.578
1.375
0.274

THA
0.677
0.692
0.917
0.224

USA
0.506
0.727
0.686
0.331
2.013
0.472

0.798
0.525
0.769
1.880
0.492

0.821
0.604
0.734
1.855
0.471

0.619
0.539
1.585
0.508

0.811
0.803
0.597
0.301
2.585
0.516

0.628
0.731
0.689
2.112
0.517

0.769
0.641
0.655
1.335
0.313

0.481
0.493
0.688
0.180
0.969
268

0.558
0.445
0.452
0.327
1.319
0.335
1.049
289

0.329
0.424
0.494
0.742
0.238
0.780
211

0.404
0.633
0.307
0.449
0.369
0.878
0.175
0.965
248

0.510
0.442
0.589
0.556
1.119
0.274
1.015
291

0.606
0.585
0.349
1.033
0.242
1.027
346

76

Kong, only the variable BC4GAPIL (percent time spent by the principal in instructional
leadership) was ultimately used as a measure of instructional leadership in the within country
regressions. For transactional leadership two variables ‘prin_obs’ and ‘tch_incentives’ were
used as mentioned earlier.
Table 5 provides reliability statistics for the transformational leadership index (TRTS II)
at the school level. Pearson Correlations for the leadership indices are provided in Appendix A2
and A3. All significant correlations in tables A2 and A3 are below 0.25 at the five percent level.
Table 5: Cronbachs Alpha for Transformational Leadership (Math & Science) –
TRTS (School)
Leadership Style

HKG

IDN

MYS

MNG

TWN

THA

USA

Transformational Leadership
Set Directions
Develop People

0.625
0.717

0.647
0.750

0.576
0.780

0.384
0.575

0.714
0.794

0.723
0.754

0.668
0.760

Community Focus 0.509 0.451 0.498 0.407
0.514 0.644 0.556
Notes:
Transformational index= 3 factors:
Set directions (teacher understanding of schools curricular goals, high teacher expectations for student
achievement, low classroom disturbance, high student attendance, principal opinion of the degree of
success teachers had in implementing the curriculum).
Develop People (pd to improve teaching, pd for technology, pd for school improvement)
Community Focus (fundraise, events, volunteer, school committees, enrichment, hwcomplete)

2.6.4. Description of Controls
An extensive set of student, teacher and school context variables were used as controls in
the analysis. A brief description of these follows. For analysis I (teacher level) I aggregated
student background controls such as whether or not the student speaks the language of the test at
home (dummy = 1 if language of test is spoken always or almost always at home and dummy=0
otherwise) and mothers’ education level (dummy=1 if education level is below a secondary
education) at the classroom level by teacher ID. These two variables were used as proxies for

77

socio-economic status in the regressions. For analysis II and III all student background controls
and teacher demographic controls such as teacher age, experience, gender, certification and
education were aggregated at the school level instead (by school ID).
Table 6 provides a picture of teachers teaching a representative sample of students in the
TIMSS data. Average teacher age in the data ranges from 37 years to 44 years with the teachers
in the United States being on the older side (44 years on average) of the age range. Average
teacher experience across countries ranges from 10-16 years. Over 91 percent of teachers in
Mongolia, Taipei, Thailand and the U.S. have high teacher qualifications and certification.
There is a preponderance of female teachers in the U.S., Malaysia and Mongolia compared to
Hong Kong, Indonesia, Taiwan and Thailand. At the school level, I controlled for school size,
small class size and an index of available school resources for instruction. This index is one of
the available TIMSS derived variables ranging from 1 to 3 where 1=High, 2=Medium and
3=Low. A value close to 1 would therefore indicate that the school resources available for
instruction are good. Values close to 3 would indicate the opposite. A dummy variable was
created from this index where 1 indicates a high level of school resources available for
instruction. Similarly, a binary variable was created for class size such that 1=small class (1-24
students) and 0=large class (>24 students).
2.7 Method
Previous research on teacher satisfaction has typically been done at the teacher level using scales
to measure teacher satisfaction (see Bogler, 2001 for example). However, as mentioned earlier
in section 2.6.1, in TIMSS 2007 the teachers are not a nationally representative sample because
the TIMSS data are primarily intended for student level analysis and therefore do not provide
appropriate weights for conducting teacher level analyses.

78

Table 6: Descriptive Statistics for Leadership Indices, Student Background, Teacher Demographic and School Controls (TIMSS 07), School Level
HKG

Countries
Dependent Variable

IDN

MYS

MNG

TWN

THA

USA

Mean

SE

Mean

SE

Mean

SE

Mean

SE

Mean

SE

Mean

SE

Mean

SE

Teacher Satisfaction (prin)*

0.645

0.055

0.574

0.067

0.829

0.035

0.637

0.059

0.539

0.1

0.704

0.061

0.793

0.052

Teacher Satisfaction (tch)*

0.400

0.044

0.327

0.049

0.657

0.034

0.391

0.051

0.415

0.076

0.423

0.047

0.632

0.041

Principal Observations

0.973

0.017

0.974

0.02

0.98

0.012

0.971

0.024

0.61

0.103

0.864

0.048

0.952

0.029

Teacher Incentives

0.041

0.022

0.325

0.061

0.428

0.048

0.61

0.051

0.05

0.025

0.346

0.057

0.022

0.008

Leadership Indices
Transactional Leadership

Transformational Leadership
Setting Directions

-0.032

0.096

-0.113

0.104

0.062

0.067

0.091

0.074

0.124

0.105

-

-

-0.009

0.136

Developing People

-0.033

0.096

-0.087

0.124

0.001

0.081

0.043

0.094

-0.229

0.145

-0.09

0.104

0.075

0.093

Community Focus

-0.003

0.076

-

-

-

-

-

-

-0.068

0.092

0.036

0.098

0.011

0.043

20.384

0.804

23.629

1.049

25.016

1.064

22.486

1.218

24.282

1.416

23.077

1.108

22.021

1.575

Language of Test Spoken at
Home (0-1)

0.882

1.795

0.34

0.048

0.674

0.027

0.95

0.01

0.741

0.063

0.535

0.056

0.934

0.013

Mothers Education (Less than
Secondary) (0-1)

0.892

0.015

0.932

0.016

0.869

0.013

0.719

0.024

0.837

0.033

0.968

0.006

0.462

0.035

12.841

0.873

10.944

0.622

11.434

0.588

15.015

1.123

10.545

0.684

12.587

0.942

15.545

1.001

0.896

0.024

0.709

0.045

0.812

0.027

0.959

0.015

0.905

0.032

0.974

0.012

0.98

0.019

Proportion Females

0.379

0.04

0.441

0.042

0.719

0.031

0.773

0.042

0.334

0.062

0.58

0.048

0.676

0.044

Teacher Age (Yrs)

37.731

0.75

37.119

0.61

36.914

0.571

37.294

1.203

38.25

0.731

40.031

0.979

43.571

1.029

0.083

0.024

0.167

0.05

0.026

0.017

0.218

0.062

0.271

0.123

0.369

0.066

0.649

0.049

1019.32

28.16

311.79

28.63

1064.7

55.41

949.8

71.37

991.34

162.31

643.2

53.3

484.97

35.47

0.735

0.045

0.032

0.012

0.456

0.049

0.038

0.017

0.5

0.089

0.041

0.011

0.522

0.06

Urban (0-1)

0.31

0.05

0.188

0.052

0.107

0.027

0.092

0.05

0.216

0.066

0.034

0.013

0.08

0.024

N
Notes:

166

Instructional Leadership (% time)
Aggregated Student Controls

Aggregated Teacher Controls
Teacher Experience (Yrs)
Teacher Certification (0-1))

Small Class size (0-1)
School Controls
School Size
School Resources for Math (0-1)

234

273

170

202

278

318

1. All means are weighted by school wgt since the numbers are generated from Analysis II & III; SE are linearized.
2. *The DV teacher satisfaction is a binary variable where 1= v.high/high teacher satisfaction and 0=medium/low satisfaction. Tch=teacher reported; Prin=principal
reported
3. Analysis is for Math and Science teachers

79

However, the TIMSS data also contain rich contextual data on schools, teachers, and classes
useful for a study on leadership. In order to maintain consistency with the research tradition on
teacher satisfaction (at the teacher level) but to also use the TIMSS data most effectively, I
therefore, conduct three different analyses. The first is an un-weighted analysis at the teacher
level (TRTS I). The dependent variable for this analysis is teacher reported satisfaction and all
the controls are at the classroom level. The dependent variable for the second analysis remains
the same but is aggregated at the school level as are the controls (TRTS II). This analysis is
weighted and uses the school weights provided in TIMSS 2007. The third analysis differs in that
it examines principal perceptions of teacher satisfaction and explores whether a leader’s
perceptions differ from teacher perceptions regarding the relationship between leadership style
and teacher satisfaction (PRTS III).
2.7.1

Analysis I: Teacher Reported Teacher Satisfaction I (Teacher Level)

Research Question 1: What aspects of a principal’s leadership style (transactional,
transformational and instructional) are associated with teacher satisfaction in a multi-country
context? In analysis I, the dependent variable teacher reported teacher satisfaction (TRTS I) is
binary (where 1=v. high/high teacher satisfaction and 0=moderate/low satisfaction). Therefore, a
logistic model is used to explore the relationship between leadership style and teacher
satisfaction. A major advantage of logistic regression for dichotomous dependent variables is
that it constrains the predicted probabilities to the range [0-1] which overcomes an obvious
limitation of the linear probability model (Long & Freese, 2006, Menard, 2002). For each
country, four models were estimated. These within country regressions were clustered by school
ID to account for the fact that teachers are typically clustered within schools.

80

Model 1 regresses satisfaction on only the leadership indices since these are the key
independent variables of interest. Model 1 is given below.

odds(Y = 1) = eα + β1(transactional) + β2(transformational) + β3(instructional)

[1]

Where Y is a measure of teacher reported teacher satisfaction for a teacher such that y=1
indicates that teachers are highly satisfied and 0 indicates that they are not satisfied;
transactional is a vector of two variables measuring transactional leadership; transformational
is a vector of factor scores measuring the principal’s transformational leadership;

is the

coefficient on the instructional leadership variable.

Model 2 includes teacher and student controls in addition to the leadership indices. In
this model ST is a vector of student background and teacher demographic variables which
include whether the test language is spoken at home, mothers’ education if less than secondary,
teacher experience, certification and gender. Due to the high correlation between teacher age
and teacher experience (0.89 in Hong Kong for example) only teacher experience was used in the
within country regressions. Model 2 is given below.

odds(Y = 1) = eα + β1(transactional) + β2(transformational) + β3(instructional)+ β4ST

[2]

Model 3 regresses satisfaction on only school controls and leadership indices where SC
is a vector of school controls such as availability of school resources for math instruction, school
size and whether the school is urban.

81

odds(Y = 1) = eα + β1(transactional) + β2(transformational) + β3(instructional)+ β4SC

[3]

Model 4 is the final model and is a complete model that uses all available information.

odds(Y = 1) = eα + β1(transactional) + β2(transformational) + β3(instructional)+ β4ST + β4SC

[4]

Running the models with a combination of different controls serves as a sensitivity analysis
because a consistent finding across all models regardless of the combination of controls used
demonstrates the robustness of the results.
2.7.2

Analysis II: Teacher Reported Teacher Satisfaction II (School Level)

This analysis is conducted using the same models outlined above.

However, the

dependent variable (teacher reported teacher satisfaction (TRTS) and all student and teacher
controls are aggregated at the school level. The advantage of this analysis is that it extends
generalizability of inferences as compared to the teacher level analysis because in TIMSS 2007
the schools are nationally representative while the teachers are not. Further, since TIMSS 2007
data provide an appropriate probability weight for school level analysis, it also addresses issues
associated with the complex sampling structure of the data. For Analysis II, the school weight
(schwgt) provided in TIMSS 2007 is applied to all the within country regressions by using the
svyset command in STATA. This command accounts for the clustering of teachers within
schools in the data.
2.7.3 Analysis III: Principal Reported Teacher Satisfaction (School Level)
Research Question 2: Does the observed relationship between principal leadership style
and teachers’ job satisfaction vary when the outcome variable (teacher satisfaction) is principal
reported rather than teacher reported? TIMSS 2007 provide the unique opportunity to compare

82

principal and teacher perceptions of teacher satisfaction. Given that teacher satisfaction is an
important construct, it is both interesting and relevant to our understanding of the relationship
between teacher satisfaction and leadership style, to unpack how principal and teacher
perceptions differ with respect to satisfaction. In order to address research question 2 therefore, I
retain the same models as for research question 1, keeping all controls and independent variables
identical in order to facilitate comparison and maintain consistency. The key difference is that
the outcome variable (teacher satisfaction) in these regressions is principal reported rather than
teacher reported.
Model 1 regressed principal reported satisfaction on only the leadership indices since
these are the key independent variables of interest. Model 1 is given below.

odds(Y = 1) = eα + β1(transactional) + β2(transformational) + β3(instructional)

[1]

Where Y is a measure of principal reported teacher satisfaction for a teacher such that y=1
indicates that teachers are highly satisfied and 0 indicates that they are not satisfied;
transactional is a vector of variables measuring transactional leadership; transformational is a
vector of factor scores that measure the principal’s transformational leadership; instructional is
the variable indicative of instructional leadership.

Model 2 included aggregated teacher and student controls in addition to the leadership
indices. In this model ST is a vector of aggregated student background and teacher demographic
variables which include whether the test language is spoken at home, mothers’ education if less
than secondary, teacher experience, certification and gender. Due to the high correlation

83

between teacher age and teacher experience (0.89 in Hong Kong for example) only teacher
experience was used in the within country regressions. Model 2 is given below.

odds(Y = 1) = eα + β1(transactional) + β2(transformational) + β3(instructional)+ β4ST

[2]

Model 3 regressed principal reported satisfaction on only school controls and leadership
indices where SC is a vector of school controls such as availability of school resources for math
instruction, school size and whether the school is urban.

odds(Y = 1) = eα + β1(transactional) + β2(transformational) + β3(instructional)+ β4SC

[3]

Model 4 is a complete model that used all available variables for analysis.

odds(Y = 1) = eα + β1(transactional) + β2(transformational) + β3(instructional)+ β4ST + β4SC

[4]

2.8 Results
2.8.1 Analysis I: Research Question 1 (Teacher Level – Unweighted)
2.8.11 Hong Kong (SAR) Results. Results are presented by country for all analyses.
Table 7 shows results of Analysis 1 organized by country. This is a teacher level analysis where
the dependent variable is teacher reported satisfaction. In Hong Kong, principal observation of
teachers for evaluative purposes is negatively correlated with teacher satisfaction in models 1, 3
and 4. The principal’s role in setting directions for the school is significantly and positively
associated with greater teacher satisfaction in the sample (p<.01) across all models controlling
for an extensive set of student, teacher background and school controls. Larger school size is

84

also positively associated with higher teacher satisfaction (p<.10) in the sample (model 4)
controlling for aggregated student, teacher and school level variables.
2.8.12 Indonesia Results. In Indonesia, the principal’s role in setting directions is
significantly associated with greater teacher satisfaction across all the models specified (p<.01)
in the sample. Mother’s education (secondary or less) is negatively associated with teacher
satisfaction in the sample (p<.10).
2.8.13 Malaysia Results. In Malaysia aspects of leadership style are not significantly
associated with teacher satisfaction for teachers teaching a representative sample of students.
What appears to matter at the teacher level is whether or not students speak the language of the
test at home and school size. Models 2 and 4 indicate that if a student speaks the test language
teacher satisfaction in the sample is higher than if they do not (p<.01). Model 3 shows that
school size is positively correlated with teacher satisfaction in the TIMSS sample. At the teacher
level there are no significant results for Mongolia.
2.8.14 Taiwan Results. In Taiwan, it appears that the percentage of time spent by the
principal on instructional leadership is significantly associated with teacher satisfaction in the
data. No other variables are significant.
2.8.15 Thailand Results. Similar to Malaysia, in Thailand leadership style is not
significantly associated with teacher satisfaction in the data. Contextual variables such as
mother’s education level, small class size and teacher gender appear to matter more for teacher
satisfaction in this sample. For example, teacher satisfaction is lower if mother’s education level
(aggregated to the classroom level) is lower secondary. Teacher satisfaction is higher for female
teachers in the sample.

85

2.8.16 U.S.A. Results. For the U.S.A., the principal’s role is developing capacity among
staff members is positively associated with teacher satisfaction in the data (p<.01). In addition,
mothers education level (secondary or below) is associated with lower teacher satisfaction
(p<.01).
2.8.2 Analysis II: Research Question 1 (School Level – Weighted)
Results for Analysis II are presented by country in table 8 and also briefly summarized
below. For this analysis the dependent variable is teacher perception of teacher satisfaction,
which is aggregated by school. The strength of this analysis lies in the fact that the results
obtained are generalizable and consistent with the sampling structure of the TIMSS data
compared to Analysis I.
2.8.21 Hong Kong (SAR) Results. Similar to Analysis I, in Hong Kong, principal
observation of teachers for evaluative purposes is negatively correlated with teacher satisfaction
in models 1, 3 and 4. The principal’s role in setting directions for the school is significantly and
positively associated with teacher satisfaction (p<.01) across all models controlling for an
extensive set of student, teacher background and school controls.
2.8.22 Indonesia Results. In Indonesia, the principal’s role in setting directions for the
staff and school (which is a key component of transformational leadership) is positively and
significantly associated with teacher satisfaction (p<.01) across all models irrespective of the
introduction of a specific set of controls in each model. No demographic variables are
statistically significant.
2.8.23 Malaysia Results. In Malaysia, aspects of the principal’s transactional leadership
(the practice of contingent reward), specifically the use of teacher incentives in the form of a
bonus, housing or pay in retaining grade 8 teachers appears to be negatively correlated with

86

teacher satisfaction (p<.01) though this association is not robust to the introduction of student,
teacher and school controls. The aspect of transformational leadership that seems to matter
positively for average teacher satisfaction in Malaysia is the principal’s role in developing staff
capacity in schools (p<.01) though even this association does not hold in the complete model
(model 4). We do see a significant positive relationship in the final model between teacher
satisfaction and test language (a proxy for socio-economic status) if the language of the test is
spoken at home (p<.10).
2.8.24 Mongolia Results. In Mongolia, the principals role in ‘developing people’ is
negatively associated with teacher satisfaction (p<.05) while the aspect of transformational
leadership that matters is the principal’s effort to engage parents in the education of their
children through participation in school committees and events (‘Community Focus’ p<.01).
The availability of good school resources for instruction also appears to be statistically
significant (p<.05) and positively correlated with teacher satisfaction in at least two models
(model 3 & 4). Additionally, contextual variables like test language and small class have a small
but significant negative association with teacher satisfaction (p<.10).
2.8.25 Taiwan Results. In Taiwan, the transactional leadership practice – management
by exception as measured by principal observation of teacher practice for evaluation is positively
associated with teacher satisfaction (p<.10) for models 1 and 3 though this association does not
hold in the complete model. Similar to Mongolia, school context variables like ‘test language’
and ‘small class size’ appear to be important in the Taiwanese context (p<.05). Teachers in
whose classes students speak the test language at home appear to have high satisfaction levels as
opposed to those whose students do not.

87

2.8.26 Thailand Results. In Thailand, student socio-economic status as measured by
mother’s education level appears to be negatively correlated with teacher satisfaction (p<.05).
Female teachers in Thai schools have a higher satisfaction level than male teachers and this
effect is found to be statistically significant (p<.01).
2.8.27 U.S.A. Results. In the U.S.A., the aspect of transformational leadership that
appears to matter is the principal’s role in setting directions. The relationship is positive and
statistically significant across all models (p<.01) much like Hong Kong and Indonesia. In
addition, the principal’s effort to ‘develop people’ is negatively correlated with teacher
satisfaction (p<.10). Socio-economic status as measured by two proxy variables –lower
secondary education level of parents and speaking the test language at home is negatively
associated with teacher satisfaction. School and classroom context variables such as small class
and urbanicity are both positively associated with teacher satisfaction (p<.10 respectively).
2.8.3 Analysis III: Research Question 2 (School Level – Weighted)
Analysis III focuses on aspects of a principal’s leadership style (transactional,
transformational and instructional) and principal perceptions of teacher satisfaction. Table 9
presents results of four logistic models for each of the 7 countries. The results of Analysis III are
also briefly summarized by country below.
2.8.31 Hong Kong (SAR) Results. Similar to the teacher level analysis, in Hong Kong,
the principal’s role in setting directions for the school is significantly and positively correlated
with principal perceptions of teacher satisfaction (p<.01) across all models. In addition,
‘community focus’ is negatively correlated with principal reported teacher satisfaction (p<.10) in
models 1 & 2 though this effect does not hold in the comprehensive model (model 4).
Additionally, teacher and school context variables like teacher experience and school size are

88

significant and positively associated with principal reported satisfaction (p<.10). In Hong Kong
it appears that principals in Hong Kong perceive that male teachers in their schools experience
higher satisfaction levels as compared to females.
2.8.32 Indonesia Results. In Indonesia, the principal’s instructional leadership is
negatively correlated with principal reported teacher satisfaction (p<0.5) across all models.
Aspects of the principal’s transformational leadership such as setting directions and developing
people are also positively and significantly (p<.01) associated with principal reported teacher
satisfaction in that country. Teacher experience in Indonesia is negatively correlated with
principal reported teacher satisfaction levels (p<.10). Similar to Hong Kong, principals perceive
that female teachers have lower satisfaction levels as compared to male teachers.
2.8.33 Malaysia Results. In Malaysia as well, aspects of a principal’s transformational
leadership that matter are the principal’s role in ‘setting directions’ and ‘developing people’
across all models (p<.05 and p<.01 respectively). In addition, principals perceive teacher
satisfaction to be lower in large schools (p<.10) though this association does not hold in model 4.
2.8.34 Mongolia Results. In Mongolia principal’s perceive that the leadership practice
management by exception as measured by principal observation of teachers for evaluation is
significantly and negatively correlated with teacher satisfaction (p<.05) and that this effect holds
across all models. They also attach importance to behaving in a more transformational manner
as is evident by the significant association between ‘setting directions’ and principal perception
of teacher satisfaction (p<.05). Principals in Mongolia also perceive school and classroom
context variables such as urbanicity and small class size to be positively associated with teacher
satisfaction (p<.05 respectively).

89

2.8.35 Taiwan Results. Principals in Taiwan prioritize aspects of transformational
leadership such as ‘setting directions’ and ‘developing people’ (p<.01; p<.10 respectively).
Principals perceive that teacher demographics such as teacher experience are positively
correlated with teacher satisfaction (p<.10). They also perceive large schools to be positively
associated with satisfaction (model 3). Finally, principals in Taiwan perceive small class sizes to
be negatively associated with teacher satisfaction (p<.01).
2.8.36 Thailand Results. The transactional leadership practice management by
exception as measured by principal observation of teachers is perceived by Thai principals to be
significantly and positively associated with teacher satisfaction (p<.10). Similarly, Thai
principals perceive the use of teacher incentives to be strongly and positively correlated with
teacher satisfaction (p<.01) and this result holds across all models for Thailand. Principals in
Thailand also believe that behaving in a more transformational manner (specifically their role in
setting directions for the school as well as providing a strong community focus) is associated
with higher teacher satisfaction in their schools (p<.01) and this result too is consistent across all
models for Thailand. In addition, contextual variables such as level of mother’s education
(secondary or less) is positively associated with principal reported teacher satisfaction (p<.01).
2.8.37 U.S.A Results. Principals in the U.S.A. perceive aspects of both transactional and
transformational leadership to be associated significantly with teacher satisfaction levels in their
schools. In particular, principal observation of teachers for evaluative purposes and the
principals role in setting directions are both perceived to be positively correlated with teacher
satisfaction (p<.05, p<.01 respectively). Principals feel that the use of teacher incentives is
negatively correlated with teacher satisfaction levels (p<.10). Principals also tend to attach
importance to school context variables such as school size, good school facilities and small class

90

size in relation to how satisfied their teachers feel. A high level of resources available for
instruction is positively associated with teacher satisfaction in their schools (p<.05) as is small
class size (p<.05). Principals in the U.S.A. also perceive their teachers to be more satisfied in
larger schools (p<.05). The variable test language spoken at home is negatively correlated with
principal perception of teacher satisfaction.
2.9 Discussion
2.9.1 Research Question 1
Of the three leadership styles (transactional, instructional and transformational) the
variables indicative of the principal’s transformational leadership are consistently and
significantly associated with principal reported teacher satisfaction (Analysis III) across all the
countries analyzed. These variables are also significantly associated with teacher reported
satisfaction (Analysis II) in five of the seven countries analyzed. Aspects of the transformational
leadership index that appear to matter most are the principal’s role in ‘setting directions’ and in
‘developing people.’ These leadership dimensions are conceptually equivalent to what Bass and
Avolio (1994) conceptualized as the 4 I’s i.e. “idealized influence, inspirational motivation,
intellectual stimulation, and individualized consideration” (Leithwood & Sun, 2012).
The direction of the association between the ‘setting directions’ dimension of
transformational leadership and teacher satisfaction (principal reported) are consistent across the
six countries analyzed in Asia and the U.S. What the data are telling us is that in general
principals in Asia perceive that their teachers are more satisfied when they behave in a more
transformational manner. Results from teacher reported satisfaction at the school level (Analysis
II) corroborate an emphasis on the ‘setting directions’ aspect of transformational leadership in
three of the seven countries analyzed. These are Hong Kong, Indonesia, and the U.S.A. Results

91

are less clear cut in the teacher level analysis (Analysis I) though this set of results needs to be
interpreted with caution because they are not generalizable.
With respect to the other independent variables, aggregated student background, teacher
demographic controls, and school level controls were introduced separately in models 2 and 3
respectively and then comprehensively in the complete model. The purpose was to determine
whether teacher and school controls are significantly associated with teacher satisfaction
regardless of the key leadership indices of interest and also to test the robustness of findings
across the models. Results of the complete model (model 4) indicate support for the recent view
that organizational factors such as principal leadership are important for teacher satisfaction
(Johnson, Kraft & Papay, 2011; Ladd, 2011).
However, contextual variables also appear to be important in many of the countries
analyzed. For example, the variable named ‘test language spoken at home,’ which is used as a
proxy for socioeconomic status is significantly (p<0.05) associated with principal reported
teacher satisfaction in Thailand and the U.S. even after the introduction of organizational
variables implying that in these countries socioeconomic composition of schools is significantly
associated with principal reported teacher satisfaction independent of the effect of leadership
style and other teacher demographic and school level controls. In other words, in Thailand and
the U.S. (which are both relatively unequal societies)15 principals perceive that student
background matters for teacher satisfaction.
Other school and classroom contextual variables such as level of school resources
available for instruction and small class size also appear to be significantly correlated with
principal perceptions of teacher satisfaction. For example, principals in the U.S.A. and
15

The gini index for the U.S.A is 45.0 (2007); the gini index for Thailand is 53.6 (2009) on a scale of 0-100. The
more unequal a country’s income distribution is, the higher the gini index. An index of 53.6 is considered high.
Source: https://www.cia.gov/library/publications/the-world-factbook/rankorder/2172rank.html

92

Malaysia perceive the presence of a high level of school resources available for instruction as
being important for teacher satisfaction. Given the variation in resources/facilities within
schools and across school districts in the U.S.A. (the Gini index for the U.S. is 45.0 (2007))16,
this is not surprising. Inner city urban schools for example might have fewer resources available
to the teachers as compared with richer suburban schools. This result is corroborated by scholars
such as Buckley, Schneider and Shang (2005), and Schneider (2003) who underscore the
importance of good school facilities as central to the process of teaching and learning.
I find that the direction of the association between small class size and math teacher
satisfaction varies by country. In Mongolia and the U.S.A. there is a positive association
between principal perception of teacher satisfaction and small class size. In Taiwan and
Thailand, on the other hand, the association is negative. This finding is fairly non-intuitive at
first glance. However, many Asian countries have a strongly imbedded collectivist culture so it
is not unlikely that Taiwanese and Thai teachers attach less importance to individualized student
attention, which is often only possible in small class sizes. These countries have moderately
large class sizes (25-40 students).
One understanding that emerges with respect to demographic controls is that there is no
clear pattern that emerges across all Asian countries and all models analyzed. The direction of
the association too appears to vary from one country to the next unlike the results for leadership
style which appear to be fairly consistent between countries.
2.9.2 Research Question 2
Interestingly, this analysis also shows that teacher perceptions of leadership behavior and
their satisfaction levels differ from principal perceptions of leadership behavior and teacher
satisfaction (Tables 8 & 9). In the case of Hong Kong and Indonesia I find that teacher and
16

Source: https://www.cia.gov/library/publications/the-world-factbook/rankorder/2172rank.html

93

principal perceptions are similar with respect to dimensions of transformational leadership that
matter for teacher satisfaction. The leadership dimension that appears to matter for teachers of a
representative sample of students in Hong Kong is the principal’s ability to ‘set directions’ for
the school and to hold high performance expectations for achievement. Given the emphasis
placed on academic achievement (Cheong, 2000) in Hong Kong and the fact that public
examinations form an important systemic component of the school system in Hong Kong this is
not surprising.
In Indonesia, the transformational leadership dimension that seems to matter from the
perspective of the principal is the principals’ ability to ‘develop people’ by investing in their staff
and providing increased opportunities for self-improvement. Given that a school-based
management (SBM) system has been instituted in Indonesia under which teachers are expected
to contribute to curriculum development, it is understandable that principals would choose to
emphasize the professional development of their staff to build capacity in their schools.
Teachers teaching in a representative sample of schools in Indonesia, however, perceive the
principals’ transformational leadership in ‘setting directions’ to be more significant for their
overall sense of well-being and satisfaction. This may be due to the fact that teacher
empowerment as a concept is still nascent in Indonesia. In order to eventually have the requisite
freedoms to effect curriculum changes geared to local needs they need their principals to exercise
their authority and behave in a transformational way that will inspire such change.
In Malaysia, principals believe that behaving in a more transformational manner (setting
directions, developing people) is associated with better odds of teacher satisfaction. Teachers
teaching a representative sample of students in Malaysia, however, perceive no significant
association with the ‘setting directions’ aspect of transformational leadership and a negative

94

correlation between a principal’s transactional leadership and teacher satisfaction (p<.10). In
addition, principals perceive school context variables such as school resources (p<.05) and
school size (p<10) to be significantly correlated with teacher satisfaction in Malaysia, while
student background in the classroom matters more from a teacher perspective (p<.01). This
finding appears to reflect their respective spheres of influence. In other words, the results mirror
the centrality of the school for the principal versus the centrality of the classroom for the teacher.
Teachers teaching a representative sample of students in Taiwan do not perceive a
significant correlation between how a principal behaves and their satisfaction levels. This may
be due to the fact that teacher incentives are generally government funded and insured and are
not controlled by school principals. It may also be the case that though education authority has
formally decentralized from central to local governments and from local governments to schools,
traditional ideology still prevails and therefore, decentralization in Taiwan does not necessarily
imply less central control. Research shows that Taiwanese principals are somewhat reluctant to
cede administrative power to teachers and likewise the teachers are reluctant to exceed the
parameters of their traditional roles (Pan & Chen, 2011). Results from analysis II appear to
provide some support for this reasoning. Interestingly, principal perceptions about their
teachers’ satisfaction differed markedly from the perceptions of the teachers discussed above.
Similar to Malaysia, student background (test language spoken at home) matters for teacher
satisfaction (p<0.05) in the Taiwanese sample. The role of teachers in Taiwan goes beyond
merely imparting subject matter knowledge. Teachers are also expected to support the personal
and social development of students and to foster good behavior. Thus, teachers play a role
similar to that of parents in terms of providing personal guidance and acting as role models for

95

the children. With this backdrop it seems reasonable that student backgrounds (and subsequent
classroom behavior) would be important from a teacher’s perspective.
In Thailand, aspects of the principal’s transactional leadership (contingent reward and
management by exception) were significantly and positively associated with principal reported
teacher satisfaction (p<0.01). This appears to imply that principals in Thailand play a more
traditional leadership role vis-à-vis their staff. Under the transactional leadership style the leader
monitors the performance of teachers and interacts with them if their performance deviates from
expected norms. This mode of operating is fairly consistent with Thailand’s more hierarchical
leadership structure as well as Thai cultural parameters which reward compliance and nonconfrontational behavior. This may also be the reason why Thai teachers in the TIMSS sample
do not perceive principals’ transformational behaviors to be significantly associated with their
satisfaction levels. Additionally, Thailand is a community based culture. Thai culture
encourages interdependence and a ‘we’ consciousness (Hallinger & Kantamara, 2000). In
keeping with Thai collectivist culture, principals in Thailand also appear to prioritize parent
school interactions as shown by the significance of the ‘community focus’ dimension within
transformational leadership.
The following section explores in greater detail how rich country context may help
explain the results for Hong Kong – a country that has in recent years begun to attract greater
research attention.
2.10

Contextual Discussion of Results in Hong Kong
2.10.1 Country Background: Hong Kong has a population of around 7 million people

living in a small area of 1,000 square kilometers (OECD, 2010). The population of Hong Kong
is unique in that it is one hundred percent urban. The average GDP per capita is above USD

96

42,000 also making it among the ten richest nations (OECD, 2010) in the world. The population
is predominantly ethnic Chinese with a few Caucasians from Western countries. A small
percentage of the population is from Indonesia and the Philippines. Chinese and English are the
official languages in Hong Kong though Cantonese (a dialect of Chinese) is spoken by the
majority of the people. According to the 2006 population by-census, approximately 90 percent
of the population age 5 and above speak Cantonese (TIMSS 2011 Encyclopedia). “The
government has adopted a biliterate (English and Chinese) and trilingual (Putonghua, Cantonese,
& English) policy for the education system” (TIMSS 2011 Encyclopedia, p.369). However,
most primary schools use Cantonese as the medium of instruction. Secondary schools wishing to
use English as a medium of instruction require government approval.
Historically, Hong Kong was a British colony for a ninety-nine year period until the year
1997 when it reverted back to China under the “one country, two systems” notion. Under this
arrangement, China resumed sovereignty over Hong Kong but Hong Kong remained a separate
jurisdiction enjoying autonomy in all areas except military defence and diplomatic relations. As
a Special Administrative Region of China (SAR), Hong Kong has an independent legislature, its
own currency and its own policies independent of the national government in Beijing.
Nonetheless, Hong Kong shares certain cultural values and similarities with China such as a long
tradition of valuing education highly and high hopes for the educational attainment of school
children. Implicit in this cultural respect for education is belief in the philosophy that hard work
can transcend difficult learning environments and is a means of social mobility. This has in turn
translated into a zest for credentials and an almost exclusive emphasis on examination results for
validating genuine learning (OECD, 2010).

97

Although cultural values and traditions overlap between Hong Kong and China, Hong
Kong’s education system is quite distinct from that of the rest of China with its own unique
structure and reform trajectory.
2.10.2 Overview of Educational System: The educational system in Hong Kong was
initially modelled on the British system with six years of primary education (grades 1-6), five
years of secondary education (Secondary 1-5), two years of pre-university study (Secondary 6-7)
culminating in three years of university education. This structure may have been in large
measure due to its colonial legacy. In recent years however, a number of policy reforms have
been implemented that have modified the system.
In 2009 a new 6-3-3-4 structure was instituted (6 years of primary, 3 years of lower
secondary, 3 years of upper secondary followed by 4 years of university education) consistent
with the system currently followed in China and many other parts of the world. It would be fair
to say that the Hong Kong education system is very much “a hybrid of Chinese culture and
British traditions” (OECD, 2010). The reform introduced in 2009 has kept the curriculum for the
first nine years of compulsory basic education essentially unchanged though the old system of
streaming by subject from grade 10 onward has been dropped. All students are now expected to
take four core subjects, which are English Language, Chinese Language, Mathematics and
Liberal Studies. A new examination to be taken at the end of grade 12 (Secondary 6) known as
the Hong Kong Diploma of Secondary Education Examination introduced in 2012 has replaced
two exams taken at the end of grades 11 and 13 previously.
The Hong Kong Education Bureau (EDB) is the body primarily responsible for
formulating policy and introducing legislation on education from the pre-primary to the tertiary
level though preprimary education in Hong Kong is not considered part of compulsory education

98

and all Kindergartens are in fact privately managed. The EDB essentially administers more than
1000 schools. All schools in Hong Kong need to be registered under the EDB’s Education
Ordinance and to observe the bureau’s regulations. Advice on curriculum development at all
levels is provided by the Curriculum Development Council, which is a freestanding advisory
body appointed by the Chief Executive in Hong Kong. Additionally, the EDB provides a list of
“acceptable” textbooks and teaching materials for all subjects and all grade levels. These lists
are not mandated. Interestingly, however, nearly all local schools revert to using them
potentially indicating that centralized influence on curriculum and instruction remains fairly
strong in schools. The EDB recommends the use of both formative and summative assessments
to collect information on student learning. Nevertheless, tests and exams are still the primary
method to inform teachers of the learning levels of their students.
In addition to public schools, schools in Hong Kong are classified as aided schools, direct
subsidy schools and international schools. Aided schools are free and funded by the government
but run by a private sponsor. Direct subsidy schools are private schools that receive some
government funding. International schools are country-specific schools that teach a curriculum
from their own country.
2.10.3 Profile of Teachers in Hong Kong: Anyone wishing to teach is required to
register under the EDB’s Education Ordinance as either a registered teacher or a permitted
teacher. Registered teachers are those with approved teaching qualifications and experience as
stated in the ordinance. Permitted teachers hold requisite academic qualifications but have no
teacher training and are given permits by the EDB to teach specified subject(s) in specified
schools. The number of permitted teachers however, appears to be small. In 2010-11 for
example, approximately 95.7 percent of primary school teachers and 94.4 percent of secondary

99

school teachers were professionally trained (TIMSS 2011). In addition, there are specialist
teachers who usually teach math and science at the secondary level and sometimes at the primary
level as well.
The TIMSS 2007 data for Hong Kong show that on average the teachers sampled are 38
years old. Average teacher experience is approximately 12.87 years and teachers with this level
of experience also have higher odds of job satisfaction as compared with teachers who are
younger. On average, teacher perceptions of their satisfaction are reportedly lower (0.40) than
principal perceptions of teacher satisfaction (0.65). One reason for this dichotomy may be the
high power distance between principals and teachers which presumably detracts from teachers’
sense of autonomy in the workplace. The data also show that the majority of teachers (90
percent) in the TIMSS sample possess teacher certification. This is consistent with the fact that
generally speaking professional training for teachers is prioritized in Hong Kong. Additionally,
the TIMSS data show that at the Secondary level the proportion of male teachers is higher than
female teachers.
2.10.4 School Leadership in Hong Kong: Hong Kong is a society that exhibits a high
“power-distance” in its relationships between leaders and followers (Walker, Lee & Bryant,
2014). A high power distance culture implies that subordinates maintain a respectable distance
from their leaders and do not question the leader’s mandates. This is because in such societies
disagreement with authority is construed as a sign of disrespect. In this respect, Hong Kong is
not very different from other Asian societies many of which (Thailand for example) are also high
power distance cultures.
Despite this power distance however, teachers in Hong Kong are likely to perceive
leadership practices around quality assurance and accountability as negative (Walker, Lee &

100

Bryant, 2014). The TIMSS 2007 data for Hong Kong appear to support this finding. The odds
of teacher satisfaction among teachers teaching a representative sample of students in Hong
Kong are lower when certain accountability mechanisms such as principal observation of
teaching are put into effect.
It is entirely possible that a cultural disconnect between externally imposed
accountability requirements and cultural realities within schools is responsible for the negative
response (Walker & Qian, 2012). This disconnect often arises from a blind adoption of policies
from societies with very different sociocultural traditions and values. Given that Hong Kong’s
accountability policy framework was originally taken from the UK, this explanation is fairly
plausible. The data show that even in hierarchically structured societies such as Hong Kong
observation of teachers’ classroom practices is interpreted as principal intrusion into teachers’
traditional domains presumably because such practices generate a negative pressure on teachers.
An area of difficulty for principals in Hong Kong is providing incentives for teachers
(Cheong, 2000). This is because the salary structure for teachers in Hong Kong is fairly rigid
and there are few opportunities available for promotion. The principal therefore has little
discretion to reward teachers. This context also helps to explain why in the case of Hong Kong
the coefficient on the variable measuring the relationship between teacher incentives and teacher
satisfaction is insignificant.
It has been argued that developing and effectively utilizing structures of communication
for school improvement is essential to the work of school leaders (Walker et al. 2014). My
results pertaining to the association between dimensions of transformational leadership and
teacher satisfaction seem to corroborate this viewpoint. Specifically, I find that when principals
in Hong Kong prioritize fostering a shared vision and common school goals among members of

101

the school community teacher reported satisfaction is higher. This is indicated by the positive
odds ratio on the setting directions aspect of transformational leadership. This finding is also
consistent with earlier work on leadership behavior in Hong Kong (see Chui, Sharpe, &
McCormick, 1996).
The setting direction dimension of transformational leadership is also indicative of the
expectations teachers hold with respect to student performance. The positive association
between setting direction and teacher reported satisfaction is again congruent with what we know
about learning environments in Hong Kong. High principal and teacher expectations for student
achievement are in keeping with the fact that Hong Kong has prioritized student learning through
all its different reform phases.
In sum, while principals’ efforts to effectively communicate with their teachers with
respect to their school’s mission and goals is important (as shown in my analysis), extant
research also indicates that in developing communication frameworks within their schools,
school leaders in Hong Kong will need to leverage the influence of societal culture in a manner
that best supports positive teacher and student outcomes. In particular, leaders would need to be
sensitive to the existence of a power distance between them and their staff and make a concerted
and conscious effort to empower their teachers to participate in school-wide decisions.
2.11 Conclusion and Limitations
Despite a growing recognition in the field of the importance of context and its influence
on practice it has not been adequately explored in the literature. By attempting to parse out
aspects of leadership that matter in diverse geographic and social contexts that differ
substantively from the U.S.A., this paper makes a valuable contribution to the existing body of
knowledge on leadership. Additionally, the analysis benefits from the use of internationally

102

recognized large-scale survey data (TIMSS 2007) with high levels of comparability and
reliability in the instruments across countries. Only a few scholars (Sargent & Hannum, 2005
and Michaelowa & Wittman, 2007) have previously attempted to use large-scale survey data to
understand and explore the notion of teacher satisfaction albeit in different geographic and
cultural contexts from the one analyzed here.
This is one of the first multi-country, large-scale analyses to examine both principal and
teacher perspectives on teacher satisfaction for several countries in Asia and the U.S. after
accounting for an extensive range of aggregated student and teacher demographic variables and
school controls. A very clear pattern that emerges from this analysis is that principals’
transformational leadership is significantly associated with principal reported teacher satisfaction
across all the Asian countries analyzed as well as the United States (p<.01). This appears to
indicate that a visionary and supportive leadership style as measured by the transformational
leadership index is important for teacher satisfaction in a multi-country context at least from the
perspective of the school principals in these countries. Specifically, two dimensions of
transformational leadership - the principal’s role in ‘setting directions’ for the school and in
‘developing people’ were found to be consistently associated with higher odds of teacher
satisfaction in the countries analyzed. This finding is also consistent with the U.S. literature on
transformational leadership in schools.
Interestingly, this pattern is not so evident when we analyze teacher perspectives on
satisfaction and leadership style (with the exception of the teachers sampled in Hong Kong,
Indonesia and Malaysia). Teachers of a representative sample of students in Mongolia, Taiwan
and Thailand seem to attach more importance to student and school background variables (with
respect to their satisfaction levels) than particular aspects of transformational or instructional

103

leadership. One possible reason for this divergence in perspectives may be that principals in
several of the countries analyzed are acting on the notions of transformational leadership
considered appropriate by their Ministries (who in turn may be influenced by the politics and
processes of globalization and policy borrowing) but that these principals are still in the process
of operationalizing these predominantly Western notions in the context of their respective
educational systems, many of which are still fairly centralized (See Steiner-Khamsi, 2004 for an
understanding of policy borrowing in Mongolia). As such, aspects of transformational
leadership such as ‘setting directions’ and ‘developing people’ which are inherently western
conceptualizations of leadership might not be well understood or easily internalized by teachers
in these countries. This may explain why principals in all the countries analyzed perceive these
dimensions to be highly significant while teachers in this sample (with the exception of Hong
Kong, Indonesia and Malaysia) do not.
An alternative explanation is that the divergence in perspectives simply mirrors the
centrality of the school for the principal versus the centrality of the classroom for the teacher.
Irrespective of the underlying reasons for differing principal and teacher perspectives on teacher
satisfaction, from a policy perspective this analysis provides an opportunity for school leaders to
gain a richer understanding of particular aspects of leadership behavior that work to make their
teachers more satisfied and therefore more likely to be productive and to stay committed to their
schools in the long term. Understanding the nuances of leadership behavior and its variations in
myriad settings has much to teach us about the task of changing schools and improving practice.
Nonetheless, the paper also has several limitations. First, because the main purpose of
the analysis is to deconstruct the type of leadership that matters in the Asian context,
investigating the complexity of interrelationships between the different types of leadership

104

behaviors17 is beyond the scope of this analysis. Additionally, although there are aspects of
transformational and instructional leadership that overlap (such as creation of a school vision,
professional development of teachers etc.) care was taken to minimize such overlaps in creating
indices of transformational and instructional leadership so as to avoid issues of multicollinearity.
Second, TIMSS 2007 mainly allows analysis of dyadic pairs of teachers and principals in
schools. This deviates from the literature on teacher satisfaction which ideally references a
larger sample of teachers than is available via TIMSS. That said, TIMSS is the only large-scale
survey data publicly available that collects rich contextual information on students, teachers and
schools for several countries in Asia. Within Asia, OECD’s TALIS 2008 contains information
for only Japan and S. Korea.
Third, the analysis is correlational in nature. Causal arguments cannot be made about the
associations between leadership and satisfaction.
Future research can be enriched further in one of two ways. First, the use of qualitative
fieldwork may provide valuable insight into the satisfaction-leadership relationship by adding
finer grained level of information at the teacher level not generally available through survey data.
Second, the use of sub-scales measuring the satisfaction construct rather than a global measure of
satisfaction as used in this analysis may facilitate a more nuanced understanding of teacher
satisfaction.

17

See for example Marks and Printy (2003) who show that instructional leadership may also be transformational in
some instances.

105

Table 7: Table of Results for Teacher perception of Teacher Satisfaction & Principal Leadership Style (Analysis I - Teacher Level)
HONG KONG (SAR)
INDONESIA
MALAYSIA
VARIABLE NAMES
(1)
(2)
(3)
(4)
(1)
(2)
(3)
(4)
(1)
(2)
(3)
Transactional Leadership
prinobs 0.286*
0.263
0.184** 0.150** 1.551
1.398
2.003
1.781
0.378
0.474
0.346
(0.216)
(0.241)
(0.131)
(0.115)
(1.182)
(1.044)
(1.462)
(1.252)
(0.425)
(0.544)
(0.365)
tch_incentives 1.680
1.599
2.064
1.741
0.968
0.943
1.039
1.080
0.731
0.673
0.704
(1.000)
(1.245)
(1.363)
(1.368)
(0.224)
(0.230)
(0.270)
(0.291)
(0.182)
(0.176)
(0.178)
Instructional Leadership
Percent time (yr) on IL 1.006
1.019
1.013
1.026
0.990
0.990
0.983
0.984
1.005
1.002
1.004
(0.019)
(0.022)
(0.021)
(0.024)
(0.012)
(0.013)
(0.013)
(0.014)
(0.011)
(0.012)
(0.011)
Transformational Leadership
set_directions 2.022*** 1.762** 2.165*** 1.975*** 2.128*** 2.177*** 2.015*** 2.071*** 1.258
1.138
1.252
(0.431)
(0.399)
(0.520)
(0.478)
(0.338)
(0.376)
(0.357)
(0.391)
(0.176)
(0.176)
(0.182)
dev_people 0.818
0.824
0.819
0.836
1.072
1.106
0.962
0.956
1.229
1.126
1.177
(0.148)
(0.158)
(0.158)
(0.170)
(0.139)
(0.152)
(0.139)
(0.147)
(0.190)
(0.177)
(0.186)
community_focus 0.830
0.862
0.756
0.763
(0.173)
(0.194)
(0.163)
(0.170)
Language of Test is Spoken at Home
1.014
1.009
0.998
0.997
1.018***
(0.025)
(0.027)
(0.003)
(0.004)
(0.004)
Mothers Education
1.102
0.474
0.479
0.256*
0.902
(1.339)
(0.625)
(0.291)
(0.199)
(0.942)
Teacher Experience (Yrs)
0.980
0.984
0.981
0.981
1.012
(0.016)
(0.018)
(0.016)
(0.018)
(0.015)
Female Teacher
0.945
0.865
0.773
0.795
0.635
(0.315)
(0.293)
(0.189)
(0.213)
(0.196)
Small Class Size
0.619
0.628
0.557
0.702
1.815
(0.317)
(0.330)
(0.223)
(0.320)
(1.928)
School Size
1.001
1.002*
1.001
1.000
0.999***
(0.001)
(0.001)
(0.000)
(0.000)
(0.000)
Good School Facility
0.914
0.792
2.050
1.560
0.974
(0.343)
(0.324)
(1.096)
(0.896)
(0.248)
Urban
1.017
0.776
1.497
1.616
0.614
(0.348)
(0.305)
(0.452)
(0.522)
(0.227)
Constant
2.004
0.645
0.618
0.532
0.509
1.779
0.299
1.841
4.556
1.765
11.59**
(1.552)
(1.811)
(0.764)
(1.670)
(0.421)
(1.923)
(0.243)
(2.144)
(5.229)
(2.838)
(12.79)
Observations
211
190
201
182
361
344
334
317
307
297
305
Notes: 1. All coefficients shown are odds ratios; 2. Std errors given in parentheses; 3. ***p<0.01, **p<0.05, *p<0.10; 4. DV=teacher perception of satisfaction.
Model 1= Teacher satisfaction on indices of leadership; Model 2=teacher satisfaction on leadership indices with student/teacher controls aggregated by teacher; Model 3 =
satisfaction on leadership with only school controls and model 4=Complete model with satisfaction regressed on all available variables.

106

(4)
0.427
(0.474)
0.675
(0.178)
1.004
(0.013)
1.147
(0.181)
1.138
(0.184)
1.016***
(0.005)
0.929
(1.037)
1.014
(0.016)
0.653
(0.202)
1.877
(2.028)
1.000
(0.000)
0.908
(0.246)
0.782
(0.329)
2.988
(4.957)
295

Table 7 (cont'd)
COUNTRY NAMES
VARIABLE NAMES
Transactional Leadership

(1)

prinobs 2.089
(2.502)
tch_incentives 1.140
(0.326)
Instructional Leadership
Percent time (yr) on IL 0.996
(0.013)
Transformational Leadership
set_directions 0.946
(0.193)
dev_people 1.126
(0.246)
community_focus Language of Test is Spoken at Home
Mothers Education
Teacher Experience (Yrs)
Female Teacher
Small Class Size

MONGOLIA
(2)
(3)

(4)

2.396
(2.980)
1.021
(0.310)

1.938
(2.267)
1.272
(0.388)

0.993
(0.014)
0.821
(0.183)
1.150
(0.286)
0.977
(0.015)
0.688
(0.514)
1.013
(0.015)
0.770
(0.301)
0.577
(0.303)

School Size

(1)

TAIWAN
(2)
(3)

(4)

2.124
(2.672)
1.119
(0.361)

1.437
(0.423)
3.258
(2.741)

1.232
(0.393)
3.168
(2.788)

1.353
(0.420)
6.261
(7.008)

1.230
(0.409)
6.320
(7.415)

0.995
(0.014)

0.996
(0.015)

0.988
(0.011)

0.978*
(0.012)

0.986
(0.011)

0.976*
(0.013)

0.879
(0.196)
1.053
(0.250)
-

0.783
(0.191)
1.036
(0.282)
0.978
(0.017)
1.287
(1.165)
1.006
(0.015)
0.622
(0.261)
0.684
(0.410)
1.000
(0.000)
3.563
(2.983)
1.232
(0.895)
2.649
(5.290)

1.063
(0.160)
1.088
(0.173)
1.048
(0.188)

1.104
(0.182)
1.048
(0.177)
1.160
(0.239)
1.009
(0.008)
1.225
(0.982)
0.979
(0.017)
0.935
(0.266)
0.608
(0.368)

1.092
(0.171)
1.190
(0.209)
1.029
(0.194)

1.088
(0.407)

0.823
(0.950)

1.000
(0.000)
1.241
(0.351)
0.846
(0.279)
0.923
(0.419)

1.122
(0.192)
1.152
(0.216)
1.133
(0.245)
1.010
(0.008)
1.384
(1.321)
0.982
(0.018)
0.847
(0.259)
0.687
(0.435)
1.000
(0.000)
1.311
(0.407)
0.793
(0.317)
0.561
(0.714)

174

248

225

230

209

Constant

0.357
(0.434)

4.768
(8.784)

1.000
(0.000)
3.272
(2.484)
1.128
(0.807)
0.287
(0.341)

Observations

204

188

189

Good School Facility
Urban

Notes:
1. All coefficients shown are odds ratios; 2. Std errors given in parentheses; 3. ***p<0.01, **p<0.05, *p<0.10; 4. DV=teacher perception of satisfaction.
Model 1= Teacher satisfaction on indices of leadership; Model 2=teacher satisfaction on leadership indices with student/teacher controls aggregated by
teacher; Model 3 = satisfaction on leadership with only school controls and model 4=Complete model with satisfaction regressed on all available
variables.

107

Table 7 (cont'd)
COUNTRY NAMES
VARIABLE NAMES
Transactional Leadership

(1)

prinobs 1.340
(0.538)
tch_incentives 0.940
(0.236)
Instructional Leadership
Percent time (yr) on IL 1.005
(0.010)
Transformational Leadership
set_directions dev_people 1.239
(0.185)
community_focus 0.965
(0.150)
Language of Test is Spoken at Home

(2)

THAILAND
(3)

(4)

(1)

(2)

1.802
(0.828)
0.942
(0.256)

1.603
(0.716)
0.879
(0.253)

1.832
(1.156)
0.672
(0.253)

1.631
(1.129)
0.807
(0.338)

1.822
(1.159)
0.692
(0.265)

1.664
(1.133)
0.852
(0.363)

0.994
(0.011)

1.002
(0.012)

0.998
(0.012)

0.996
(0.007)

0.996
(0.008)

0.997
(0.007)

0.997
(0.008)

1.110
(0.176)
0.956
(0.150)
0.994
(0.003)
0.039***
(0.040)
1.002
(0.013)
1.900**
(0.518)
0.481*
(0.196)

1.153
(0.178)
1.003
(0.162)

0.930
(0.096)
1.869***
(0.202)
1.128
(0.179)

0.917
(0.104)
1.657***
(0.198)
1.131
(0.200)
0.997
(0.006)
0.295***
(0.105)
1.003
(0.009)
1.016
(0.197)
1.279
(0.239)

0.919
(0.097)
1.920***
(0.213)
1.121
(0.180)

0.692
(0.428)

1.568
(1.459)

1.000
(0.000)
0.929
(0.166)
1.146
(0.260)
0.783
(0.501)

0.901
(0.105)
1.726***
(0.212)
1.126
(0.198)
0.995
(0.006)
0.277***
(0.104)
0.999
(0.010)
1.029
(0.203)
1.252
(0.243)
1.000
(0.000)
0.862
(0.167)
1.003
(0.246)
2.455
(2.402)

626

564

608

548

Constant

0.679
(0.308)

19.60**
(24.07)

Observations

289

287

281

Teacher Experience (Yrs)
Female Teacher
Small Class Size

(4)

1.255
(0.524)
0.819
(0.227)

1.113
(0.178)
0.999
(0.163)
0.994
(0.004)
0.071**
(0.085)
1.005
(0.014)
1.728**
(0.475)
0.539
(0.227)
1.000*** 1.000
(0.000) (0.000)
1.061
0.943
(0.403) (0.406)
0.526
0.538
(0.244) (0.261)
0.319** 6.737
(0.172) (9.513)

Mothers Education

USA
(3)

School Size
Good School Facility
Urban

279

Notes:
1. All coefficients shown are odds ratios; 2. Std errors given in parentheses; 3. ***p<0.01, **p<0.05, *p<0.10; 4. DV=teacher perception of satisfaction.
Model 1= Teacher satisfaction on indices of leadership; Model 2=teacher satisfaction on leadership indices with student/teacher controls aggregated
by teacher; Model 3 = satisfaction on leadership with only school controls and model 4=Complete model with satisfaction regressed on all available
variables.

108

Table 8: Table of Re sults for Te ache r Pe rce ption of Te ache r Satisfaction & Principal Le ade rship Style (Analysis II - School Le ve l)
HONG KONG (SAR)
INDONESIA
MALAYSIA
VARIABLE NAMES
(1)
(2)
(3)
(4)
(1)
(2)
(3)
(4)
(1)
(2)
(3)
Transactional Leadership
Principal observation 0.113*
0.124
0.087** 0.088** 2.754
2.912
3.112
3.201
1
1
1
(0.145)
(0.169)
(0.097)
(0.093)
(3.607)
(3.931)
(3.815)
(4.005)
Teacher incentives 2.366
2.542
2.470
2.160
0.895
0.725
0.856
0.717
0.570*
0.584
0.595
(1.717)
(2.391)
(1.880)
(2.121)
(0.356)
(0.311)
(0.351)
(0.317)
(0.188)
(0.197)
(0.193)
Instructional Leadership
Percent time (yr) on IL 0.997
1.007
1.007
1.020
1.012
1.012
1.002
1.004
0.994
0.985
0.993
(0.020)
(0.023)
(0.021)
(0.024)
(0.019)
(0.019)
(0.020)
(0.022)
(0.013)
(0.015)
(0.013)
Transformational Leadership
Set directions 1.844** 1.765** 1.927** 1.920** 2.816*** 2.921*** 2.836*** 3.037*** 1.115
1.052
1.108
(0.440)
(0.483)
(0.512)
(0.527)
(0.848)
(0.863)
(0.922)
(0.953)
(0.248)
(0.213)
(0.234)
Develop people 0.857
0.872
0.825
0.826
1.259
1.396
1.167
1.242
1.364*
1.254
1.376*
(0.184)
(0.204)
(0.185)
(0.197)
(0.233)
(0.294)
(0.240)
(0.292)
(0.241)
(0.228)
(0.238)
Community focus 0.742
0.772
0.712
0.748
1.101
1.146
1.153
(0.173)
(0.189)
(0.160)
(0.172)
(0.209)
(0.228)
(0.222)
Language of Test is Spok en at Home
1.005
0.999
0.994
0.995
1.020***
(0.023)
(0.024)
(0.007)
(0.007)
(0.006)
Mothers Education
2.321
1.247
0.208
0.267
1.583
(3.200)
(1.945)
(0.229)
(0.331)
(2.071)
Teacher Experience (Yrs)
0.972
0.971
0.961
0.957
0.997
(0.018)
(0.019)
(0.029)
(0.031)
(0.017)
Female Teacher
0.981
0.844
0.853
0.756
0.765
(0.381)
(0.341)
(0.393)
(0.388)
(0.325)
Small Class Size
0.865
0.968
0.435
0.559
0.954
(0.655)
(0.772)
(0.239)
(0.329)
(0.599)
School Size
1.001
1.002
1.001
1.001
1.000
(0.001)
(0.001)
(0.001)
(0.001)
(0.000)
Good School Facility
0.952
0.829
0.983
1.037
0.874
(0.400)
(0.361)
(0.737)
(0.844)
(0.266)
Urban
0.927
0.875
1.072
1.090
0.468*
(0.358)
(0.410)
(0.491)
(0.514)
(0.202)
Constant
6.213
2.148
1.484
1.141
0.157
1.618
0.128
1.046
3.035*** 0.870
3.248**
(8.220)
(6.107)
(2.528)
(3.472)
(0.223)
(2.924)
(0.177)
(1.894)
(1.157)
(1.079)
(1.640)

(4)
1
0.609
(0.203)
0.988
(0.014)
1.111
(0.213)
1.294
(0.232)
1.184
(0.233)
1.021***
(0.006)
0.936
(1.334)
0.995
(0.017)
0.743
(0.327)
0.995
(0.610)
1.000
(0.000)
0.825
(0.257)
0.650
(0.291)
0.972
(1.377)

Observations
195
177
186
169
254
254
235
235
278
272
276
270
Notes:
1. All coefficients shown are odds ratios; 2. Std errors given in parentheses; 3. ***p<0.01, **p<0.05, *p<0.10; 4. DV=teacher perception of satisfaction.
Model 1= Teacher satisfaction on indices of leadership; Model 2=teacher satisfaction on leadership indices with student/teacher controls aggregated by
teacher; Model 3 = satisfaction on leadership with only school controls and Model 4=Complete model with satisfaction regressed on all available variables.

109

Table 8 (cont'd)
COUNTRY NAMES
VARIABLE NAMES
Transactional Leadership

(1)

prinobs 4.036
(4.926)
tch_incentives 1.345
(0.526)
Instructional Leadership
Percent time (yr) on IL 1.002
(0.021)
Transformational Leadership
set_directions 0.985
(0.333)
dev_people 0.654
(0.203)
community_focus 2.454**
(0.853)
Language of Test is Spoken at Home
Mothers Education
Teacher Experience (Yrs)
Female Teacher
Small Class Size

(2)

MONGOLIA
(3)

(2)

TAIWAN
(3)

(4)

3.271
(3.523)
1.624
(0.668)

3.445
(3.635)
1.334
(0.569)

2.573*
(1.464)
2.490
(2.170)

1.660
(0.664)
2.303
(2.381)

2.455*
(1.240)
3.609
(4.110)

1.810
(0.773)
3.326
(4.299)

0.996
(0.022)

0.988
(0.020)

0.987
(0.021)

1.009
(0.019)

0.996
(0.019)

1.008
(0.019)

0.995
(0.021)

1.227
(0.397)
0.613
(0.184)
2.775***
(1.041)
0.937*
(0.033)
0.836
(0.927)
1.021
(0.020)
1.077
(0.487)
0.306*
(0.216)

0.859
(0.250)
0.540**
(0.150)
3.129***
(1.052)

0.866
(0.212)
1.342
(0.301)
1.087
(0.300)

1.012
(0.219)
1.133
(0.255)
1.335
(0.404)
1.021*
(0.011)
3.961
(4.227)
0.986
(0.021)
0.770
(0.315)
0.180**
(0.125)

0.966
(0.228)
1.291
(0.302)
1.049
(0.291)

1.088
(0.265)
1.124
(0.262)
1.264
(0.435)
1.023**
(0.011)
9.128
(12.89)
0.985
(0.024)
0.656
(0.297)
0.192**
(0.152)
1.000
(0.000)
1.135
(0.518)
1.408
(0.827)
0.022**
(0.039)

Good School Facility
Urban
0.121*
(0.148)

(1)

5.031
(6.059)
1.184
(0.465)

School Size

Constant

(4)

57.45
(204.0)

0.987
(0.314)
0.478**
(0.142)
3.482***
(1.340)
0.945*
(0.028)
1.888
(2.581)
1.012
(0.019)
0.774
(0.358)
0.411
(0.264)
1.000
1.000
(0.000) (0.000)
10.78*** 14.42**
(9.012) (15.06)
1.963
2.325
(2.116) (2.433)
0.115** 22.22
(0.123) (67.14)

0.372
(0.280)

0.071*
(0.106)

1.000
(0.000)
0.742
(0.300)
1.144
(0.684)
0.328
(0.263)

Observations
202
187
189
174
242
220
224
204
Notes:
1. All coefficients shown are odds ratios; 2. Std errors given in parentheses; 3. ***p<0.01, **p<0.05, *p<0.10; 4.
DV=teacher perception of satisfaction.
Model 1= Teacher satisfaction on indices of leadership; Model 2=teacher satisfaction on leadership indices with
student/teacher controls aggregated by teacher; Model 3 = satisfaction on leadership with only school controls and
Model 4=Complete model with satisfaction regressed on all available variables.

110

Table 8 (cont'd)
COUNTRY NAMES
VARIABLE NAMES
Transactional Leadership

(1)

prinobs 1.041
(0.652)
tch_incentives 1.001
(0.343)
Instructional Leadership
Percent time (yr) on IL 1.020
(0.019)
Transformational Leadership
set_directions 1.292
(0.307)
dev_people 1.182
(0.242)
community_focus 1.263
(0.343)
Language of Test is Spoken at Home
Mothers Education
Teacher Experience (Yrs)
Female Teacher
Small Class Size

(2)

THAILAND
(3)

(2)

USA
(3)

(4)

1.280
(0.813)
1.042
(0.367)

1.096
(0.674)
1.049
(0.393)

1.786
(1.383)
0.923
(0.497)

0.784
(0.806)
1.468
(0.814)

1.827
(1.426)
0.943
(0.597)

0.821
(0.805)
1.168
(0.703)

1.000
(0.019)

1.028
(0.020)

1.007
(0.020)

1.014
(0.015)

1.022
(0.014)

1.020
(0.015)

1.021
(0.014)

1.081
(0.291)
1.098
(0.244)
1.251
(0.300)
0.999
(0.005)
0.125
(0.175)
0.996
(0.021)
3.921***
(1.670)
0.576
(0.269)

1.287
(0.340)
1.159
(0.234)
1.440
(0.379)

1.177
(0.329)
1.119
(0.242)
1.366
(0.339)
1.000
(0.005)
0.0196**
(0.038)
0.996
(0.021)
3.754***
(1.634)
0.564
(0.270)
1.000
(0.000)
0.581
(0.380)
1.119
(0.989)
19.83
(43.23)

1.730***
(0.289)
0.799
(0.181)
1.261
(0.302)

1.495**
(0.271)
0.678
(0.164)
1.312
(0.389)
0.970**
(0.015)
0.153**
(0.132)
1.003
(0.022)
2.037
(1.050)
2.477**
(1.102)

1.794***
(0.297)
0.740
(0.170)
1.107
(0.260)

1.563**
(0.277)
0.652*
(0.161)
1.244
(0.363)
0.972*
(0.015)
0.260
(0.221)
1.006
(0.023)
1.941
(1.029)
2.458*
(1.170)
1.000
(0.000)
0.961
(0.326)
1.806
(0.671)
16.06
(33.39)

Good School Facility
Urban
0.510
(0.420)

(1)

0.881
(0.519)
0.983
(0.371)

School Size

Constant

(4)

3.721
(6.463)

1.000
(0.000)
0.796
(0.445)
0.936
(0.885)
0.312
(0.255)

1.132
(0.877)

32.50*
(65.86)

1.000
(0.000)
1.255
(0.401)
2.137*
(0.849)
0.877
(0.723)

Observations
289
287
281
279
339
328
328
318
Notes:
1. All coefficients shown are odds ratios; 2. Std errors given in parentheses; 3. ***p<0.01, **p<0.05, *p<0.10; 4.
DV=teacher perception of satisfaction.
Model 1= Teacher satisfaction on indices of leadership; Model 2=teacher satisfaction on leadership indices with
student/teacher controls aggregated by teacher; Model 3 = satisfaction on leadership with only school controls and
Model 4=Complete model with satisfaction regressed on all available variables.

111

Table 9: Table of Results for Principal Perception of Teacher Satisfaction & Principal Leadership Style (Analysis III - School Level)
HONG KONG (SAR)
INDONESIA
MALAYSIA
VARIABLE NAMES
(1)
(2)
(3)
(4)
(1)
(2)
(3)
(4)
(1)
(2)
(3)
(4)
Transactional Leadership
Principal observation 3.079
1.518
2.421
1.430
5.151
4.774
8.825
7.675
(4.292) (1.781) (3.255) (1.647) (5.892) (5.971) (12.04) (11.76)
Teacher incentives 0.233
0.178
0.189
0.0911
0.788
0.516
0.769
0.528
0.691
0.824
0.687
0.850
(0.272) (0.268) (0.240) (0.158) (0.430) (0.285) (0.441) (0.321) (0.386) (0.460) (0.391) (0.492)
Instructional Leadership
Percent time (yr) on IL 1.016
1.047
1.027
1.056
0.915*** 0.913*** 0.905*** 0.903** 0.999
1.006
1.010
1.013
(0.030) (0.033) (0.033) (0.035) (0.031) (0.031) (0.034) (0.035) (0.027) (0.031) (0.029) (0.031)
Transformational Leadership
Set directions 3.950*** 5.607*** 3.204*** 5.480*** 12.31*** 15.56*** 12.24*** 15.71*** 1.809* 1.828* 2.238** 2.207**
(1.266) (2.087) (1.048) (2.239) (5.915) (7.653) (6.044) (7.715) (0.641) (0.656) (0.823) (0.825)
Develop people 1.135
1.016
1.292
1.156
3.024*** 3.389*** 2.853*** 3.248*** 2.578*** 2.563*** 2.574*** 2.500***
(0.350) (0.328) (0.429) (0.419) (0.819) (0.977) (0.801) (0.987) (0.702) (0.678) (0.699) (0.690)
Community focus 0.479* 0.421* 0.578
0.489
0.823
0.900
0.690
0.800
(0.201) (0.193) (0.219) (0.213)
(0.265) (0.281) (0.234) (0.262)
Language of Test is Spoken at Home
0.978
0.972
1.004
1.005
0.997
0.996
(0.018)
(0.019)
(0.010)
(0.012)
(0.010)
(0.010)
Mothers Education
0.227
0.081
0.200
0.280
0.187
0.443
(0.772)
(0.334)
(0.400)
(0.618)
(0.557)
(1.339)
Teacher Experience (Yrs)
1.045**
1.039*
0.942**
0.947*
0.969
0.967
(0.023)
(0.023)
(0.026)
(0.027)
(0.026)
(0.026)
Female Teacher
0.288***
0.232***
0.442*
0.381*
0.794
0.839
(0.133)
(0.114)
(0.204)
(0.192)
(0.350)
(0.412)
Small Class Size
0.819
0.960
0.300
0.289
0.131
0.132
(0.602)
(0.667)
(0.223)
(0.217)
(0.175)
(0.166)
School Size
1.002
1.003*
1.001
1.001
0.999* 0.999
(0.001) (0.002)
(0.000) (0.001)
(0.000) (0.000)
Good School Facility
1.460
1.330
0.561
0.517
0.305** 0.294**
(0.758) (0.690)
(0.744) (0.769)
(0.154) (0.149)
Urban
1.488
0.924
1.396
1.587
6.582
5.095
(1.069) (0.684)
(1.058) (1.191)
(7.666) (5.972)
Constant
0.600
17.37
0.0564 3.936
4.174
83.36* 2.433
36.68
6.986*** 54.91
18.04*** 75.27
(0.862) (59.22) (0.117) (16.06) (5.327) (201.5) (3.532) (104.0) (4.846) (152.6) (17.59) (212.8)
Observations
195
177
186
169
255
255
236
236
279
273
277
271
Notes:
1. All coefficients shown are odds ratios; 2. Std errors given in parentheses; 3. ***p<0.01, **p<0.05, *p<0.10; 4. DV=principal perception of
teacher satisfaction.
Model 1= Teacher satisfaction on indices of leadership; Model 2=teacher satisfaction on leadership indices with student/teacher controls
aggregated by teacher; Model 3 = satisfaction on leadership with only school controls and model 4=Complete model with satisfaction regressed
on all available variables.

112

Table 9 (cont'd)
COUNTRY NAMES
VARIABLE NAMES
Transactional Leadership

(1)

MONGOLIA
(2)
(3)
(4)

prinobs 0.118** 0.102** 0.104** 0.0628**
(0.106) (0.103) (0.093) (0.071)
tch_incentives 1.372
1.187
1.236
1.391
(0.634) (0.547) (0.579) (0.698)
Instructional Leadership
Percent time (yr) on IL 0.977
0.977
0.967
0.951*
(0.021) (0.021) (0.023) (0.025)
Transformational Leadership
set_directions 3.660*** 3.001*** 3.049*** 2.383**
(1.367) (1.211) (1.198) (1.018)
dev_people 1.631
1.366
1.426
1.005
(0.589) (0.528) (0.556) (0.419)
community_focus 1.502
1.549
1.597
1.838
(0.562) (0.641) (0.666) (0.854)
Language of Test is Spoken at Home
0.991
0.999
(0.029)
(0.025)
Mothers Education
1.250
11.28
(1.530)
(20.63)
Teacher Experience (Yrs)
1.024
1.031
(0.021)
(0.022)
Female Teacher
1.094
0.827
(0.599)
(0.484)
Small Class Size
14.88**
23.89***
(16.30)
(27.27)
School Size
1.000
1.001
(0.000) (0.001)
Good School Facility
Urban
9.700
28.30**
(13.97) (46.41)
Constant
18.48*** 20.60
27.02*** 2.826
(18.27) (63.08) (28.97) (7.503)

TAIWAN
(3)

(1)

(2)

2.597
(2.281)
1.122
(1.270)

2.159
(1.522)
0.915
(1.187)

2.274
(1.943)
1.771
(2.612)

(4)
2.828
(2.520)
1.492
(2.636)

0.997
0.980
0.992
0.988
(0.021) (0.025) (0.031) (0.032)
2.437**
(0.956)
2.642**
(1.162)
0.478*
(0.187)

5.788*** 8.626*** 16.84***
(3.715) (5.741) (14.80)
1.258
2.036* 1.331
(0.473) (0.850) (0.508)
0.621
0.411* 0.510
(0.264) (0.200) (0.297)
1.007
1.007
(0.017)
(0.017)
0.0668
0.207
(0.112)
(0.695)
1.080**
1.099*
(0.039)
(0.053)
1.711
1.836
(0.862)
(1.050)
0.009***
0.020***
(0.011)
(0.027)
1.002*** 1.001
(0.001) (0.000)
0.423
0.670
(0.303) (0.410)
1.151
0.714
(1.251) (1.029)
0.961
6.500
0.253
0.617
(0.922) (16.19) (0.314) (2.111)

Observations
203
187
180
165
241
219
223
203
Notes:
1. All coefficients shown are odds ratios; 2. Std errors given in parentheses; 3. ***p<0.01, **p<0.05,
*p<0.10; 4. DV=principal perception of teacher satisfaction.
Model 1= Teacher satisfaction on indices of leadership; Model 2=teacher satisfaction on leadership indices
with student/teacher controls aggregated by teacher; Model 3 = satisfaction on leadership with only school
controls and model 4=Complete model with satisfaction regressed on all available variables.

113

Table 9 (cont'd)
COUNTRY NAMES
VARIABLE NAMES
Transactional Leadership

(1)

THAILAND
(2)
(3)
(4)

prinobs 4.627
5.794* 4.793
5.435*
(4.470) (5.546) (4.687) (5.443)
tch_incentives 3.417* 8.022*** 3.893* 8.259***
(2.355) (5.356) (2.867) (5.748)
Instructional Leadership
Percent time (yr) on IL 0.988
0.984
0.992
0.984
(0.041) (0.035) (0.045) (0.038)
Transformational Leadership
set_directions 5.980*** 6.150*** 6.508*** 6.275***
(2.812) (2.932) (3.381) (3.156)
dev_people 1.026
0.925
1.038
0.947
(0.352) (0.383) (0.364) (0.393)
community_focus 3.089*** 3.712*** 2.979*** 3.594***
(1.080) (1.376) (1.087) (1.374)
Language of Test is Spoken at Home
0.979**
0.979**
(0.009)
(0.009)
Mothers Education
204.7***
51.08
(395.6)
(142.1)
Teacher Experience (Yrs)
1.076**
1.077**
(0.032)
(0.032)
Female Teacher
0.497
0.499
(0.260)
(0.262)
Small Class Size
0.085***
0.077***
(0.066)
(0.062)
School Size
1.000
1.000
(0.001) (0.001)
Good School Facility
0.420
1.400
(0.359) (1.588)
Urban
0.307
0.550
(0.430) (0.782)
Constant
1.804
0.0571 1.697
0.280
(1.989) (0.139) (2.108) (0.883)

(1)

(2)

USA
(3)

3.075
1.556
(2.782) (1.699)
0.152*** 0.165**
(0.102) (0.145)

8.122**
(8.444)
0.167
(0.183)

(4)
8.522**
(9.057)
0.108*
(0.122)

1.005
1.019
1.033
1.045
(0.021) (0.022) (0.025) (0.030)
4.575*** 5.140*** 6.278*** 7.935***
(1.793) (2.034) (2.305) (3.381)
1.207
0.941
0.901
0.681
(0.391) (0.319) (0.294) (0.217)
2.007
2.236
1.764
1.963
(1.774) (2.507) (1.589) (1.928)
0.942**
0.911***
(0.023)
(0.031)
0.118
0.156
(0.207)
(0.321)
1.011
1.006
(0.015)
(0.015)
2.246
2.051
(1.151)
(0.931)
5.047**
2.191
(3.792)
(1.385)
0.999
0.998**
(0.001) (0.001)
4.565* 6.737**
(3.730) (5.732)
1.305
1.399
(1.091) (1.367)
2.270
577.3** 0.587
5,098**
(2.060) (1,668) (0.687) (19,150)

Observations
289
287
281
279
339
328
328
318
Notes:
1. All coefficients shown are odds ratios; 2. Std errors given in parentheses; 3. ***p<0.01, **p<0.05,
*p<0.10; 4. DV=principal perception of teacher satisfaction.
Model 1= Teacher satisfaction on indices of leadership; Model 2=teacher satisfaction on leadership indices
with student/teacher controls aggregated by teacher; Model 3 = satisfaction on leadership with only school
controls and model 4=Complete model with satisfaction regressed on all available variables.

114

APPENDIX

115

Table A2: Pearson Correlations for Principal and Teacher Perceptions of Teacher Satisfaction
ANALYSIS I – TRTS TEACHER LEVEL: Pearson Correlations for Dependent Variables
HKG
Variable Name

tsatistch

tsatis_teacher

1

tsatis_principal

0.146*

tsatisprin

IDN
tsatistch

tsatisprin

1
1

0.019

MYS
tsatistch

tsatisprin

1

0.196*

1

0.001

MNG
tsatistch

tsatisprin

1

0.061

1

0.277

TWN
tsatistch

tsatisprin

1

0.103

1

0.099

THA
tsatistch

tsatisprin

1

0.205*

1

0

US A
tsatistch

tsatisprin

1

0.084

1

0.145

0.255*

1

0

ANALYSIS II – TRTS SCHOOL LEVEL: Pearson Correlations for Dependent Variables
HKG
Variable Name

tsatistch

tsatis_teacher

tsatisprin

IDN
tsatistch

1

tsatis_principal

0.152*

tsatisprin

MYS
tsatistch

1
1

0.022

0.223*

tsatis_prin

MNG
tsatis_tch

1
1

0

0.043

tsatis_prin

TWN
tsatis_tch

1
1

0.456

0.103

tsatis_prin

THA
tsatis_tch

1
1

0.104

0.202*

tsatis_prin

US A
tsatistch

1
1

0.001

0.084

tsatisprin
1

1

0.145

0.293*

1

0

ANALYSIS III – PRTS SCHOOL LEVEL: Pearson Correlations for Dependent Variables
HKG
Variable Name

tsatistch

tsatis_teacher

tsatisprin

IDN
tsatistch

1

tsatis_principal

0.152*
0.022

tsatisprin

MYS
tsatistch

1
1

0.223*
0

tsatisprin

MNG
tsatistch

1
1

0.043
0.456

tsatisprin

TWN
tsatistch

1
1

0.103
0.104

tsatisprin

THA
tsatistch

1
1

0.202*
0.001

US A
tsatistch

1
1

0.084
0.145

Notes:
1.

*Indicates significance level p<0.05

2.

The dependent variable tsatis_teacher = teacher perception of teacher satisfaction; tsatis_principal=principal perception of teacher satisfaction

116

tsatisprin

tsatisprin
1

1

0.293*
0

1

Table A3: Pearson Correlations for Leadership Indices – Analysis I (Teacher Level)

HKG:

prinobs

prinobs

1

tch_incent

0.1706*

tch_incent

BC4GAPIL

set_directions

dev_people

comm_focus

1

0.006
BC4GAPIL

0.1942*

-0.055

0.0022

0.3928

0.0208

0.0778

-0.1996*

0.7574

0.2475

0.0036

0.0861

-0.0317

0.0824

0.0285

0.2003

0.6377

0.2332

0.6719

0.0185

-0.0195

-0.1599*

0.0626

0.0626

0.784

0.7724

0.0202

0.3524

0.3518

IDN:

prinobs

tch_incent

BC4GAPIL

set_directions

dev_people

prinobs

1

tch_incent

-0.0228

set_directions

dev_people

comm_focus

1

1

1

1

0.6441
BC4GAPIL

set_directions

dev_people

0.1520*

0.0492

1

0.0021

0.3265

0.1474*

0.106*

0.084

0.0039

0.0405

0.1067

0.1432*

0.0798

-0.0114

0.0665

0.0051

0.1236

0.827

0.1953

1

Notes:
*indicates sig p <0.05

117

1

1

Table A3 (cont'd)
MYS:

prinobs

prinobs

1

tch_incent

-0.0265

tch_incent

BC4GAPIL

set_directions

dev_people

1

0.6357
BC4GAPIL

0.1349*

0.0976

0.0154

0.0818

-0.069

0.1565*

-0.1894*

0.2219

0.0056

0.0008

-0.0832

-0.1056

0.0093

0.0072

0.1406

0.0624

0.8707

0.8986

MNG:

prinobs

tch_incent

BC4GAPIL

set_directions

prinobs

1

tch_incent

0.0345

set_directions

dev_people

1

1

1

dev_people

1

0.5715
BC4GAPIL

set_directions

dev_people

0.0453

0.0629

1

0.4585

0.3065

0.0225

0.0251

-0.0352

0.7438

0.7182

0.6128

0.0455

0.0442

-0.1468*

0.0708

0.5086

0.5249

0.0339

0.3034

Notes:
*indicates sig p <0.05

118

1

1

Table A3 (cont'd)
TWN:

prinobs

Prinobs

1

tch_incent

0.1814*

tch_incent

BC4GAPIL

set_directions

dev_people

comm_focus

1

0.0015
BC4GAPIL

-0.0809

-0.0296

0.1665

0.6136

0.2106*

0.1058

0.0957

0.0007

0.0932

0.1311

0.2290*

0.0763

0.0521

0.1182

0.0002

0.2264

0.4122

0.06

0.2158*

-0.0617

0.1185

0.0483

-0.0075

0.0005

0.3283

0.0614

0.4437

0.9053

THA:

prinobs

tch_incent

BC4GAPIL

dev_people

comm_focus

Prinobs

1

tch_incent

0.1455*

set_directions

dev_people

comm_focus

1

1

1

1

0.012
BC4GAPIL

dev_people

comm_focus

0.0491

0.0664

1

0.3972

0.2529

0.1781*

0.174

* 0.0093

0.0023

0.003

0.8748

0.1227*

-0.0025

-0.0475

0.1230*

0.0365

0.9669

0.4197

0.036

Notes:
*indicates sig p <0.05

119

1

1

1

Table A3 (cont'd)
USA:

prinobs

prinobs

1

tch_incent

0.0339

tch_incent

BC4GAPIL

set_directions

dev_people

comm_focus

1

0.3363
BC4GAPIL

set_directions

dev_people

comm_focus

0.1257*

-0.1980*

1

0.0006

0

0.1178*

0.0549

0.1848*

0.0027

0.1637

0

0.1054*

-0.0992*

0.0830*

0.1039*

0.0074

0.0117

0.0376

0.0083

-0.0418

0.0499

0.1142*

0.023

-0.0257

0.2891

0.206

0.0041

0.559

0.5152

Notes:
*indicates sig p <0.05

120

1

1

1

Table A4: Pearson Correlations for Leadership Indices - Analysis II (School Level)
HKG:

prinobs

prinobs

1

tch_incent

BC4GAPIL

0.2483*
0.0001

tch_incent

BC4GAPIL

set_directions

dev_people

1

0.1703*

-0.0797

0.0114

0.2412

0.0437

0.0261

-0.1919*

0.534

0.7105

0.0072

0.0792

-0.0455

0.0735

0.026

0.2587

0.5175

0.307

0.7114

0.0316

-0.0584

-0.1703*

0.0599

0.0572

0.6529

0.4053

0.0173

0.3938

0.415

IDN:

prinobs

tch_incent

BC4GAPIL

set_directions

dev_people

prinobs

1

tch_incentives

-0.0137

set_directions

dev_people

comm_focus

1

1

1

1

0.8165
BC4GAPIL

set_directions

dev_people

comm_focus

0.1504*

0.0402

1

0.0112

0.5032

0.1510*

0.0909

0.0937

0.0133

0.1416

0.132

0.1261*

0.0818

-0.0215

0.0695

0.039

0.1862

0.7301

0.2571

1

Notes:
*indicates sig p <0.05

121

1

1

Table A4 (cont'd)
MYS:

prinobs

prinobs

1

tch_incent

BC4GAPIL

set_directions

dev_people

0.0484
0.4057

0.1008

0.0556

0.0838

0.0705
0.2324
0.0767
0.1934

prinobs

Prinobs

1
0.0329

BC4GAPIL

set_directions

dev_people

1

0.111

MNG:

tch_incent

tch_incent

1

-0.0838

-0.0262

1

0.1575

0.6586

0.1862

* -0.1837

* -0.0092

0.0016

0.0018

0.8765

tch_incent

BC4GAPIL

set_directions

1

dev_people

1

0.5931
BC4GAPIL

set_directions

dev_people

0.0459

0.0556

1

0.4579

0.3723

0.045

0.0391

-0.1304

0.5155

0.5762

0.0612

0.0246

0.0227

-0.0526

0.0907

0.7226

0.7452

0.4515

0.1894

Notes:
*indicates sig p <0.05

122

1

1

Table A4 (cont'd)
TWN:

prinobs

prinobs

1

tch_incent

0.1836*

tch_incent

BC4GAPIL

set_directions

dev_people

comm_focus

1

0.0015
BC4GAPIL

-0.0899

-0.0319

0.1281

0.59

0.2219*

0.0767

0.0445

0.0004

0.2296

0.4893

0.1971*

0.1096

0.095

0.1192

0.0018

0.0856

0.1388

0.0609

0.2144*

-0.0592

0.1307

* -0.0067

0.0471

0.0007

0.354

0.0414

0.9163

0.4603

THA:

prinobs

tch_incent

BC4GAPIL

dev_people

comm_focus

prinobs

1

tch_incent

0.1455*

set_directions

dev_people

comm_focus

1

1

1

1

0.012
BC4GAPIL

dev_people

comm_focus

0.0491

0.0664

1

0.3972

0.2529

0.1781*

0.174

* 0.0093

0.0023

0.003

0.8748

0.1227*

-0.0025

-0.0475

0.1230*

0.0365

0.9669

0.4197

0.036

Notes:
*indicates sig p <0.05

123

1

1

1

Table A4 (cont'd)
USA:

prinobs

prinobs

1

tch_incent

0.0355

tch_incent

BC4GAPIL

set_directions

dev_people

comm_focus

1

0.4661
BC4GAPIL

set_directions

dev_people

comm_focus

0.1087*

-0.1653*

0.0316

0.001

1

0.0572

-0.0759

0.0646

0.2887

0.1588

0.2354

0.0981

0.0424

0.1553*

0.0784

0.0683

0.432

0.0041

0.1455

-0.0263

0.0501

0.1178*

-0.0349

0.0196

0.626

0.3533

0.0301

0.5172

0.7165

1

1

Notes:
*indicates sig p <0.05; Analysis III correlations are very similar to Analysis II

124

1

Table A5: Descriptive Statistics TRTS1 - Teacher Level
HKG

Countries
Dependent Variable

IDN

MYS

MNG

TWN

THA

USA

Mean

SE

Mean

SE

Mean

SE

Mean

SE

Mean

SE

Mean

SE

Mean

SE

0.426

0.495

0.378

0.485

0.625

0.485

0.388

0.488

0.530

0.500

0.500

0.501

0.518

0.500

Principal Observations

0.973

0.162

0.974

0.160

0.982

0.135

0.982

0.133

0.645

0.479

0.900

0.301

0.985

0.121

Teacher Incentives

0.054

0.227

0.387

0.488

0.406

0.492

0.529

0.500

0.056

0.230

0.379

0.486

0.071

0.256

Setting Directions

0.000

0.851

0.000

0.769

0.000

0.869

0.000

0.713

0.000

0.888

-

-

0.000

0.849

Developing People

0.000

0.821

0.000

0.874

0.000

0.768

0.000

0.670

0.000

0.814

0.000

0.839

0.000

0.817

Community Focus
0.000
Instructional Leadership (% time) 19.617

0.687

-

-

-

-

-

-

0.000

0.755

0.000

0.772

0.000

0.711

7.825

25.271

9.293

24.901

11.830

24.162

11.119

25.051

11.938

25.787

11.621

24.302

13.281

90.243

11.829

33.593

37.255

62.666

32.260

95.179

10.854

83.208

17.989

66.790

36.923

89.767

15.552

0.895

0.144

0.911

0.201

0.879

0.123

0.703

0.208

0.767

0.192

0.900

0.173

0.530

0.271

13.033

9.662

12.135

7.585

11.694

8.266

16.283

10.667

12.162

8.469

14.440

9.974

12.896

9.746

0.903

0.296

0.755

0.431

0.809

0.394

0.962

0.191

0.906

0.292

0.977

0.152

0.999

0.035

Proportion Females

0.399

0.491

0.520

0.500

0.730

0.445

0.818

0.386

0.468

0.500

0.652

0.477

0.633

0.482

Teacher Age (Yrs)

37.776

9.692

38.126

8.059

37.094

8.077

38.600

10.997

39.169

8.444

40.394

10.312

41.630

10.821

0.115

0.320

0.092

0.290

0.019

0.136

0.111

0.314

0.048

0.214

0.123

0.329

0.537

0.486

1051.690
School Resources for Math (0-1) 0.734
Urban (0-1)
0.345

175.059

553.197

858.581 1906.202 1214.769 1341.747 1070.423

834.725

402.750

0.443

0.058

0.233

0.426

0.495

0.058

0.235

0.381

0.486

0.141

0.349

0.499

0.500

0.476

0.201

0.401

0.117

0.322

0.088

0.284

0.213

0.410

0.102

0.303

0.185

0.389

Teacher Satisfaction (tch)*
Leadership Indices
Transactional Leadership

Transformational Leadership

Aggregated Student Controls
Language of Test Spoken at
Home (0-1)
Mothers Education (Less
than Secondary) (0-1)
Aggregated Teacher
Controls
Teacher Experience (Yrs)
Teacher Certification (0-1))

Small Class size (0-1)
School Controls
School Size

N
Notes:

182

317

335.066 1428.156

686.501 1413.964

295

174

209

279

1. All means are unweighted and at the teacher level
2. *The DV teacher satisfaction is a binary variable where 1= v.high/high teacher satisfaction and 0=medium/low satisfaction. Tch=teacher reported
3. Analysis is for Math and Science teachers

125

548

BIBLIOGRAPHY

126

BIBLIOGRAPHY
Abu-Saad, I., & Hendrix, V. L. (1995). Organizational climate and teachers’ job satisfaction in a
multi-cultural milieu: The case of the Bedouin Arab schools in Israel. International Journal
of Educational Development, 15(2), 141-153.
Allensworth, E., Ponisciak, S., & Mazzeo, C. (2009). The schools teachers leave: Teacher
mobility in Chicago public schools. Consortium on Chicago School Research Report,
University of Chicago.
Barnes, G., Crowe, E., & Schaefer, B. (2007). The cost of teacher turnover in five school
districts: A pilot study. National Commission on Teaching and America’s Future (NCTAF).
Bass, B. M., & Avolio, B. J. (1994). Improving organizational effectiveness through
transformational leadership. Sage Publications: California, U.S.A.
Bogler, R. (2001). The influence of leadership style on teacher job satisfaction. Education
Administration Quarterly, 37(5), 662-683.
Bogler, R. (2002). Two profiles of schoolteachers: a discriminant analysis of job satisfaction.
Teaching and Teacher Education 18, 665–673.
Boyd, D., Grossman, P., Ing, M., Lankford, H., Loeb, S. & Wyckoff, J. (2011). The influence of
school administrators on teacher retention decisions. American Education Research Journal,
48(2), 303-333. DOI: 10.3102/0002831210380788.
Branch, G.F., Hanushek, E.A., & Rivkin, S.G. (2013). School leaders matter. Education Next,
Winter 2013.
Brown, K.M., & Wynn, S.R. (2013). Finding, supporting and keeping: The role of the principal
in teacher retention issues. Leadership and Policy in Schools, 8(1), 37-63.
Buckley, J., Schneider, M., & Shang, Y. (2005). Fix it and they might stay: School facility
quality and teacher retention in Washington, DC. Teachers College Record, 107, 1107-1123.
Cheng, K., & Wong, K. (1996). School effectiveness in East Asia, concepts, origins and
implications. Journal of Educational Administration, 34(5), 32-49.
Cheong, C. (2000). The characteristics of Hong Kong school principals’ leadership: The
influence of societal culture, Asia Pacific Journal of Education, 20(2), 68-86.
Chui, H.S., Sharpe, F.G., & McCormick, J. (1996). Vision and leadership of principals in Hong
Kong. Journal of Educational Administration, 34(3), 30-48.

127

Dinham, S. (2005). Principal leadership for outstanding educational outcomes. Journal of
Educational Administration, 43(4), 338-356.
Dinham, S. & Scott, C. (2002), “The international Teacher 2000 Project: an international study
of teacher and school executive satisfaction, motivation and health in Australia, England,
USA, Malta and New Zealand”, paper presented at the Challenging Futures Conference,
University of New England, Armidale.
Firestone, W. A., & Pennell, J.R. (1993). Teacher commitment, working conditions and
differential incentive policies. Review of Educational Research, 63(4), 489-525.
Foy, P., & Olson, J. F. (2009). The TIMSS 2007 User Guide for the International Database.
TIMSS & PIRLS International Study Center, Lynch School of Education, Boston.
Grissom, J.A. (2011). Can good principals keep teachers in disadvantaged schools? Linking
principal effectiveness to teacher satisfaction and turnover in hard-to-staff environments.
Teachers College Record, 113(11), 2552-2585.
Grissom, J.A. & Loeb, S. (2011). Triangulating principal effectiveness: How perspectives of
parents, teachers and assistant principals identify the central importance of managerial skills.
American Educational Research Journal, 48(5), 1091–1123,
DOI: 10.3102/0002831211402663.
Guarino, C.M., Santibanez, L., & Daley, G.A. (2006). Teacher recruitment and retention: A
review of the recent empirical literature. Review of Education Research, 76 (2), 173–208,
DOI: 10.3102/00346543076002173.
Hallinger, P. (2011). Developing a knowledge base for educational leadership and management
in East Asia. School Leadership & Management, 31(4), 305-320.
Hallinger, P. (2005). Instructional leadership and the school principal: A passing fancy that
refuses to fade away. Leadership and Policy in Schools, 4(3), 221-239.
Hallinger, P. & Heck, R. (2002). What do you call people with visions? The role of vision,
mission and goals in school improvement, in K. Leithwood, P. Hallinger, G. Furman, P.
Gronn, J. Macbeath, B. Mulford & K. Riley (Eds), The Second International Handbook of
Educational Leadership and Administration (Dordrecht, The Netherlands, Kluwer).
Hallinger, P. & Heck, R. (1996). Reassessing the principal’s role in school effectiveness: a
review of empirical research, 1980-1995, Educational Administration Quarterly, 32(1), 5-44.
Hallinger, P. & Kantamara, P. (2000). Leading at the confluence of tradition and globalization:
The challenge of change in Thai schools. Asia Pacific Journal of Education, 20(2), 24-57.

128

Hariri, H., Monypenny, R., & Prideaux, M. (2012). Principalship in an Indonesian school
context: Can principal decision-making styles significantly predict teacher job satisfaction?
School Leadership & Management: Formerly School Organization, 32(5), 453-471.
Heck, R. H., Larsen, T. J., & Marcoulides, G. A. (1990). Instructional leadership and school
achievement: Validation of a causal model. Education Administration Quarterly, 26(2), 94125.
Hui, S. K., & Cheung, H. (2006). A re-examination of leadership style for Hong Kong schoolbased management (SBM) schools. Asia Pacific Journal of Education, 26(2), 173-187.
Hulpia, H., Devos, D., & Rosseel, Y. (2009). The relationship between the perception of
distributed leadership in secondary schools and teachers' and teacher leaders' job satisfaction
and organizational commitment. School Effectiveness and School Improvement, 20(3), 291317.
Ingersoll, R.M. (2001). Teacher turnover and teacher shortages: An organizational analysis.
Amercian Educational Research Journal, 38(3), 499-534.
Johnson, S. M., Kraft, M. A., & Papay, J.P. (2011). How context matters in high needs schools:
The effects of teachers’ working conditions on their professional satisfaction and their
students’ achievement. Project on the Next Generation of Teachers, Harvard Graduate
School of Education.
Joncas, M. (2008a). TIMSS 2007 sample design. Chapter 5 in J.F. Olson, M.O. Martin, & I.V.S.
Mullis, (Eds.). TIMSS 2007 technical report. Chestnut Hill, MA: TIMSS & PIRLS
International Study Center, Boston College.
Joncas, M. (2008b). TIMSS 2007 sampling weights and participation weights. Chapter 9 in J.F.
Olson, M.O. Martin, & I.V.S. Mullis, (Eds.). TIMSS 2007 technical report. Chestnut Hill,
MA:TIMSS & PIRLS International Study Center, Boston College.
Kloep, M., & Tarifa, F. (1994). Working conditions, work style, and job satisfaction among
Albanian teachers. International Review of Education, 40 (2), 159-172.
Koh, W.L., Steers, R.M., & Terborg, J.R. (1995). The effects of transformational leadership on
teacher attitudes and student performance in Singapore. The Journal of Organizational
Behavior, 16, 319-333.
Ladd, H. (2011). Teachers’ perceptions of their work environments: How predictive of planned
and actual teacher movement? Educational Evaluation and Policy Analysis, 33(2), 235-261.
DOI: 10.3102/0162373711398128
Lee, V.E., Dedrick, R. F., & Smith, J.B. (1991). The Effect of the Social Organization of Schools
on Teachers' Efficacy and Satisfaction. Sociology of Education, 64(3), 190-208.

129

Lee, M., Walker, A., & Chui, L. (2012). Contrasting effects of instructional leadership practices
on student learning in a high accountability context. Journal of Educational Administration,
50(5), 586-611.
Leithwood, K. & Day, C. (2008). The impact of school leadership on pupil outcomes: Editorial.
School Leadership & Management, 28(1), 1-4.
Leithwood, K., & Jantzi, D. (2000). The effects of transformational leadership on student
engagement with school, Journal of Educational Administration, 38(2), 112-129.
Leithwood, K., & Sun, J. (2012). The nature and effects of transformational school leadership:
A meta-analytic review of unpublished research. Educational Administration Quarterly,
48(3), 387-423.
Long, J.S., & Freese, J. (2006). Regression models for categorical dependent variables using
stata. Second edition, Stata Press: College Station, Texas, ISBN 1-59718-011-4.
Lortie, D. C. 2002. Schoolteacher, Chicago: University of Chicago Press.
Marks, H.M., & Printy, S.M. (2003). Principal leadership and school performance: An
integration of transformational and instructional leadership. Educational Administration
Quarterly, 39(3), 370-397.
May, H., Huff, J., & Goldring, E. (2012). A longitudinal study of principals’ activities and
student performance, School Effectiveness and School Improvement: An International
Journal of Research, Policy and Practice, 23(4), 417-439.
Menard, S. (2002). Applied logistic regression analysis 2nd Edition. California, CA: Sage
Publications Inc.
Michaelowa, K., & Wittmann, E. (2007). The cost, satisfaction and achievement of primary
education – Evidence from Francophone Sub-Saharan Africa. The Journal of Developing
Areas, 41(1), 51-78.
Moore, C. M. (2012). The role of the school environment in teacher dissatisfaction among U.S.
public school teachers. SAGE Open, published online 24 February 2012.
DOI: 10.1177/2158244012438888.
Moosung, L., & Hallinger, P. (2012). National contexts influencing principals’ time use and
allocation: economic development, societal culture, and educational system. School
Effectiveness and School Improvement: An International Journal of Research, Policy and
Practice, 23(4), 461-482.
Mowday, R. T., Steers, R.M., & Porter, L.M. (1979). The measurement of organizational
commitment. Journal of Vocational Behavior, 14, 224-247.

130

Neumerski, C. M. (2013). Rethinking instructional leadership, a review: What do we know
about principal, teacher, and coach instructional leadership, and where should we go from
here? Educational Administration Quarterly, 49, 310-347.
OECD (2010), Strong performers and successful reformers in education: Lessons from PISA for
the United States, OECD Publishing.
Ostroff, C. (1992). The relationship between satisfaction, attitudes, and performance: An
organizational level analysis. Journal of Applied Psychology, 77(6), 963-974.
Perie, M., & Baker, D. P. (1997). Job satisfaction among America’s teachers: Effects of
workplace conditions, background characteristics, and teacher compensation (NCES 97-471).
Washington, DC: U.S. Government Printing Office.
Pan, H., & Chen, P. (2011). Challenges and research agenda of school leadership in Taiwan,
School Leadership & Management, 31(4), 339-353.
Pogodzinski, B., Youngs, P., Frank, K. A., Belman, D. (2012). Administrative climate and
novices’ intent to remain teaching. Elementary School Journal, 113(2), 252-275.
Printy, S.M. (2008). Leadership for teacher learning: A community of practice perspective.
Educational Administration Quarterly, 44(2), 187-226.
Robinson, V. Lloyd, C. & Rowe, K. (2008). The impact of leadership on student outcomes. An
analysis of the differential effects of leadership types. Educational Administration
Quarterly, 44(5), 564-588.
Rosenholtz, S. J., & Simpson, C. (1990). Workplace conditions and the rise and fall of teachers’
commitment. Sociology of Education, 63(4), 241-257.
Sargent, T., & Hannum, E. (2005). Keeping teachers happy: Job satisfaction among primary
school teachers in rural Northwest China. Comparative Education Review, 49(2), 173-204.
Schneider, M. (2003). Linking school facility conditions to teacher satisfaction and success.
National Clearinghouse for Educational Facilities: Washington DC.
Schwille, J., Ingvarson, L. C., & Holdgreve-Resendez, R. (2013). TEDS-M 2008 Encyclopedia:
A Guide to Teacher Education Context, Structure, and Quality Assurance in the Seventeen
TEDS-M 2008 Countries.
Scribner, J.P., Sawyer, R.K., Watson, S.T., & Myers, V.L. (2007). Teacher teams and distributed
leadership: A study of group discourse and collaboration. Education Administration
Quarterly, 43(1), 67-100.
Silins, H. (1994). The relationship between transformational and transactional leadership and
school improvement outcomes, School Effectiveness and School Improvement, 5(3), 272-298.
131

Solo, F., Fitzgerald, R., & Jawas, U. (2012). Instructional leadership in Indonesian school
reform: Overcoming the problems to move forward, School Leadership & Management,
32(5), 503-522.
Steiner-Khamsi, G., & Stolpe, I. (2004). Decentralization and recentralization reform in
Mongolia: Tracing the swing of the pendulum. Comparative Education, 40(1), 29-53.
Stockard, J., & Lehman, M. B. (2004). Influences on the satisfaction and retention of 1st-year
teachers: The importance of effective school management. Educational Administration
Quarterly, 40(5), 742-771. DOI: 10.1177/0013161X04268844
Taylor, D., & Tashakkori, A. (1995). Decision participation and school climate as predictors of
job satisfaction and teachers’ sense of efficacy. Journal of Experimental Education, 63, 217–
230.
TIMSS 2011 Encyclopedia: Education Policy and Curriculum in Mathematics and Science,
Volumes 1 and 2. Edited by Mullis, I.V.S., Martin, M.O., Minnich, C.A., Stanco, G.M.,
Arora, A., Centurino, V.A.S., & Castle, C.E. (2012). Chestnut Hill, MA: TIMSS & PIRLS
International Study Center, Boston College. Downloaded from:
http://timss.bc.edu/timss2011/downloads/TIMSS2011_Enc-v1.pdf
U.S. Central Intelligence Agency (2013). Country background information.
Retrieved from https://www.cia.gov/library/publications/the-world-factbook/geos/hk.html
Wahlstrom, K.L., & Louis, K.S. (2008). How teachers experience principal leadership: The roles
of professional community, trust, efficacy and shared responsibility. Educational
Administration Quarterly, 44(4), 458-495.
Walker, A. D., Lee, M., & Bryant, D. A. (2014). How much of a difference do principals make?
An analysis of between-schools variation in academic achievement in Hong Kong public
secondary schools, School Effectiveness and School Improvement: An International Journal
of Research, Policy and Practice. DOI: 10.1080/09243453.2013.875044
Walker, A., Hu, Rongkun, & Qian, H. (2012). Principal leadership in China: An initial review,
School Effectiveness and School Improvement, 23(4), 369-399.
Walker, A., & Qian, H. (2012). Reform disconnection in China. Peabody Journal of Education,
87, 162-177.

132

CHAPTER 3
TEACHER COLLABORATION IN INTERNATIONAL CONTEXTS: DOES
TRANSFORMATIONAL LEADERSHIP PLAY A ROLE?
3.1 Introduction
This paper unpacks the types of teacher collaboration associated with a principal’s
transformational leadership style and then explores whether teacher engagement in a breadth of
collaborative activities is associated with teacher satisfaction. Chapter 2 of this dissertation
explored in detail the association between teachers’ job satisfaction and principals’
transformational leadership in Asia and the United States. This paper will focus on exploring the
association between teacher collaboration and transformational leadership, as well as the
association between collaboration and teacher satisfaction in a multi-country context.
Teacher collaboration is a term which has been previously defined as cooperative actions
that take place primarily in the workplace and that relate to improving instructional practice and
school outcomes (Kelchtermans, 2006). Collaboration has been credited in the literature with
benefits such as providing moral support for teachers (Brundett, 1998; Bolam, McMahon, Stoll,
Thomas & Wallace, 2005; Shulman, cited in Hargreaves, 1994), promoting school effectiveness
(Heck & Hallinger, 2010; Moolenaar, Daly & Sleegers, 2010) and promoting teacher
effectiveness via teacher reflection and teacher learning (Brundett, 1998; Cha & Ham, 2012;
Hargreaves, 1994; Parise & Spillane, 2010; Youngs, 2007). Further, teacher collaboration has
also been viewed by some scholars as an important prerequisite for the effective implementation
of reform initiatives (Coburn, 2001). Since teachers are the agents who interpret and implement
reform it stands to reason that reform initiatives are likely to be more effectively implemented in
schools which have a collaborative climate.
133

Researchers have found that the perceived benefits of collaboration are to a large extent
social and emotional. Collaboration has been linked positively in previous work to teachers’
morale and sense of self-efficacy (Bolam et al, 2005; Demir, 2008; Tschannen-Moran & Hoy,
2007; Ross & Gray, 2006; Wahlstrom & Louis, 2008). Collaborative work helps teachers feel
better about their jobs, their students and themselves, which in turn contributes to greater teacher
commitment and job enthusiasm. Thus, collaborative activity is perceived by many scholars as
being central to the decision-making process in schools as well as being an integral part of
effective schooling. Indeed, it has been suggested that developing and sustaining teacher
collaboration in schools is an important aspect of organizational capacity building because it
enables school staff to perform their tasks more effectively by pooling knowledge and technical
expertise (Cosner, 2009; Printy, 2010; Sergiovanni, 1994).
It is well-established that school leaders play an important role in establishing a school’s
organizational context (Parise & Spillane, 2010; Urick & Bowers, 2013). Despite wide
acknowledgement of the benefits of teacher collaboration in the literature, however, there has
been little systematic effort to understand the effect of leadership style on teacher collaboration
particularly in a multi-country setting. In order to develop a comprehensive understanding of
collaboration in the school context, it is important to go beyond the perceived benefits of
collaboration and try to understand how different types of teacher collaboration vary based on
leadership style, teacher demographics and school organizational characteristics.
As a systematic investigation in this direction, this study attempts to explore the
hypothesis that a principal’s leadership behavior can alter teachers work conditions such that
teachers are empowered to engage in a breadth of different forms of collaboration. Specifically,
I examine the frequency and breadth of collaboration (termed the “coll_index”) within schools

134

situated in Europe, Central America and South America and explore how different types of
collaboration are or are not associated with transformational leadership behavior in diverse
country context. Additionally, my study adds to the body of literature on teacher collaboration
by exploring the association between teacher collaboration and teachers’ job satisfaction, which
has hitherto not received much scholarly attention. Much of the previous literature on
collaboration has been either focused on teacher learning or on student outcomes. Given
previous evidence that the perceived benefits of collaboration are to a large extent social and
emotional I argue that it is important to explore the links between teacher satisfaction, which is
largely an affective reaction to workplace conditions, and teacher collaboration within the
workplace.
3.2 Literature Review
During the past two decades teacher collaboration has captured the attention of scholars
such as Hargreaves (1994), Little (2003), Talbert and McLaughlin (1994), Tschannen-Moran
(2001; 2010) Levine and Marcus (2010) and Cha and Ham (2012) among others. These scholars
posit that school-based teacher collaboration has the potential to stimulate improvement in
teaching and learning, as well as to facilitate effective change within schools. In addition, based
on shared reflections in the workplace, teacher collaboration provides possibilities for new
models of professional development and higher teacher self-efficacy.
In the literature that follows, I first review studies on teacher collaboration, followed by a
review of research on principals’ transformational leadership and studies at the nexus of
leadership and collaboration. Finally, I provide a brief overview of the literature on teacher
satisfaction and collaboration.
3.2.1 Collaboration. The discourse on teacher collaboration is far from homogenous.
Authors such as Hargreaves (1994) and Little (1990) have approached the study of teacher
135

collaboration as a culturally oriented phenomenon in which the beliefs, norms, and values
constructed by teachers are related to their personal and professional relationships. Scholars
such as Achinstein (2002) and Lortie (2002) on the other hand problematize the purpose and
values underlying a given approach to teachers’ work and relationships. This body of work
depicts collaboration as a micro-political activity in which issues of educational goals, role
expectations and power relationships are fought over and where teacher collaboration is viewed
as a tradeoff between professional interdependence and individual autonomy.
Other scholars understand collaboration from the perspective of the social and
professional relationships inside schools and between schools and their environments.
Sergiovanni (1994) for example, posits that teachers experience a sense of interdependency and
mutual obligation in schools where collaboration is fostered. He postulates a theory of
community such that “schools can become places where relationships are familylike, where
space and time resemble a neighborhood and where a code of values and ideas is shared”
(Sergiovanni, 1994, preface, xv-xvi). Lavie (2006) also emphasizes shared goals, values, and
beliefs as a necessary condition for teacher collaboration within a learning community.
Underlying the notion of a professional community is the belief that teachers benefit as
professionals when they have better relationships in the workplace because they are less isolated
and uncertain. Professional communities provide teachers the support they need to collaborate in
sharing their practices and engaging in reflective dialogue regarding curriculum and instruction
(McLaughlin & Talbert, 2001). The term “professional communities” thus exemplifies a
collaborative school context where teachers share norms, values, visions and beliefs concerning
their students, their teaching, and their work within an organizational culture that encourages

136

interdependence among members of the organization (Printy, 2008; Vescio, Ross, & Adams,
2007).
To date, much of the literature on professional learning communities has focused
extensively on impact of collaboration on student achievement. Yasumoto, Uekawa, and
Bidwell (1990) for example, found that when teachers in high schools organize along collegial
lines within problem-solving communities, student achievement growth is intensified. Similarly,
Louis and Marks (1998) documented that the presence of a professional community in a school
was indicative of higher levels of social support for achievement as well as higher levels of
authentic pedagogy.18 More recently, Bolam, McMahon, Stoll, Thomas, and Wallace (2005)
posited that better professional knowledge among teachers participating in professional learning
communities tends to enhance student learning. Their survey data from 393 schools also
suggests an increase in teacher collaboration and an improvement in teacher practice as a result
of working in teacher learning communities. Additionally, Vescio, Ross, and Adam’s (2008)
review on professional learning communities suggests that such communities have a positive
impact on both teaching practice and student achievement. Several of the studies they reviewed
specifically referred to teachers’ use of techniques (developed as a result of teacher participation
in professional learning communities) such as flexibility in classroom arrangements and changes
in the pace of instruction based on student mastery of the content taught. Similar support for the
relationship between professional learning communities and student achievement can be found in
other literature on school communities (Strahan, 2003; Supovitz, 2002; Printy, 2008; Wahlstrom
& Louis, 2008).

18

Authentic Pedagogy is a term developed by Louis and Marks (1998) which implies teaching that promotes higher
order thinking and the development of a depth of knowledge which is intrinsically valuable.

137

Additionally, there is a separate stream of work on teacher collaboration that specifically
focuses on the importance of trust as a resource in schools (Adams & Forsythe, 2013; Bryk &
Schneider, 2003; Cosner, 2009; Daly, 2009; Tschannen-Moran, 2001). In general, this body of
work posits that trust contributes to positive performance culture in schools. Tschannen-Moran
(2001) for example, noted that trust is an essential prerequisite for greater collaboration within
schools. Her study showed that when there was a high level of trust in parents and teachers,
there was a greater likelihood of collaboration among faculty. Similarly, Cosner’s (2009)
qualitative study of 11 high school principals found that collegial trust was an important feature
of principals’ capacity-building work. Recently, Adams and Forsythe (2013) found that mean
math and reading achievement were higher in schools with a stronger culture of collective
faculty trust. A caveat to these findings is substantive variability19 in trust across schools within
the same school district. This large between school variance indicates that schools within a
district vary considerably in their ability to generate trust among faculty, parents and students. It
is therefore quite possible that variables not accounted for in this work may be mediating the
trust-achievement relationship.
3.2.2 Transformational Leadership & Collaboration. The concept of transformational
leadership was originally developed in the business literature as a means of transforming
organizations (Bass & Avolio, 1994; Burns, 1978) and was applied to the context of schools as a
strategy to support reform (Leithwood & Jantzi, 2005). In contrast to transactional leadership in
which leaders attend to managerial tasks and closely monitor staff, transformational leadership
according to Bass and Avolio (1994) consists of the “four I’s,” which include individualized
consideration (the leader creates a supportive climate and motivates workers to achieve higher

19

Adams and Forsythe (2013) found that around forty-five percent of the variability in collective faculty trust is
between schools.

138

potential), intellectual stimulation (the leader provides opportunities for creativity and staff
development), inspirational motivation (the leader builds consensus for a shared vision and
mission), and idealized influence (the leader acts as a role model and engages in community
building). These four components of transformational leadership guide the restructuring of
organizations for increased effectiveness. Leithwood and Sun (2012) further extended the
conceptual framework provided by Bass and Avolio (1994) to include measures such as holding
high performance expectations, strengthening the school culture, building collaborative
structures to encourage staff participation in school decisions, providing a community focus and
improving the instructional program.
Scholars have found that transformational leadership restructures and prepares schools
for an increase in shared leadership with improved opportunities for innovation and change
(Moolenaar et al, 2010; Printy, 2003). Additionally, prior research indicates that the more
closely connected principals are to their teachers, the more willing teachers are to invest in
change and new practices (Moolenaar et al, 2010). Schools where the teachers feel more
effective are more likely to be schools where teachers share beliefs and values consistent with
the central aims of the school (‘inspirational motivation’- an important aspect of transformational
leadership) and where they feel valued and respected (the notion of ‘individualized
consideration’). Thus, it may be argued that transformational leadership seeks to establish the
kind of school climate necessary for teacher collaboration to flourish. This leadership style is
therefore, an appropriate choice in a study of teacher collaboration.
Much of the leadership literature specifically examines the influence of principal
leadership on student achievement. A major contribution of this research is the strong and
significant indirect relationships (of efficacy, community, and trust) which mediate school

139

leadership and student learning (Dumay, Boonen, & Van Damme, 2013; Goddard, Miller,
Larson & Goddard; 2010; Heck & Hallinger, 2010; Robinson, Lloyd, & Rowe, 2008; Supovitz,
Sirinides & May, 2010; Youngs & King, 2002; Waters, Marzano, & McNulty, 2003; Witziers,
Bosker & Kruger, 2003). Waters et al. (2003) for example, identified the fostering of shared
beliefs and a sense of community and cooperation to be one of the most significant leadership
predictors of student learning outcomes. In their meta-analysis, Robinson et al. (2008)
concluded that instructional leadership had an impact on student achievement that was three to
four times that of transformational leadership. The work of Heck and Hallinger (2010) provides
additional support for the link between collaborative leadership and schools’ organizational
capacity to increase student outcomes.
To date, there has been comparatively less focus on the direct effect of leadership (and
specifically transformational leadership) on teacher collaboration. Gumus, Bulut and Bellibas
(2013), and Duyar, Gumus and Bellibas (2013) have researched the association between teacher
collaboration and principals’ instructional and administrative leadership. However, these studies
are limited geographically in that they are single country studies that examine the association for
Turkey only. It is likely that the association is different for countries in Europe. Further,
Gummus et al. (2013) explore only administrative and instructional leadership styles, thus
leaving room for further analysis of these relationships.
This study builds on extant literature pertaining to principal leadership in two ways.
First, it explores the direct effect of transformational leadership on different types of teacher
collaboration. Second, it uses principal reported measures of transformational leadership rather
than teacher reported measures. Little research has been done to examine principal perceptions
of their own leadership practice and how those practices generate the conditions necessary for

140

the growth of teacher collaboration controlling for school context and school climate (Urick &
Bowers, 2013). Bowers (2013) argues that principal perceptions of leadership are important
because principal perceptions and behaviors determine the extent to which school leaders shape
the school environment and effect organizational change. Principals can also be instrumental in
helping teachers improve their practice by supporting them and establishing an orderly school
climate and through a cohesive schools’ vision and mission.
3.2.3 Teacher Satisfaction and Collaboration. Organizational psychologists have long
maintained that work attitudes drive employee behaviors and performance (Judge, Bono,
Thoresen & Patton, 2001). If we think of the school as a workplace then teacher satisfaction
(work attitude) should be associated with teacher collaboration (employee behavior). There is
emerging support in the scholarly community for the notion that school cultures promoting
teacher collaboration are positively associated with teacher satisfaction (Johnson, 2006; Kraft &
Papay, 2011; Ladd, 2011; Liu & Ramsey, 2008). Johnson and Birkland (2003) for example,
found that teachers were more likely to stay in schools with “integrated professional cultures”
organized around collaborative efforts. This may be because teacher collaboration affords
teachers the opportunity for professional sharing that may not otherwise occur (Woods &
Weasmer, 2004).
Despite the emerging literature on the connection between teacher collaboration in
schools and school improvement, there has been little systematic effort to understand the
association between collaboration and teacher satisfaction. Additionally, there has been little
effort to use large scale nationally representative data to explore teacher collaboration in the
workplace for more generalizable results. This analysis attempts to address this gap in the
literature by analyzing the association between satisfaction and a breadth of collaborative

141

activities across five countries participating in OECD’s Teaching and Learning International
Survey (TALIS 2008).
3.3 Theoretical Framework
Since the purpose of this paper is to analyze the interrelationships between teacher
collaboration, transformational leadership and teacher satisfaction, I draw on the literatures on
school effectiveness and school climate to provide theoretical underpinnings to this work. The
school effectiveness discourse on collaboration is an appropriate choice because it highlights the
centrality of the principal in developing a shared vision as well as developing consensus in
practices and expectations. These are also key aspects of transformational leadership (Bass &
Avolio, 1994; Leithwood & Sun, 2012).
There is also a body of literature which examines principal teacher interaction within
schools, in which the importance of a clear, shared sense of vision is emphasized in the context
of a strong organizational culture that holds high expectations for student and teacher
performance, focuses on academic learning, and encourages discipline and order within the
school (Barnett & McCormick, 2004; Price, 2012). Supportive principal leadership is a key
component of this perspective. The underlying assumption is that the principal’s actions set the
tone of the school and shape the organizational conditions in which teachers work. Studies that
adopt this theoretical stance view leadership as a powerful predictor of teacher effectiveness.
In sum, school leaders who support knowledge sharing among teachers and create
internal structures that promote collaboration are most effective at fostering change within their
schools (Bryk & Schneider, 2002; Parise & Spillane, 2010; Youngs & King, 2002). Based on
this understanding of the relationship between leadership and collaboration, I put forward two
hypotheses:

142

Hypothesis 1: Teachers’ collaborative interaction with other teachers is more likely in
schools with a more positive organizational climate where the principal chooses to exercise
transformational leadership.
Hypothesis 2: Teachers who are engaged in a variety of collaborative experiences are
more satisfied with their jobs than those that do not engage in different types of collaboration.

Collaboration

Teacher
Satisfaction

Transformational
Leadership

Figure 3: Conceptual Interrelationships between Transformational Leadership, Teacher
Collaboration and Teachers’ Job Satisfaction

Figure 3 is a conceptual diagram for the analysis proposed. Although the arrows show
hypothesized directionality this should not be mistaken for causality because the analysis
proposed is purely correlational in nature and indicates associations rather than causal
relationships.
The remaining sections of the paper sequentially examine the country context, research
questions, data and methods. This is followed by a summary of results and discussion of key

143

findings. Additionally, I provide a detailed look at Mexico’s country context given the scale of
education reform there and the breadth of work undertaken by the OECD in that country. I also
discuss briefly the policy implications and limitations of this work.
The following section provides a brief description of country location, educational
background and teacher satisfaction. The purpose of this information is to help the reader
contextualize the results of the within country regressions in the analyses to follow.
Comparative education analyses have traditionally focused on geographic entities as units of
comparison (Bray, Adamson & Mason, 2007). The five countries analyzed in this paper are
Denmark, Poland, Brazil, Mexico and Hungary. My rationale for selection of this subset of
countries from TALIS 2008 was to have sufficient variability in terms of geographic location and
the levels of teacher satisfaction in the countries of interest. Thus, two out of the five countries
selected are in Central Europe (Hungary and Poland), one is located in Northern Europe
(Denmark), one is in North America (Mexico) and the last is South America (Brazil). In three of
these countries – Brazil, Mexico and Hungary teachers report lower levels of job satisfaction
than in Denmark and Poland. Yet these five countries also have certain similarities. All the
countries selected have literacy levels above 90 percent. Similarly, across four of the five
countries selected the level of educational expenditure is quite similar (around five percent of the
national budget). The level of educational expenditure in Denmark is slightly higher
(approximately 8 percent of the budget).
3.4 Country Context20
3.4.1 Denmark

20

Teacher information on all countries is obtained from TALIS 2008 overview of country results at the following
URL: http://www.oecd.org/edu/school/oecdsteachingandlearninginternationalsurveytalisbriefingnotesandoverviewsofcountries.htm

144

3.4.11 Background. Denmark is located in Northern Europe and is a member of the
European Union. Once the seat of the Vikings, Denmark has evolved into a prosperous nation.
Danish is the primary language spoken though English is a predominant second language. Danes
enjoy a high standard of living and the Danish economy is characterized by extensive welfare
measures and an equitable distribution of income as indicated by a Gini index of 24.7 (2011)21.
There is a high literacy rate (99%) in Denmark and the educational expenditure in Denmark is
8.9% of the GDP, which is the highest among the countries selected for analysis.
3.4.12 Teacher Characteristics. In Denmark, teacher satisfaction appears to be above
the TALIS average22 for all countries participating in TALIS 2008. Based on reports from
teachers of lower secondary education and the principals of their schools, on average, teachers in
Denmark view the disciplinary climate in their schools somewhat negatively. However, the
percentage of lesson time lost due to disruptive student behavior or administrative
responsibilities is relatively low as compared to the other TALIS countries.
Approximately 75% of teachers in Denmark participated in professional development
activities during the survey period of 18 months. Teacher experience in Denmark is a little
above average with 39% of teachers working for twenty years or more as compared with the
TALIS average of 36%.
3.4.2

Poland

3.4.21 Background. Poland is located in Central Europe. Poland joined the European
Union in 2004. It has also transformed into a democratic market economy. In the 1990’s
Poland’s economy was considered one of the most robust in Central Europe. The official

21
22

Gini index for all countries taken from https://www.cia.gov/library/publications/the-world-factbook/
The score for job satisfaction (as reported by TALIS) represents the extent of agreement on average with the
statement “All in all I am satisfied with my job”, where strongly agree=4; agree=3; disagree=2 and strongly
disagree=1. The TALIS average is between 3.1-3.2

145

language of the country is Polish with 97.8% of the population speaking it. The Gini index for
Poland is 34.1 (2009) so it is relatively more equal than other countries like Mexico and Brazil.
Poland has a literacy rate that matches that of Denmark (99.7%). Educational expenditures in
Poland however, constitute 5.10% of the GDP which is somewhat lower than Denmark.
3.4.22 Teacher Characteristics: In Poland, teacher satisfaction is approximately the
same as the TALIS average (which is between 3.1-3.2 on a scale from 1-4 where 4 = strong
agreement with the statement that the teacher is very satisfied). Teachers in Poland have a fairly
positive view of the classroom disciplinary climate. Like Denmark, the percentage of lesson
time lost due to disruptive behavior or administrative demands is relatively low compared to
other countries in TALIS.
Teacher participation rate in professional development over the 18 months of survey data
collection was approximately the same (90%) as the TALIS average of 89%. Teacher
experience (defined as 20 or more years of experience) in Poland is slightly below the average
found in the TALIS countries (31% as compared to 36%). Notably, Poland has a relatively
higher degree of professional collaboration than the other TALIS countries.
3.4.3

Brazil

3.4.31 Background. Brazil is located in South America and is considered a leading
economic power and regional leader in South America. Portugese is the official language of the
country and is widely spoken. The country is characterized by a highly unequal income
distribution. The Gini index in Brazil is 51.9 (2009) and is the highest among the countries
selected for analysis. Educational expenditure in Brazil constitutes 5.6% of the GDP, similar to
that of Mexico.

146

3.4.32 Teacher Characteristics: In Brazil, teachers’ job satisfaction is lower than the
TALIS average. On average, teachers in Brazil have a negative opinion of the classroom
disciplinary climate as compared to the other TALIS countries. The percentage of lesson time
lost to disruptive student behavior or administrative commitments is the highest in Brazil of all
TALIS countries.
Around 83% of teachers in Brazil participated in professional development activities in
the survey period as compared to the TALIS average of 89%. The percentage of teachers in
Brazil who have 20 or more years of experience is well below the TALIS average of 36%. Only
19% of Brazilian teachers have this level of experience.
Notably, the extent of instructional leadership reported is strongest in Brazil compared
with the other TALIS countries. Teacher participation in professional development forms an
important component of teacher appraisals in schools where instructional leadership is practiced.
3.4.4

Mexico

3.4.41 Background. Mexico was a Spanish colony until the nineteenth century. 92.7%
of the population speak Spanish which is the primary language. Ongoing economic concerns
include an inequitable distribution of wealth as indicated by a relatively high Gini index of 48.3
(2008). Educational expenditure in Mexico is 5.1 % of the GDP.
3.4.42 Teacher Characteristics. Teachers’ job satisfaction was lower than the TALIS
average. Teachers in Mexico related their overall job satisfaction not to beliefs about teaching
but to other factors such as classroom climate, teacher student relations and their self-efficacy.
In Mexico, teachers have a relatively good opinion of the disciplinary climate in classrooms yet
correspondingly report a lower amount of time spent teaching. Notably, a large proportion of
instructional time in Mexico is devoted to administrative activities.

147

In Mexico, 92% of teachers report participation in professional development during the
18 months preceding the survey as compared to the TALIS average of 89%. Approximately
two-thirds of the teachers in Mexico have ten or more years of experience. In terms of teacher
collaboration, Mexico appears to have higher than average professional collaboration.
3.4.5

Hungary

3.4.51 Background. Hungary is located in Central Europe. The country embarked on a
program of economic liberalization in 1968 followed by the adoption of a free-market economy
in 1990. 84.6 percent of the population in Hungary speak Hungarian. Income distribution in
Hungary is relatively more equitable and the Gini index for Hungary is in the same range as that
for Denmark (24.7 – 2009).
3.4.52 Teacher Characteristics. Teachers in Hungary have a rather positive view of
classroom disciplinary climate. Additionally, the percentage of time lost due to disruptive
student behavior or administrative responsibilities is among the lowest. Despite this, teacher
satisfaction in Hungary is low compared to the other TALIS countries.
The percentage of teachers in Hungary participating in professional development
activities during the survey period of 18 months is approximately the same as the TALIS average
(87% versus 89%). Hungarian teachers have more years of experience than the TALIS average
with 47% of teachers who have 20 or more years of experience compared with the TALIS
average of 36%.
Notably, professional collaboration is relatively more frequent in Hungary as compared
with other TALIS countries.
It is evident from the above review that the five countries chosen for analysis vary with
respect to geographical location and the teachers in these countries display varying levels of

148

satisfaction with their jobs. There is also variation in terms of equitable distribution of wealth
with the European countries being more equitable than either Brazil or Mexico. Nevertheless,
these countries all display similar levels of educational expenditure and a high literacy level of
90 percent or greater. Table 10 on the next page provides a quick summary of this information
on country context
3.5 Research Questions
The research questions are briefly stated as follows:
1. Is frequent teacher collaboration across a breadth of collaborative activities among teachers
associated with transformational leadership in a multi-country context controlling for the
effect of teacher demographics, principal demographics and school climate?

2. What are the kinds of collaborative activity associated with different dimensions of
transformational leadership?

3. Is frequent participation in a breadth of collaborative activities associated with teachers’ job
satisfaction in a multi-country context?

.

149

Table 10: Summary of Country Context - TALIS 2008

Size
Location
Gini Index
Education Expenditure
Literacy
Teacher Satisfaction

Teacher Professional Development
(%)
% Teacher Experience (20yrs+)

DNK

POL

BRA

MEX

HUN

43,094
N.Europe
24.8
8.70%
99%

312,685
C.Europe
34.1
5.10%
99.7%

8,514,877
S. America
51.9
5.60%
90.40%

1,964,375
N.America
48.3
5.30%
93.50%

93,028
C. Europe
24.7
4.90%
99%

> Avg

Avg

< Avg

< Avg

< Avg

76%

90%

83%

92%

87%

39%

31%

19%

N/A

47%

Sources:
http://www.oecd.org/edu/school/oecdsteachingandlearninginternationalsurveytalisbriefingnotesandoverviewsofcountries.htm
https://www.cia.gov/library/publications/the-world-factbook/geos/hu.html

150

3.6 Data
3.6.1

Sample design & weights

The OECD’s Teaching and Learning International Survey (TALIS, 2008) is the first
international survey to focus on teachers working conditions and the role of school leadership.
As such, this data is well-suited to analyze the interrelationships between teacher satisfaction,
teacher collaboration, and a principal’s transformational leadership in an international context.
TALIS 2008 is a cross sectional, observational, and non-experimental study. Information is
available from more than 4,000 schools and 70,000 teachers. The target population for TALIS is
lower secondary school teachers and principals. A minimum of 200 schools per country and 20
teachers per school were sampled. Exclusion restrictions include schools catering exclusively to
special needs children or to adult education.
All the data are derived from random samples of schools and teachers. Consequently, the
results of all analyses using this data hold not only for the sampled schools and teachers but for
the country as a whole. To make correct inferences using the TALIS data, however, it is
necessary to take into account the complex sampling structure of the data. I do this by using
Stata’s svyset command together with the brr option to generate appropriate standard errors in all
the analyses.
The sampling plan for TALIS is a stratified two-stage probability sampling design. This
means that the teachers (who constitute the Secondary Sampling Unit - SSU) were randomly
selected from a list of teachers within each of the randomly selected schools (which constitute
the Primary Sampling Unit - PSU). Each school is regarded as a cluster and all teachers are
nested within these clusters. The use of appropriate sampling weights compensates for the
disproportional selection probabilities of schools and teachers. Because the analysis proposed

151

requires data from the school as well as teacher level, school level data are added to the teacher
data such that the school information becomes an attribute of the teacher. A detailed description
of all the variables used in the analysis follows.
3.6.2

Description of Dependent Variables

For the first research question my dependent variable is an index measuring the
frequency of teacher collaboration across different types of collaborative activity. The index was
constructed in two steps. First, each of eight different collaborative activities (numbers 5-12 in
table 11) were recoded as binary variables (where 1= teacher participation in collaborative
activity 3-4 times/year or more frequently and 0= teacher participation less than 3-4 times/year).
Next these eight measures of collaboration were averaged for each teacher using Stata’s
“rowmean” command. This produced fractional numbers between 0 and 1 (inclusive of 0 and 1)
for each teacher, such that a fractional number closer to 1 implies more frequent collaborative
activity for that teacher across a breadth of collaboration variables. The index of collaboration
thus measures frequent participation in a variety of different types of collaboration. Notably, the
index of collaboration contains measures of collaboration that pertain to collaboration within the
school. The collaboration variables (numbers 1-4 in table 11) measuring participation in a PD
network, dialogue with teachers were not included in this index because they are binary Y/N
variables that simply measure participation or lack thereof in these collaborative activities. They
do not measure frequent participation in a collaborative activity (unlike variables 5-12 in table
11).
For my second research question pertaining to the types of collaborative activity
associated with dimensions of principal leadership, the dependent variable differs by type of
collaborative activity. Thus, for each of five countries analyzed, I ran twelve within country

152

logistic regressions where the dependent variable for each regression was a different type of
collaborative activity. These twelve collaborative activities as well as their recodes are listed in
table 11.
My third research question examines the association between teacher collaboration and
teachers’ job satisfaction. The dependent variable for this question is binary (1= teacher
satisfaction; 0=otherwise) and reflects teacher agreement with the statement “I am satisfied with
my job.” Teacher responses to the satisfaction question on the TALIS questionnaire were
organized on a scale going from ‘Strongly Disagree’ to ‘Strongly Agree.’ However, there were
very few observations in the ‘Strongly Disagree’ category so the variable was recoded as binary
by combining the ‘Strongly Disagree’ and ‘Disagree’ categories as 0 and the ‘Strongly Agree’
‘Agree’ categories as 1.
3.6.3

Description of Key Independent Variables

For research question 1, the key independent variables are standardized factor scores
named ‘factor A,’ ‘factor B,’ and ‘factor C,’ indicative of a principal’s transformational
leadership style. These scores were obtained through an explanatory factor analysis (EFA).
Table 12 lists all the variables that were used in the EFA to create the transformational leadership
factor scores. These variables are consistent with Leithwood & Sun’s (2012) conceptualization
of transformational leadership, which is based on the Nature of School Leadership survey
(NSL).23 The Kaiser criterion (which uses the decision rule of eigenvalues greater than 1) as well
as, a visual examination of the scree plot in each country were used to decide how many factors

23

Related leadership practices – Management by exception and Contingent reward were excluded from the factor
analysis because the variable BCG20F measuring the extent to which teacher bonuses were tied to evaluations
has 38% missing data for Denmark. Similarly the variable BCG15C (principal observation of classrooms where
1=very often and 0=not very often) has only 18% in the ‘very often’ category.

153

Table 11: Types of Teacher Collaboration
No. Variable Name

Variable Desciption /Recode

1

pd_network

Teacher participation (over past 18 mo) in a network put together specifically for teacher professional development
(1=Y; 0=N)

2

collab_research

Teacher participation in collaborative research on a topic of interest (1=Y; 0=N)

3

Teacher participation in mentoring, peer observation, and coaching as part of a formal school arrangement (1=Y;
mentoring_coaching 0=N)

4

dialogue_tchg

Teachers engage in informal dialogue on how to improve teaching (last 18 mo) (1=Y; 0=N)

5

coll_textbook

Frequency with which teachers discuss & decide on selection of textbooks (1=3-4 times/yr or more; 0=less than 34 times/yr)

6

tchg_materials

Frequency with which teachers exchange teaching materials with colleagues (1=3-4 times/yr or more; 0=less than
3-4 times/yr)

7

team_conf

Frequency with which teachers attend team conferences for age group taught (1=3-4 times/yr or more; 0=less than
3-4 times/yr)

8

discuss_learning

Frequency with which teachers engage in discussions about learning (1=3-4 times/yr or more; 0=less than 3-4
times/yr)

9

team_tchg

Frequency with which teachers teach jointly as a team in the same class (1=3-4 times/yr or more; 0=less than 3-4
times/yr)

10

obs_feedback

Frequency with which a teacher observes another teacher's class and offers feedback ((1=3-4 times/yr or more;
0=less than 3-4 times/yr)

11

joint_activities

Frequency with which a teacher engages in joint activities with other teachers across different classes and age
groups (1=3-4 times/yr or more; 0=less than 3-4 times/yr)

hw_practice

Frequency with which teachers engage in homework practice across subjects ((1=3-4 times/yr or more; 0=less than
3-4 times/yr)

12

154

Table 12: List of Variables Used in EFA for Transformational Leadership Indices –
TALIS 2008 24
No. Leadership Practices

Variable Name

1.

staff_goals

Inspirational Motivation

The principal defines goals to be accomplished by
the staff of the school
The principal ensures clarity concerning the
responsibility for coordinating the curriculum

curr_clarity
2.

3.

Hold High Performance
Expectations

Provide Individual
Support

Variable description25

monitor_stuwork

The principal monitors student work

enforce_goals

The principal ensures that teachers work according
to the school's educational goals

discuss_probs

The principal takes initiative to discuss problems
with teachers when a teacher has problems in
his/her classroom

Induction

The principal provides for a formal induction
process for a beginning teacher at his school

4.

Provide Intellectual
Stimulation

upgrade_skills

The Principal informs teachers about opportunities
to upgrade their skills

5.

Strengthening School
Culture

disruptive_class

Principal pays attention to disruptive behavior in
classrooms
principal and teachers solve classroom problems
together
Principal ensures that the pd activities of teachers
are in accordance with teaching goals of school
An important part of my job is to create an orderly
atmosphere in the school.

cl_issues
pd_schgoals
Schorder
6.
7.

8.

Building Collaborative
Structures
Engaging Parents

sch_devplan

Focus on instructional
development

tch_skills

Principal and teachers work jointly on goals and a
school development plan
The principal feels it is an important part of his job
to present new ideas to parents in a convincing
way

parent_outreach

Principal gives teachers suggestions as to how to
improve their teaching
Principal ensures a task-oriented atmosphere is
fostered in the school

stimulate_task
Prinobs

Principal observes instruction in classrooms

24

All variables for EFA were selected based on Leithwood and Sun’s (2012) conceptualization of transformational
leadership.
25
All variables are principal reported measures of leadership practices and are therefore at the school level.

155

should be retained. Additionally, a cutoff of .30 for the factor loadings was used for inclusion.
Three factors were identified via this method – factor A, factor B, and factor C. Regression
factor scores were then computed for each factor using STATA’s ‘predict’ command.
The factor scores (A, B, & C) thus identified did not all load for each country. However,
each factor loaded on at least two of the countries in the analysis. For example, factor A loaded
on Denmark, Poland and Brazil. Factor B loaded on Denmark, and Mexico. Factor C loaded on
Poland and Hungary. Individual factor loadings for each country as well as the percent of
variance explained by each factor are shown in table 13 below.
Table 13: Factor Loadings for Transformational Leadership Indices - TALIS 2008
DNK
Variable Names
pd_schgoals
enforce_goals
upgrade_skills
schorder
stimulate_task
sch_devplan
parent_outreach
discuss_probs
disruptive_class
cl_issues
Eigenvalue Factor
Proportion Variance
Cumulative Variance
N

A
0.58
0.72
0.48

POL
B

A
0.80
0.81

C

BRA
A
0.62
0.62

0.45
-0.38
-0.64
0.42

1.76
0.55

1.01
0.32
0.87
112

MEX
B

HUN
C

0.49
-0.68
-0.48
0.58
0.40
0.69
1.41
0.36

1.61
0.41
0.77
161

1.87
0.63
0.63
329

1.88
0.54
0.82
179

0.66
0.47
0.61
1.74
0.53
0.82
171

Table 13 shows that factor A consists of three variables – pd_schgoals (The principal ensures
that teacher PD is in accordance with th e teaching goals of the school), enforce_goals (The
principal ensures that teachers work according to the school’s educational goals), and
upgrade_skills (The principal informs teachers about the possibilities for updating their
knowledge and skills). These three variables appear to measure staff development consistent
156

with school goals. Factor B consists of four variables that appear to pertain to the principal’s
role in establishing a positive school climate. Factor B includes the variables schorder (An
important part of my job is to create an orderly atmosphere in the school), stimulate_task (The
principal ensures a task-oriented atmosphere is fostered in the school), sch_devplan (The
principal and teachers work jointly on goals /a school development plan), and parent_outreach
(The principal feels it is an important part of his job to present new ideas to parents in a
convincing way). Factor C consists of three variables that are all classroom focused. These
include discuss_probs (The principal takes initiative to discuss problems with teachers when a
teacher has problems in his/her classroom), disruptive_class (Principal pays attention to
disruptive behavior in classrooms), and cl_issues (principal and teachers solve classroom
problems together). Factor C appears to measure the principal’s role in maintaining order and
discipline in the school.
Table 14 below shows standardized reliability coefficients for the three factors indicative
of transformational leadership. Cronbach’s Alpha for these factors is between 0.57 and 0.73
across all the countries.
Table 14: Scale Reliability Coefficient (Cronbachs Alpha) for Factor Scores of Transformational
Leadership – TALIS 2008
Factor A

Factor B

Factor C

DENMARK

0.6339

0.5711

-

POLAND

0.7293

-

0.5955

BRAZIL

0.6918

-

-

MEXICO

-

0.6075

-

HUNGARY

-

-

0.6135

For research question 2 the key independent variables are the same as for research
question 1, namely the transformational leadership factor scores. The key independent variable
157

for question 3 is the index of teacher collaboration constructed as the dependent variable in
question 1.
3.6.4 Description of Controls
School controls include principal demographic variables such as principal gender which
is recoded as a binary variable where 1=female; 0=male and principal experience in years
recoded as a binary variable such that 1=principal experience greater than or equal to 10 years
and 0=principal experience less than 10 years. The variable ‘public’ is a binary variable that is
=1 if the school is a public school and it is = 0 if the school is a private school. ‘langdiff’
represents the percentage of students whose first language differs from the language of
instruction such that 1=less than 10%; 2=more than 10% but less than 20%; 3=20% or more but
less than 40%; 4=40% or more but less than 60%; 5=60% or more. ‘langdiff’ is calculated by
TALIS at the school level as the mean of these response categories rather than the percentages
that they are meant to represent. ‘hpareduc’ is a measure of parent education which has been
recoded as a binary variable such that 1= 60% or more of the class has at least one parent who is
educated to the level of ISCED 5 (higher secondary) or more.
Additionally, there are some school climate controls such as ‘lackpers’, ‘autcurr’,
BTG41B and BTG08C. The index ‘lackpers’ is constructed by TALIS to indicate a lack of
school personnel. Component variables in ‘lackpers’ include lack of teachers (BCG29A);
laboratory technicians (BCG29B); instructional support personnel (BCG29C) and other support
personnel (BCG29D). A high value on this index is indicative of high levels of inadequacy with
respect to support staff in the school. The index ‘autcurr’ was derived from a principal
component analysis by TALIS and consists of three items measuring teacher autonomy with
respect to curriculum. These items are choosing which textbooks are to be used (BCG31);

158

determining course content (BCG31K) and deciding which courses are offered (BCG31K).
Higher values on the index indicate relatively higher levels of school responsibility in this area
(The index ranges from approximately -1 to 1 across the countries analyzed). Cronbach’s Alpha
is acceptable or good for all measures of autonomy (TALIS 2008, Technical Report).The
variable BTG41B is a continuous variable indicating the percentage of class time spent
maintaining discipline in the classroom. The variable BTG08C is a continuous variable
indicating the number of hours spent by teachers per week in administrative duties and clerical
paperwork. The rationale for inclusion is the reasoning that more time spent on clerical work
implies less time available to the teachers for collaborative work.
There are in addition a couple of demographic teacher controls such as ‘teacher gender’
(tfem) which equals one if teacher is female and 0 otherwise and ‘teacher experience’ which is
recoded such that the texps =1 if the teacher has over 10 years of experience and 0 otherwise.
The variable ‘tefficacy2’ reflects teacher concurrence with the statement “If I try really hard I
can make progress with even the most difficult and unmotivated students.” This variable was
recoded as binary (1=SA/A; 0=D/SD).
Table 15 provides a complete list of all of the variables used in the analyses to follow as
well as the respective levels (teacher or school) at which they were administered in the TALIS
survey. Table 16 shows descriptive statistics for the same.

159

Table 15: List of Variables Used in Analyses - TALIS 2008
VARIABLE NAME

DESCRIPTION

LEVEL

DV:
tsatis

Teacher Satisfaction with Job (1= satisfied)

Teacher

Summative index of 8 collaboration variables

Teacher

Teacher participation (over past 18 months) in a network
put together specifically for teacher professional
development

Teacher

collab_research

Teacher participation in individual or collaborative
research on a topic of interest

Teacher

mentoring_coaching

Teacher participation in mentoring/peer observation and
coaching as part of a formal school arrangement

Teacher

dialogue_tchg

Did you engage in informal dialogue with colleagues on
how to improve your teaching in the past 18 months?

Teacher

coll_textbook

How often do you discuss and decide on the selection of
instructional media (textbooks)

Teacher

tchg_materials

Frequency exchange teaching materials with colleagues

Teacher

team_conf

Frequency attend team conferences for age group taught

Teacher

discuss_learning

Frequency engage in discussion about the learning
development of specific students

Teacher

team_teaching

Frequency teach jointly as a team in the same class

Teacher

obs_feedback

Frequency teacher observes other teachers’ classes and
provide feedback

Teacher

joint_activities

Frequency teacher engages in joint activities across
different classes and age groups (e.g. projects)

Teacher

hw_practice

Frequency discuss and coordinate homework practice
across subjects

Teacher

Index of Teacher
Collaboration
pd_network

160

Table 15 (cont’d)
VARIABLE NAME

DESCRIPTION

LEVEL

Factor A

Factor score measuring transformational leadership

School

Factor B

Factor score measuring transformational leadership

School

Factor C

Factor score measuring transformational leadership

School

No. of hours spent in administrative duties, and clerical,
paperwork as a teacher

Teacher

CONTROLS:
School Climate Controls:
BTG08C

Teacher
BTG41B

% class time spent maintaining discipline in the classroom

lackpers

TALIS index of lack of personnel in school

School

autcurr

TALIS index of curriculum autonomy

School

Teacher Controls:
tfem

Teacher Gender

Teacher

texp

Teacher experience in years

Teacher

tefficacy

If I try really hard I can make progress with even the most
difficult and unmotivated students

Teacher

School Controls:
prinfem

Principal gender

School

prinexp

Principal experience

School

public

Is school public/private?

School

161

Table 16: Descriptive Statistics for Analysis on Teacher Collaboration – TALIS 2008
Countries
Dependent Variable
Teacher Satisfaction (0-1)
Index of Collaboration
Collaboration Variables
PD Network
Collaborative Research
Mentoring/Coaching
Dialogue about Teaching
Textbook Selection/Discussion
Teaching Materials
Team Conferences
Discuss Student Learning
Team Teaching
Joint Activities
Homework Practice
Transformational Leadership
Indices
Factor A
Factor B
Factor C
School Controls
Principal Gender (Female)
Principal Experience
Public
Language Difference
Parent Education

DNK

POL

BRA

MEX

HUN

Mean
0.90
0.73

SE
0.00
0.00

Mean
0.93
0.56

SE
0.00
0.00

Mean
0.84
0.49

SE
0.01
0.01

Mean
0.90
0.73

SE
0.00
0.00

Mean
0.81
0.45

SE
0.01
0.00

0.46
0.55
0.18
0.91
0.81
0.93
0.93
0.87
0.88
0.64
0.57

0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01

0.63
0.41
0.69
0.96
0.09
0.73
0.76
0.94
0.46
0.58
0.45

0.01
0.01
0.01
0.00
0.00
0.01
0.01
0.00
0.01
0.01
0.01

0.24
0.55
0.50
0.95
0.20
0.62
0.76
0.87
0.42
0.52
0.40

0.01
0.01
0.01
0.00
0.01
0.01
0.01
0.01
0.01
0.01
0.01

0.46
0.55
0.18
0.91
0.81
0.93
0.93
0.87
0.88
0.64
0.57

0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01

0.45
0.16
0.47
0.79
0.07
0.67
0.75
0.68
0.12
0.45
0.55

0.01
0.01
0.01
0.01
0.00
0.01
0.01
0.01
0.01
0.01
0.01

-0.02
0.03
-

0.05
0.04
-

-0.02
0.06

0.09
0.04

-0.02
-

0.04
-

-0.02
0.03
-

0.05
0.04
-

0.04

0.03

0.36
0.62
0.73
1.40
0.48

0.02
0.03
0.01
0.05
0.03

0.69
0.30
0.93
1.92
0.09

0.02
0.02
0.01
0.02
0.01

0.75
0.22
0.85
1.41
0.09

0.02
0.01
0.01
0.01
0.01

0.36
0.62
0.73
1.40
0.48

0.02
0.03
0.01
0.05
0.03

0.48
0.37
0.83
1.13
0.14

0.03
0.02
0.02
0.01
0.01

162

Table 16 (cont’d)
Countries

DNK
Mean

POL
SE

Mean

BRA
SE

Mean

School Climate
Index Lack of School Personnel
-0.73
0.03
-0.74
0.03
0.40
Index Teacher Autonomy
0.77
0.02
-0.26
0.04
0.03
% Time Spent on Administration
3.87
0.12
3.24
0.05
3.23
% Time on Classroom Disruption
11.94
0.25
9.03
0.12
17.75
Teacher Controls
Teacher Gender (Female)
0.57
0.01
0.75
0.00
0.73
Teacher Experience
0.52
0.01
0.60
0.01
0.49
Teacher Efficacy
0.19
0.01
0.20
0.01
0.23
N
1021
2188
3302
Notes:
1. Brr Standard Errors
2. All descriptive statistics are weighted
3. The range for the index measuring a lack of school personnel ranges from -1 to 1.

163

MEX
SE

Mean

HUN
SE

Mean

SE

0.05
0.04
0.21
0.26

-0.73
0.77
3.87
11.94

0.03
0.02
0.12
0.25

-0.36
0.63
5.18
9.99

0.04
0.02
0.10
0.14

0.02
0.02
0.01

0.57
0.52
0.19
2034

0.01
0.01
0.01

0.77
0.74
0.07
2141

0.01
0.01
0.00

3.7 Method
3.7.1 Method – Research Question 1
Research Question 1: Is frequent teacher collaboration across a breadth of collaborative
activities among teachers associated with transformational leadership in a multi-country context
controlling for the effect of teacher demographics, principal demographics, school background
and school climate? In order to address the first question I constructed an index of collaboration
(described in the previous section) and then regressed this index of collaboration on factor scores
measuring dimensions of transformational leadership. The index of collaboration is a summative
index that takes on fractional values in the unit interval such that

. Given that the

index takes on values between [0, 1] inclusive of [0, 1] I use a fractional response model
developed by Papke and Wooldridge (1996) rather than a logistic regression. The Fractional
response model (FRM) overcomes several of the problems encountered when other methods are
used with fractional response variables. Logistic regressions for example would require arbitrary
adjustment for all observations that are not either 0, or 1. The coefficients in the FRM model
cannot be interpreted as marginal effects or slopes. However, the sign and significance of the
coefficients will be the same as the sign and significance of the marginal effects evaluated at the
mean value of the independent variables. Because of this, and the difficulty of interpreting
changes in some of the independent variables (factor scores), the sign and significance of the
association between the independent and dependent variables are the results of interest here
(Table 17).
The FRM is a quasi-maximum likelihood estimation (QMLE) method.

The binomial

GLM variance assumption required for a generally robust inference may be fulfilled in STATA

164

by specifying a binomial distribution family and a logit link function26. The model specification
is given below.
coll_indexi = δ + β1 (factor A) + β2 (factor B) + β3 (factor C) + Sβ4 + SCβ5 + Tβ6 + ε

[1]

where the dependent variable coll_index is an index measuring the breadth of teacher
collaboration and taking on values between 0 and 1.
transformational leadership.

is a factor score measuring principals’

is a second factor score for transformational leadership.

is a

third factor score for transformational leadership. For any given country at most two of the three
factor scores load. Table 14 provides a detailed look at the factor loadings for each country.

is

a vector of school controls such as principal gender and experience, whether the school is public
or private, whether the language of instruction is different from the language spoken at home and
parents’ education level (ISCED 5 or more for at least 1 parent).

is a vector of school climate

which includes an index measuring a shortfall in school personnel, level of curricular autonomy
among teachers, percent of time spent by teachers quelling classroom disruption and percent of
time spent by teachers in administrative tasks.

is a vector of teacher controls such as teacher

gender, teacher experience and teacher efficacy. Since teacher age and experience were found to
be highly collinear, only teacher experience is used in all regressions.
3.7.2 Method – Research Question 2
Research Question 2: Specifically, what are the kinds of collaborative activity that are
associated with different dimensions of transformational leadership? In order to identify the
kinds of collaboration associated with the factor scores indicative of transformational leadership
in this analysis (namely factor A, factor B, and factor C), I ran twelve separate logistic
regressions per country with a different dependent variable each time measuring some form of
26

In stata the code for a fractional response model (FRM) is glm y x1….xk fam(bin) logit(link) vce(robust)

165

teacher collaboration (see table 16 for a list of all the collaboration variables used as dependent
variables). I used Stata’s svy brr prefix to compensate for the complex survey structure of
TALIS and to adjust for the clustering of teachers within schools. Since the dependent variables
measuring type of collaboration were binary, the logistic specification seemed an appropriate
choice. Model 2.1 shows the first of these twelve regressions. Models 2.2-2.12 are not shown in
the interest of parsimony.
Odds (Y=1) = eα + β1(factor A) + β2(factor B) + β3(factor C)+ S β4 + SC β5+ Tβ6

[2.1]

In the first model Y measures teacher participation (over the course of the past 18
months) in a network of teachers formed specifically for the professional development of
teachers.

is one of three factor scores created using Exploratory Factor Analysis measuring

principals’ transformational leadership.
leadership.

is the second factor score for transformational

is a third factor score for transformational leadership.

is a vector of school

controls such as principal gender and experience, whether the school is public or private, whether
the language of instruction is different from the language spoken at home and parents’ education
level (ISCED 5 or more for at least 1 parent).

is a vector of school climate which includes

an index measuring a shortfall in school personnel, level of curricular autonomy among teachers,
percent of time spent by teachers quelling classroom disruption and percent of time spent by
teachers in administrative tasks.

is a vector of teacher controls such as teacher gender, teacher

experience and teacher efficacy.
3.7.3 Method – Research Question 3
Research Question 3: Is participation in a breadth of collaborative activities associated
with high teacher job satisfaction in a multi-country context?

166

For my third research question, I created a binary variable measuring teacher satisfaction
where 1 is indicative of teacher agreement with the statement “I am satisfied by my job” and 0 is
indicative of disagreement with this statement. A description of this variable is provided under
section 3.6.2. This measure of teacher satisfaction was regressed on the index of collaboration27
controlling for a range of school, school climate and teacher demographic variables associated in
the literature with teacher satisfaction. Since my dependent variable for this question is binary, I
used logistic regression with Stata’s svy brr prefix to compensate for the complex survey
structure of TALIS and to adjust for the clustering of teachers within schools. This clustering
implies that teacher responses within schools are more similar than teacher responses across
schools, which violates the OLS assumption of independence and therefore needs to be
addressed. The logistic model specification used for research question 3 is shown below.
Odds (Y=1) = eα +β1 (coll_index)+ S β2 + SC β3 + T β2

[3]

In the above equation, Y is a binary variable measuring teacher perceptions of teachers’ job
satisfaction.

is an index that measures the richness of teacher collaboration across a

breadth of collaborative activities.

is a vector of school controls such as whether the school is

public or private, whether the language of instruction is different from the language spoken at
home and parents’ education level (ISCED 5 or more for at least 1 parent).

is a vector of

school climate variables which include the level of curricular autonomy among teachers, percent
of time spent by teachers quelling classroom disruption and percent of time spent by teachers in
administrative tasks.

is a vector of teacher controls such as teacher gender, teacher experience

and teacher efficacy.

27

Also self-created. See section 3.6.2

167

The next section provides a detailed look at the results of these analyses. Tables 17 - 23
summarize these results. I organize a discussion of results by research questions and by country
within each research question.
3.8 Results & Discussion
3.8.1

Results – Research Question 1

3.8.11 Denmark. Table 17 shows results for the association between the index of
teacher collaboration and principals’ transformational leadership. We do not observe a
consistently positive relationship across countries between the factor scores measuring
transformational leadership and the summative index measuring breadth of teacher collaboration
in the countries being analyzed. What we do see however is that teacher efficacy is consistently
and positively associated with teacher participation in collaborative activity.
In Denmark, the principal’s role in staff development consistent with the schools goals as
embodied by factor A is not significantly associated with a breadth of teacher collaboration.
However, teaching in a public school is positively associated with the index of collaboration. As
expected, the percentage of class time spent maintaining discipline is negatively associated with
frequent collaboration. Presumably, the teacher has less time available for collaborative
activities in a situation where much of the teacher’s energy is taken up maintaining discipline in
class. Additionally, female teachers in Denmark are more likely to engage in a breadth of
collaborative activities than male teachers. Frequent collaboration across a breadth of
collaboratively activities is positively and significantly associated with teacher efficacy (p<0.01).
3.8.12 Poland. In Poland on the other hand, the principal’s role in pursuing goal driven
staff development (Factor A) is positively and significantly associated with the index of
collaboration (p<0.05). Teaching in a public school is also positively associated with the index of

168

collaborative activity. Notably, female teachers are more likely to engage in a breadth of
collaborative activities than male teachers. Teacher efficacy in Poland as in the other countries is
positively associated with teacher participation in a breadth of collaborative activities. There is a
positive relationship between percent time spent disciplining the class and the index of
collaboration which does not at first glance make intuitive sense. Perhaps teachers in Poland
tend to engage in conversations with colleagues more when they face disciplinary issues in the
classroom.
3.8.13 Brazil. In Brazil, the transformational leadership factor (factor A which seems to
measure staff development consistent with school goals) is negatively associated with the index
of collaborative activity. Taking country context into account, we know that in Brazil principals
seem to display strong instructional rather than transformational leadership. Thus, the coefficient
on factor A may just be capturing the lack of transformational leadership in Brazil. Teachers in
Brazil spend a higher percent of class time maintaining discipline than in other countries (We
know from table 16 that this number is 17 percent on average). Nevertheless, the data show a
positive association between time spent maintaining discipline and collaboration. It is possible
that teachers collaborate with one another over classroom management to address this discipline
problem in the schools where they work. Being female and having a high sense of teacher
efficacy are both positively and significantly associated with frequent teacher participation in a
breadth of collaborative activities.
3.8.14 Mexico. In Mexico, principals’ role in developing a positive school climate
through reaching out to the parents and working jointly with the teachers on creating a school
development plan seems to matter for the index of collaboration in schools. We see a positive
and statistically significant association between factor B and the index of collaboration.

169

Additionally, teacher experience and teacher self-efficacy are both positively associated with the
index of collaboration (p<0.05 and p<0.001 respectively).
3.8.15 Hungary. Hungary is the only country where there is not a significant association
between a transformational leadership factor and the index of collaboration. It is also the only
country in the analysis where principal experience is positively associated with teacher
collaboration. The variable language difference is also positively associated with teacher
collaboration. This variable represents the percentage of students whose first language differs
from the language of instruction such that 1=less than 10%; 2=more than 10% but less than 20%
and so on. From table 16 we see that the average value of the response categories for this
variable is 1.13. This indicates that approximately 10% of students in Hungarian schools speak a
language different from the language of instruction. Presumably, a more diverse group of
students in class catalyzes collaboration across classes. Although percent of time spent by
Hungarian teachers maintaining order in the classrooms is among the lowest of all TALIS
countries, there is a significant negative association between time spent on disciplinary actions
and teacher participation in a breadth of collaborative activities. As in the other countries,
female teachers who are more experienced and teachers with a sense of self-efficacy are more
likely to engage in a breadth of collaborative activity.
3.8.16 Summary of Results – RQ1. Results of the association between transformational
leadership and the index of teacher collaboration vary by country. Across countries, school
factors appear to be less important than teacher variables. Two teacher specific characteristics
that seem to matter consistently for participation in a breadth of collaborative activity across
contexts in addition to school leadership are a strong sense of teacher efficacy and teacher

170

Table 17: Table of Results for Collaboration Index and Transformational Leadership
Index of Collaboration (DV)

DNK

Transformational Leadership
Factor A 0.051
(0.041)
Factor B 0.091**
(0.044)
Factor C
School Controls
Principal Gender (Female)
Principal Experience
Public
Language Difference
Parent Education

0.010
(0.077)
0.102
(0.077)
0.581***
(0.098)
-0.018
(0.034)
-0.114
(0.073)

POL

BRA

0.058**
(0.029)

-0.056*
(0.032)

MEX

HUN

0.072*
(0.039)
-0.012
(0.033)

0.056
(0.050)

0.030
(0.056)
0.069
(0.060)
0.266**
(0.127)
-0.051
(0.057)
0.074
(0.135)

0.065
(0.076)
-0.091
(0.073)
-0.036
(0.136)
0.129*
(0.076)
-0.057
(0.125)

0.094
(0.069)
-0.088
(0.064)
-0.075
(0.115)
0.122*
(0.069)
0.131
(0.128)

0.086
(0.055)
0.108**
(0.055)
-0.115
(0.081)
0.219***
(0.081)
0.042
(0.071)

0.008
(0.034)
0.008***
(0.002)

-0.000
(0.034)

0.011
(0.030)

-0.005*
(0.003)

-0.018
(0.033)
0.012***
(0.002)

-0.001
(0.002)

-0.004**
(0.002)

0.116**
(0.053)
0.069
(0.056)
0.281***
(0.071)

0.222***
(0.051)
0.037
(0.042)
0.293***
(0.045)

0.209***
(0.040)
0.045
(0.036)
0.289***
(0.048)

0.399***
(0.150)
1,257

-0.083
(0.178)
2,606

-0.333*
(0.186)
4,170

-0.004
(0.041)
0.104**
(0.046)
0.250***
(0.039)
0.550***
(0.208)
2,681

0.130**
(0.053)
0.127***
(0.049)
0.444***
(0.077)
0.578***
(0.128)
2,440

School Climate
Index Lack of School Personnel

% Time on Classroom Disruption
Teacher Controls
Teacher Gender (Female)
Teacher Experience
Teacher Efficacy

Constant
N

0.153***
(0.057)

Notes:
1. All numbers reported are coefficients from an FRM. 2. Numbers in parentheses represent standard errors
3. Significance is given by p<0.01 ***; p<0.05 **; p<0.10*

171

gender. The association between teacher gender (female) and collaboration as well as teacher
efficacy and collaboration is positive and significant across countries.
3.8.2

Results – Research Question 2

My second research question unpacks the type of collaborative activity (there are 12)
associated with aspects of transformational leadership. Results for this analysis are shown in
tables 18-22 by country.
3.8.21 Denmark. In Denmark, the transformational leadership factor score (factor A) is
positively and significantly (p<0.01) associated with collaborative activity that focuses on the
core task of teaching. Specifically, the odds of teacher collaboration related to an exchange of
teaching materials between teachers, teacher discussions about improving student learning, team
teaching, as well as, peer observation of classes and feedback on teaching are greater when the
principal engages in staff development connected with school goals (factor A). Teacher
participation in a network, collaborative research, mentoring and holding a dialogue about
teaching is not significantly associated with factor A.
The transformational leadership factor (factor B) which appears to relate more to the
principal’s role in strengthening school culture is positively and significantly associated with
teacher participation in team conferences, team teaching, teacher discussions about student
learning, teacher willingness to teach a class jointly and frequent teacher coordination of
homework practices across subjects.
The odds of participating in a professional development network, engaging in
collaborative research, engaging in an informal dialogue about improving teaching, and
frequently coordinating homework practices across subjects are higher in Denmark if the
principal is a female. Conversely, the odds of teacher participation in discussion around the

172

selection of textbooks, exchange of teaching materials, participation in team conferences and
engagement in discussions on student learning are lower if the principal is female. When the
principal is more experienced, the odds of teacher participation in more instruction focused tasks
such as engaging in dialogue about teaching, selecting textbooks, and teaching the same class
jointly are higher. Teachers are also more likely to engage in collaborative research, mentoring
and coaching. For eight out of the twelve models the odds of teacher participation in activities
focused on student learning and instruction/teaching are higher if the teacher teaches in a public
school. With respect to school climate, teachers in Denmark have lower odds of collaboration if
percent of time spent managing classroom disruptions goes up. This makes sense particularly in
the case of time-intensive activities such as frequently holding joint activities with other teachers
across classes/ages, and frequently coordinating homework across subjects. Finally, the odds of
teacher participation in ten of the twelve collaborative activities are higher if teacher efficacy is
high.
3.8.22 Poland. In Poland the odds of participation in eight of the twelve different types of
collaborative activity are higher when the principal exercises transformational leadership (factor
A). Factor A has a significant positive association with teacher participation in collaborative
research, dialogue about teaching, textbook selection, team conferences, discussion about
learning, teacher classroom observation and feedback, and coordinating homework across
subject areas. Factor C which appears to capture classroom management by the principal and
teacher working together is significantly associated with only two types of collaborative activity
– teacher participation in a professional development network and teacher participation in
collaborative research.

173

The odds of teachers participating in a teaching network, collaborative research,
mentoring/coaching, team conferences, and opportunities for teachers to observe and provide
feedback to other teachers are higher if the teacher teaches in a public school in Poland. There is
a positive and significant relationship between all twelve types of collaboration and the
percentage of time spent by teachers on administrative tasks. The odds of participation are lower
for ten out of the twelve types of collaboration when the teachers spend more time on classroom
disruptions. Female teachers in Poland are more likely to engage in a breadth of collaborative
activities and the odds of participation in ten out of twelve types of collaboration are higher when
the teachers have high self-efficacy.
3.8.23 Brazil. In Brazil, the factor (factor A – staff development organized around
school and teaching goals) measuring transformational leadership is negatively associated with
teacher participation in collaborative activities pertaining to instructional matters such as teacher
exchange of teaching materials, team conferences, discussions about learning, and teacher
observation/feedback of another teachers classroom.
The odds of teacher participation in 11 of the 12 collaborative activities are higher when
the number of hours spent by a teacher per week on administrative tasks increases. Since
administration involves working with people, this is not entirely surprising. This result is
consistent with what we observe in Poland. The percent of time spent by Brazilian teachers on
managing classroom disruptions is negatively associated with teacher participation in
collaborative activities. This is consistent with the fact that the percentage of lesson time lost to
disruptive student behavior in Brazil is the highest of all TALIS countries.
The odds of participation in a professional development network, informal dialogue about
teaching, exchanging teacher materials, participating in team conferences, discussing learning,

174

team teaching and teaching classes jointly are higher in Brazil if the teacher is female.
Additionally, teacher efficacy is positively and significantly (p<0.01) associated with all twelve
types of teacher collaboration.
3.8.24 Mexico. In Mexico, the transformational leadership factor (factor B) which
appears to relate more to the principal’s role in strengthening school culture is positively and
significantly associated with teacher participation in mentoring, informal dialogue in teaching,
selection of textbooks, team conferences, and teacher discussions about learning. Mexican
teachers teaching in public schools have higher odds of engaging in dialogue about teaching, and
participating in team conferences. The odds of teacher collaboration increase in Mexico when
the percent of students whose first language is different from the language of instruction goes up.
Presumably, this catalyzes teachers to pool their resources about how to handle language barriers
in the classroom. Similar to Poland and Brazil, the number of hours spent on administrative
tasks by teachers in Mexico is significantly and positively associated with nine of the twelve
collaborative activities. Finally, the odds of participation in all but one of the types of
collaborative activity are higher with high teacher efficacy.
3.8.25 Hungary. In Mexico, the transformational leadership factor (factor C) which
appears to relate more to the principal’s joint role with the teacher with respect to classroom
management is positively and significantly associated with teacher participation in textbook
selection, team conferences, discussions about learning, and teaching a class jointly. Factor C is
negatively associated with teacher participation in professional networks, and mentoring. In
Hungary the odds of participation in collaborative activity are also significantly and positively
associated with principal experience (more than 10 years). There is also a positive and
significant association between the number of hours per week spent on administrative duties and

175

participation in nine of the twelve types of collaboration. For eight out of the twelve
collaborative activities, the odds of teacher participation in collaboration are lower if time spent
on managing classroom disruption goes up. For seven of the twelve activities, the odds of
teacher participation are higher if the teacher is female. Specifically, these seven activities are
mentoring, dialogue about teaching, exchanging teaching materials, team conferences,
discussions about learning, teaching a joint class, and coordinating homework practice. Finally,
the odds of teacher participation in all twelve types of collaboration are higher in Hungary as in
all the other countries with high teacher efficacy.
3.8.26 Summary of Results – RQ2. Factor A (which gets at staff development
consistent with school goals) appears to be systematically associated with three different types of
collaborative activity across three countries – Brazil, Poland and Denmark. These activities are
teachers exchanging teaching materials, engaging in discussions about learning and participating
in teacher observation and feedback. Factor B (strengthening school culture) is systematically
associated with frequent participation in team conferences and discussions about learning in
Denmark and Mexico. Factor C (classroom management) is systematically negatively associated
with teacher participation in professional development networks in both Poland and Hungary.
Additionally, the finding that across all countries analyzed the odds of teacher
participation in the majority of different types of collaborative activity are higher when teachers
have a high teacher efficacy is fairly robust. School climate factors such as time spend on
administrative tasks and time spent maintaining discipline in class are also significantly
associated with teacher participation for many of the collaborative activities. This makes
intuitive sense given that time spent on classroom discipline implies less time for collaborative
professional development activities.

176

Table 18 : Results for Types of Collaboration & Transformational leadership in Denmark
(1)
(5)
(6)
(7)
(8)
(9)
(10)
(11)
(12)
Dependent Variables (0-1)
(2)
(3)
(4)
Leadership Indices (IV)
Factor A 1.059
0.965
0.927
1.096
1.050
1.352*** 1.088
1.187** 1.239*
1.289*** 0.939
1.018
(0.073)
(0.048)
(0.074)
(0.086)
(0.058)
(0.141)
(0.164)
(0.079)
(0.156)
(0.110)
(0.073)
(0.049)
Factor B 0.916
0.985
1.253** 1.141
0.900
1.282
2.310*** 1.246** 1.253** 0.970
1.229*** 1.099**
(0.060)
(0.060)
(0.112)
(0.099)
(0.061)
(0.202)
(0.325)
(0.105)
(0.131)
(0.056)
(0.072)
(0.052)
School Controls
Female Principal
1.430*** 1.303** 0.788
1.314*
0.677*** 0.452*** 0.502** 0.564*** 0.758
1.104
1.177
1.172*
(0.138)
(0.136)
(0.132)
(0.181)
(0.071)
(0.067)
(0.135)
(0.078)
(0.159)
(0.105)
(0.125)
(0.102)
Principal Experience
1.068
1.321*** 1.297** 1.320** 1.348*** 1.325
0.745
0.922
1.677*** 1.010
1.066
1.110
(0.090)
(0.121)
(0.161)
(0.172)
(0.148)
(0.227)
(0.158)
(0.077)
(0.216)
(0.109)
(0.108)
(0.086)
Public
1.201** 1.046
1.168
0.904
2.158*** 4.715*** 10.58*** 3.913*** 5.857*** 0.793** 0.922
1.505***
(0.100)
(0.096)
(0.192)
(0.142)
(0.243)
(0.861)
(2.290)
(0.479)
(0.767)
(0.077)
(0.126)
(0.125)
Language Difference
1.156*** 1.068** 1.218*** 1.246** 1.004
0.821
0.913
0.957
1.261** 1.378*** 0.734*** 0.997
(0.043)
(0.035)
(0.072)
(0.119)
(0.073)
(0.098)
(0.083)
(0.053)
(0.136)
(0.098)
(0.038)
(0.038)
Parent Education
0.891
0.905
1.129
0.725** 0.993
1.969*** 1.154
1.247*
1.294*
0.746** 0.473*** 1.007
(0.077)
(0.076)
(0.128)
(0.109)
(0.092)
(0.375)
(0.226)
(0.145)
(0.190)
(0.083)
(0.056)
(0.076)
School Climate
% Time on Administrative 1.022*
1.036** 0.990
0.963*** 0.994
0.969** 0.979** 0.946*** 1.016
1.035*** 0.994
1.000
(0.012)
(0.014)
(0.014)
(0.012)
(0.009)
(0.014)
(0.010)
(0.008)
(0.015)
(0.010)
(0.009)
(0.010)
% Time on Classrm
0.992** 0.986*** 1.002
0.991
0.993*
0.981*** 0.970*** 0.990*
0.993
1.004
0.986*** 0.980***
(0.003)
(0.004)
(0.004)
(0.006)
(0.004)
(0.007)
(0.007)
(0.006)
(0.005)
(0.004)
(0.004)
(0.004)
Teacher Background
Female Teacher
0.978
1.093
1.054
2.191*** 2.372*** 1.695*** 1.044
1.218
1.362*** 0.881
0.822*** 1.244***
(0.087)
(0.079)
(0.137)
(0.255)
(0.244)
(0.252)
(0.150)
(0.153)
(0.141)
(0.069)
(0.061)
(0.093)
Teacher Experience
1.351*** 1.681*** 0.884
1.003
0.864
0.633*** 0.897
0.791** 0.787
1.186** 1.307*** 1.093
(0.117)
(0.132)
(0.090)
(0.157)
(0.082)
(0.107)
(0.137)
(0.085)
(0.128)
(0.101)
(0.088)
(0.077)
Teacher Efficacy
2.111*** 1.450*** 1.012
1.045
1.439*** 1.351*
2.789*** 2.306*** 1.699*** 1.334*** 1.392*** 1.587***
(0.183)
(0.149)
(0.110)
(0.158)
(0.184)
(0.244)
(0.708)
(0.350)
(0.274)
(0.125)
(0.146)
(0.137)
Constant
0.406*** 0.600*** 0.124*** 6.836*** 1.827*** 7.310*** 11.10*** 4.761*** 1.118
0.172*** 4.594*** 0.854
(0.064)
(0.103)
(0.028)
(2.029)
(0.311)
(2.407)
(4.142)
(0.898)
(0.270)
(0.029)
(0.855)
(0.104)
N
1,125
1,125
1,125
1,114
1,114
1,127
1,125
1,123
1,119
1,125
1,127
1,115
Notes:
1. All numbers are Odds ratios 2. std errors are brr 3. sig levels p<0.01 ***, p<0.05 **, p<0.10 *
4. (1)=PD Network, (2)=Collaborative Research, (3)=Mentoring, (4)=Dialogue about teaching, (5)=Textbook discussion, (6)=Exchange teaching materials
(7)=Team Conferences, (8)=Discuss Learning, (9)= team teaching, (10)=observation/feedback, (11)=Joint Class, (12)=Coordinate HW Practice

177

Table 19 : Results for Types of Collaboration & Transformational leadership in Poland
(1)
(5)
(6)
(7)
(8)
(9)
(10)
(11)
(12)
Dependent Variables (0-1)
(2)
(3)
(4)
Leadership Indices (IV)
Factor A 0.964*** 1.044*** 1.002
1.056*** 1.039*** 1.092*** 1.054*** 1.173*** 0.992
1.083*** 0.983
1.076***
(0.010)
(0.009)
(0.012)
(0.017)
(0.011)
(0.013)
(0.018)
(0.017)
(0.009)
(0.012)
(0.012)
(0.010)
Factor C 1.203*** 1.067** 1.028
0.900
0.951
0.977
0.993
0.980
0.979
1.039
0.992
1.055
(0.036)
(0.030)
(0.041)
(0.057)
(0.039)
(0.042)
(0.057)
(0.059)
(0.026)
(0.040)
(0.035)
(0.037)
School Controls
Female Principal
0.862** 0.967
1.357*** 1.508*** 0.989
0.963
0.951
0.801*
1.149** 0.915
0.976
0.753***
(0.055)
(0.058)
(0.107)
(0.177)
(0.077)
(0.083)
(0.096)
(0.099)
(0.067)
(0.064)
(0.069)
(0.048)
Principal Experience
0.840*** 1.017
1.198** 1.215
0.892
1.026
0.682*** 1.120
0.884*
1.193** 1.065
1.036
(0.050)
(0.050)
(0.100)
(0.173)
(0.084)
(0.104)
(0.082)
(0.129)
(0.056)
(0.081)
(0.073)
(0.069)
Public
1.858*** 1.319** 1.682** 2.407*** 1.182
1.212
1.630*
0.792
0.899
1.612** 0.449*** 0.955
(0.131)
(0.182)
(0.376)
(0.605)
(0.271)
(0.253)
(0.433)
(0.152)
(0.141)
(0.318)
(0.062)
(0.181)
Language Difference
0.981
0.874*** 1.035
0.803** 0.919
0.801** 0.747** 0.931
1.258*** 0.823*** 1.006
1.242***
(0.050)
(0.040)
(0.076)
(0.076)
(0.073)
(0.081)
(0.103)
(0.112)
(0.077)
(0.050)
(0.053)
(0.066)
Parent Education
1.063
1.360** 1.502*
1.707*** 0.803
0.764
0.986
0.648** 1.078
1.856*** 0.727*** 0.817
(0.064)
(0.188)
(0.321)
(0.321)
(0.171)
(0.138)
(0.254)
(0.109)
(0.113)
(0.318)
(0.062)
(0.120)
School Climate
% Time on Administrative
Tasks
1.033*** 1.031*** 1.102*** 1.070*** 1.031*** 1.047*** 1.085*** 1.118*** 1.041*** 1.068*** 1.048*** 1.041***
(0.006)
(0.006)
(0.010)
(0.020)
(0.008)
(0.010)
(0.012)
(0.037)
(0.006)
(0.007)
(0.008)
(0.007)
% Time on Classrm
Disruption
0.988*** 1.000
0.999
0.988** 0.990*
0.981*** 0.976*** 0.981*** 0.993** 0.987*** 0.981*** 0.974***
(0.003)
(0.003)
(0.003)
(0.005)
(0.005)
(0.003)
(0.003)
(0.005)
(0.003)
(0.003)
(0.002)
(0.003)
Teacher Background
Female Teacher
1.192*** 1.028
1.234*** 1.631*** 0.522*** 1.657*** 1.926*** 2.176*** 1.201*** 0.823*** 1.175** 1.191***
(0.071)
(0.074)
(0.074)
(0.176)
(0.069)
(0.120)
(0.121)
(0.206)
(0.070)
(0.054)
(0.076)
(0.073)
Teacher Experience
1.593*** 1.267*** 0.632*** 1.216*
0.752*** 0.987
1.371*** 1.763*** 1.134*** 0.616*** 1.050
1.365***
(0.079)
(0.073)
(0.031)
(0.130)
(0.059)
(0.055)
(0.083)
(0.184)
(0.054)
(0.035)
(0.049)
(0.068)
Teacher Efficacy
1.110*
1.414*** 1.051
1.272
1.920*** 1.261*** 1.121
2.323*** 1.659*** 1.414*** 1.446*** 1.726***
(0.063)
(0.085)
(0.066)
(0.197)
(0.217)
(0.075)
(0.092)
(0.369)
(0.112)
(0.088)
(0.083)
(0.118)
Constant
0.706*** 0.473*** 0.806
6.513*** 0.164*** 2.476*** 2.136*** 9.877*** 0.403*** 0.892
2.545*** 0.510***
(0.080)
(0.074)
(0.214)
(2.173)
(0.045)
(0.593)
(0.610)
(2.730)
(0.064)
(0.208)
(0.433)
(0.104)
N
2,532
2,532
2,532
2,500
2,518
2,510
2,514
2,519
2,391
2,514
2,509
2,476
Notes:
1. All numbers are Odds ratios 2. std errors are brr 3. sig levels p<0.01 ***, p<0.05 **, p<0.10 *
4. (1)=PD Network, (2)=Collaborative Research, (3)=Mentoring, (4)=Dialogue about teaching, (5)=Textbook discussion, (6)=Exchange teaching materials
(7)=Team Conferences, (8)=Discuss Learning, (9)= team teaching, (10)=observation/feedback, (11)=Joint Class, (12)=Coordinate HW Practice

178

Table 20 : Results for Types of Collaboration & Transformational leadership in Brazil
(1)
(5)
(6)
(7)
(8)
(9)
(2)
(3)
(4)

(10)
(11)
(12)
Dependent Variables (0-1)
Leadership Indices (IV)
Factor A 1.090
0.937
0.932
0.915
0.984
0.891** 0.841*** 0.809*** 1.043
0.881** 0.962
0.928
(0.066)
(0.059)
(0.050)
(0.069)
(0.040)
(0.041)
(0.053)
(0.052)
(0.043)
(0.045)
(0.044)
(0.049)
School Controls
Female Principal
1.356** 0.912
0.805*** 1.044
0.923
1.166*
1.415*** 1.038
1.066
0.951
1.211
0.970
(0.200)
(0.064)
(0.064)
(0.154)
(0.087)
(0.107)
(0.167)
(0.151)
(0.105)
(0.128)
(0.160)
(0.108)
Principal Experience
0.858
0.989
1.031
1.350*
0.825*
1.000
0.797*
0.712** 0.704*** 0.863
0.732*** 0.776*
(0.104)
(0.123)
(0.100)
(0.214)
(0.095)
(0.106)
(0.096)
(0.105)
(0.065)
(0.172)
(0.062)
(0.104)
Public
1.067
0.670*** 1.461*** 1.270
0.786
1.690*** 0.862
1.073
0.758** 0.921
1.436*** 0.917
(0.221)
(0.088)
(0.188)
(0.400)
(0.133)
(0.204)
(0.130)
(0.208)
(0.082)
(0.185)
(0.189)
(0.132)
Language Difference
0.955
0.992
1.131
0.720*
1.203*
0.927
0.708*** 0.759** 0.940
1.175
1.029
1.112
(0.124)
(0.124)
(0.105)
(0.134)
(0.132)
(0.105)
(0.071)
(0.087)
(0.095)
(0.127)
(0.107)
(0.121)
Parent Education
1.767*** 0.827
1.290** 1.914
0.828
1.254
1.447*
1.683
0.860
0.618** 1.063
1.213
(0.240)
(0.139)
(0.159)
(0.764)
(0.192)
(0.297)
(0.281)
(0.531)
(0.179)
(0.131)
(0.279)
(0.258)
School
%
TimeClimate
on Administrative
Tasks
1.017** 1.041*** 1.034*** 0.995
1.015** 1.018*** 1.022*** 1.011
1.048*** 1.045*** 1.035*** 1.044***
(0.008)
(0.011)
(0.006)
(0.018)
(0.006)
(0.006)
(0.007)
(0.010)
(0.008)
(0.006)
(0.009)
(0.007)
% Time on Classrm
Disruption
1.000
0.995
1.001
0.988*** 0.992*** 0.986*** 0.985*** 0.981*** 0.991*** 0.994*
0.992*** 0.994**
(0.003)
(0.003)
(0.004)
(0.004)
(0.003)
(0.003)
(0.003)
(0.005)
(0.003)
(0.003)
(0.003)
(0.003)
Teacher Background
Female Teacher
1.136** 0.889
1.080
1.963*** 1.056
2.065*** 1.621*** 1.538*** 1.274*** 0.775*** 1.295*** 1.042
(0.070)
(0.143)
(0.062)
(0.223)
(0.069)
(0.134)
(0.112)
(0.147)
(0.069)
(0.050)
(0.080)
(0.063)
Teacher Experience
1.355*** 0.976
1.219*** 1.219
1.110
1.170*** 1.153** 1.062
1.261*** 0.789*** 1.131*
0.899
(0.099)
(0.053)
(0.078)
(0.227)
(0.084)
(0.066)
(0.075)
(0.103)
(0.077)
(0.070)
(0.070)
(0.062)
Teacher Efficacy
1.740*** 1.313*** 1.464*** 1.495*** 1.292*** 1.462*** 1.322*** 1.317** 1.515*** 1.454*** 1.623*** 1.882***
(0.222)
(0.079)
(0.101)
(0.212)
(0.087)
(0.106)
(0.126)
(0.149)
(0.109)
(0.158)
(0.134)
(0.136)
Constant
0.148*** 1.826*
0.462*** 14.99*** 0.239*** 0.622** 3.385*** 8.607*** 0.719
0.171*** 0.489*** 0.522**
(0.043)
(0.657)
(0.113)
(5.758)
(0.093)
(0.143)
(0.919)
(2.149)
(0.167)
(0.048)
(0.122)
(0.144)
N
3,641
3,641
3,641
3,587
3,622
3,619
3,631
3,626
3,608
3,632
3,633
3,622
Notes:
1. All numbers are Odds ratios 2. std errors are brr 3. sig levels p<0.01 ***, p<0.05 **, p<0.10 *
4. (1)=PD Network, (2)=Collaborative Research, (3)=Mentoring, (4)=Dialogue about teaching, (5)=Textbook discussion, (6)=Exchange teaching materials
(7)=Team Conferences, (8)=Discuss Learning, (9)= team teaching, (10)=observation/feedback, (11)=Joint Class, (12)=Coordinate HW Practice

179

Table 21 : Results for Types of Collaboration & Transformational leadership in Mexico
(1)
(5)
(6)
(7)
(8)
(9)
(10)
(11)
(12)
Dependent Variables (0-1)
(2)
(3)
(4)
Leadership Indices (IV)
Factor B 1.122
1.071
1.113** 1.310*** 1.111*
1.091*
1.245*** 1.196*** 1.050
1.037
1.080
1.089
(0.080)
(0.054)
(0.048)
(0.092)
(0.060)
(0.055)
(0.072)
(0.068)
(0.066)
(0.071)
(0.058)
(0.072)
School Controls
Female Principal
0.979
1.200** 1.099
1.371** 1.200** 1.245
1.201
1.427*** 0.820*** 0.521*** 1.153** 0.900
(0.112)
(0.108)
(0.078)
(0.199)
(0.104)
(0.190)
(0.138)
(0.117)
(0.056)
(0.062)
(0.083)
(0.083)
Principal Experience
0.898
1.035
0.967
0.962
0.834** 0.896
1.020
0.765*** 1.036
0.901
0.677*** 0.896
(0.062)
(0.087)
(0.070)
(0.109)
(0.072)
(0.066)
(0.082)
(0.059)
(0.059)
(0.101)
(0.078)
(0.069)
Public
1.104
0.939
1.084
2.585** 0.627*** 0.966
1.411*** 0.660
0.927
0.470** 0.965
0.910
(0.168)
(0.172)
(0.179)
(1.038)
(0.108)
(0.132)
(0.170)
(0.171)
(0.182)
(0.143)
(0.132)
(0.183)
Language Difference
1.164** 1.034
0.874*
0.753*** 0.801*** 1.341*** 1.059
0.866*
1.159*
1.446*** 1.179** 1.191**
(0.068)
(0.068)
(0.064)
(0.076)
(0.065)
(0.089)
(0.110)
(0.071)
(0.089)
(0.112)
(0.090)
(0.104)
Parent Education
0.910
0.930
1.141
2.356
0.972
2.141*** 0.894
1.005
0.798
1.018
1.083
1.125
(0.124)
(0.275)
(0.169)
(1.256)
(0.163)
(0.416)
(0.155)
(0.310)
(0.127)
(0.333)
(0.127)
(0.218)
School
Climate
% Time on Administrative
Tasks
1.017** 1.025*** 1.029*** 1.006
1.010
1.014** 1.016*
1.005
1.017*** 1.070*** 1.040*** 1.047***
(0.008)
(0.007)
(0.005)
(0.008)
(0.008)
(0.006)
(0.009)
(0.006)
(0.006)
(0.006)
(0.009)
(0.005)
% Time on Classrm
Disruption
0.996
0.986*** 0.996
0.986** 0.999
0.988*** 0.996
0.990*** 1.010*** 0.998
0.999
1.004
(0.004)
(0.004)
(0.004)
(0.006)
(0.004)
(0.003)
(0.004)
(0.003)
(0.003)
(0.008)
(0.003)
(0.004)
Teacher Background
Female Teacher
0.914
0.999
1.071
1.295** 0.925
1.333*** 1.202** 0.971
0.812** 0.679*** 1.094
1.162**
(0.100)
(0.057)
(0.087)
(0.132)
(0.057)
(0.119)
(0.085)
(0.072)
(0.069)
(0.064)
(0.067)
(0.073)
Teacher Experience
1.087
0.876
1.145** 1.178*
1.084
1.004
1.262*** 1.200** 1.065
0.919
1.236** 1.099
(0.101)
(0.073)
(0.064)
(0.109)
(0.097)
(0.116)
(0.110)
(0.089)
(0.120)
(0.075)
(0.106)
(0.069)
Teacher Efficacy
1.238*** 1.177*** 1.238*** 1.005
1.360*** 1.343*** 1.128
1.546*** 1.225** 1.430*** 1.434*** 1.401***
(0.058)
(0.071)
(0.078)
(0.116)
(0.097)
(0.088)
(0.102)
(0.128)
(0.103)
(0.154)
(0.082)
(0.084)
Constant
0.274*** 2.042*** 0.571** 6.252*** 0.562** 0.644** 1.664
3.713*** 0.838
0.132*** 0.390*** 0.163***
(0.060)
(0.486)
(0.153)
(2.408)
(0.144)
(0.122)
(0.547)
(0.983)
(0.241)
(0.060)
(0.076)
(0.036)
N
2,211
2,211
2,211
2,177
2,204
2,199
2,211
2,204
2,187
2,210
2,208
2,205
Notes:
1. All numbers are Odds ratios 2. std errors are brr 3. sig levels p<0.01 ***, p<0.05 **, p<0.10 *
4. (1)=PD Network, (2)=Collaborative Research, (3)=Mentoring, (4)=Dialogue about teaching, (5)=Textbook discussion, (6)=Exchange teaching materials
(7)=Team Conferences, (8)=Discuss Learning, (9)= team teaching, (10)=observation/feedback, (11)=Joint Class, (12)=Coordinate HW Practice

180

Table 22 : Results for Types of Collaboration & Transformational leadership in Hungary
(1)
(5)
(6)
(7)
(8)
(9)
(10)
(11)
Dependent Variables (0-1)
(2)
(3)
(4)
Leadership Indices (IV)
Factor C 0.839*** 0.984
0.911** 1.034
1.101*
0.962
1.245*** 1.268*** 1.027
0.966
1.090*
(0.037)
(0.042)
(0.042)
(0.054)
(0.057)
(0.048)
(0.062)
(0.049)
(0.051)
(0.060)
(0.055)
School Controls
Female Principal
1.367*** 0.937
1.706*** 1.097
0.667*** 1.093
1.040
1.035
1.283** 1.761*** 1.217***
(0.112)
(0.068)
(0.158)
(0.076)
(0.054)
(0.082)
(0.082)
(0.069)
(0.130)
(0.163)
(0.079)
Principal Experience
1.405*** 0.898
1.417*** 1.569*** 0.950
1.013
1.334*** 1.295*** 1.398*** 1.447*** 1.341***
(0.099)
(0.060)
(0.132)
(0.118)
(0.092)
(0.063)
(0.100)
(0.083)
(0.125)
(0.145)
(0.090)
Public
1.337*** 0.786** 1.189
0.815** 0.822
1.182*
0.826*
1.283*** 0.573*** 0.661*** 0.918
(0.106)
(0.095)
(0.144)
(0.079)
(0.111)
(0.109)
(0.092)
(0.121)
(0.086)
(0.075)
(0.082)
Language Difference
1.320*
1.142
1.250*
1.222
0.676** 0.979
1.765*** 1.737*** 3.031*** 1.056
1.871***
(0.213)
(0.173)
(0.150)
(0.330)
(0.121)
(0.142)
(0.356)
(0.275)
(0.441)
(0.203)
(0.230)
Parent Education
1.021
1.263** 1.388*** 1.056
1.552*** 1.599*** 1.327*** 0.881
1.476*** 1.231*
0.869
(0.082)
(0.113)
(0.133)
(0.099)
(0.150)
(0.121)
(0.142)
(0.095)
(0.165)
(0.145)
(0.081)
School Climate
% Time on Administrative Tasks 1.034*** 1.005
1.052*** 1.005
1.053*** 1.031*** 1.048*** 1.042*** 1.005
1.089*** 1.023***
(0.006)
(0.005)
(0.005)
(0.007)
(0.007)
(0.006)
(0.009)
(0.007)
(0.005)
(0.007)
(0.006)
% Time on Classrm Disruption
0.998
0.990** 0.989*** 1.006*
1.030*** 0.997
0.990*** 0.993*** 0.992*
0.978*** 0.985***
(0.003)
(0.004)
(0.003)
(0.004)
(0.004)
(0.004)
(0.003)
(0.002)
(0.005)
(0.003)
(0.003)
Teacher Background
Female Teacher
1.048
0.579*** 1.233*** 1.275*** 0.734** 1.403*** 1.491*** 1.374*** 0.610*** 1.068
1.304***
(0.067)
(0.045)
(0.078)
(0.099)
(0.088)
(0.105)
(0.100)
(0.077)
(0.050)
(0.085)
(0.098)
Teacher Experience
1.641*** 0.947
1.061
0.738*** 1.003
0.917
1.494*** 1.044
1.036
1.305*** 0.945
(0.096)
(0.102)
(0.061)
(0.048)
(0.120)
(0.070)
(0.077)
(0.064)
(0.094)
(0.082)
(0.050)
Teacher Efficacy
1.588*** 2.761*** 2.268*** 1.338*
1.829*** 1.522*** 2.258*** 2.254*** 1.728*** 1.610*** 2.324***
(0.158)
(0.360)
(0.208)
(0.197)
(0.258)
(0.216)
(0.406)
(0.271)
(0.191)
(0.158)
(0.239)
Constant
0.199*** 0.318*** 0.266*** 2.805*** 0.0976*** 1.174
0.753
0.571** 0.0603*** 0.252*** 0.303***
(0.052)
(0.076)
(0.063)
(1.009)
(0.031)
(0.239)
(0.177)
(0.127)
(0.015)
(0.072)
(0.060)
N
2,393
2,393
2,393
2,353
2,403
2,374
2,383
2,372
2,350
2,385
2,368
Notes:
1. All numbers are Odds ratios 2. std errors are brr 3. sig levels p<0.01 ***, p<0.05 **, p<0.10 *
4. (1)=PD Network, (2)=Collaborative Research, (3)=Mentoring, (4)=Dialogue about teaching, (5)=Textbook discussion, (6)=Exchange teaching materials
(7)=Team Conferences, (8)=Discuss Learning, (9)= team teaching, (10)=observation/feedback, (11)=Joint Class, (12)=Coordinate HW Practice

181

(12)
1.046
(0.057)
1.169***
(0.063)
1.142**
(0.070)
0.920
(0.065)
1.051
(0.156)
0.860
(0.078)
1.009**
(0.004)
0.993**
(0.003)
1.454***
(0.086)
1.610***
(0.103)
2.219***
(0.241)
0.589**
(0.121)
2,353

3.8.3 Results – Research Question 3
3.8.31 Denmark. In Denmark, the index of teacher collaboration is positively associated
with teachers’ job satisfaction. What this implies is that the odds of teacher satisfaction are
higher when teachers frequently engage in a variety of collaborative activities (p<0.05).
Teachers in Denmark who teach in public schools have lower odds of satisfaction (p<0.01). In
Denmark, teacher autonomy with respect to curriculum is associated with higher odds of teacher
satisfaction. Given that on average teachers in Denmark report a high level of autonomy28this
result is consistent with the literature that claims high levels of teacher autonomy are associated
with higher teacher satisfaction (Moore, 2012; Perie & Baker, 1997; Renzulli, 2011). The odds
of teacher satisfaction are lower when teachers spend more time on classroom discipline issues.
This finding is also consistent with the literature on teacher satisfaction. In fact, student
discipline issues have been cited as a major reason for teacher dissatisfaction (Kennedy, 2005;
Liu & Meyer, 2005; Stockard & Lehman, 2004). Finally, there appears to be a positive and
statistically significant association between teacher efficacy and teacher satisfaction in Denmark
and this finding is consistent across the countries analyzed.
3.8.32 Poland. In Poland, the odds of teacher satisfaction are higher if teachers engage
frequently in a breadth of collaborative activity. Poland appears to have a shortage of school
personnel29. Reflecting this school climate, the odds of teacher satisfaction are lower when there
is a shortage of school personnel. The odds of teacher satisfaction are lower when teachers
spend more of their class time maintaining discipline. Teachers in Poland are overwhelmingly
female (75%). Female teachers in Poland report higher odds of job satisfaction. Finally, there is

28

See table of descriptives (table 16). Level of autonomy is 0.77 on a scale from -1 to 1 where a higher value
indicates more autonomy.
29
The average shortage of school personnel in Poland is -0.74 on a scale of -1 to 1 where 1 indicates this is not a
problem.

182

a positive and statistically significant relationship between teachers’ job satisfaction in Poland
and teacher efficacy.
3.8.33 Brazil. In Brazil, the odds of teacher satisfaction are higher when teachers engage
frequently in a breadth of collaborative activities. Teachers who work in public schools in Brazil
report lower odds of being satisfied with their jobs. Teachers in Brazil who spend more of their
class time maintaining discipline report lower odds of being satisfied with their jobs.
Additionally, female teachers in Brazil report higher odds of job satisfaction. From table 16 we
know that in Brazil 73% of teachers are female. Further, Brazilian teachers who are more
experienced (>10 yrs) have lower odds of job satisfaction. Finally, the odds of teacher
satisfaction are higher among teachers who report high level of teacher efficacy.
3.8.34 Mexico. In Mexico, the odds of teacher satisfaction are higher if teachers engage
frequently in a breadth of collaborative activity. Teachers teaching classes in which 60% or
more of the students have at least one parent who has completed higher secondary or higher
(ISCED level 5) report higher odds of job satisfaction. In other words, the odds of teacher
satisfaction are higher when the level of parent education is high. The odds of satisfaction are
lower if teachers spend more class time maintaining order. This finding is fairly intuitive and
consistent across the countries analyzed. Finally, the odds of teacher satisfaction in Mexico are
higher when teachers report high teacher efficacy.
3.8.35 Hungary. In Hungary, the odds of teacher satisfaction are higher if teachers
frequently engage in a breadth of collaborative activities. Among school factors, the odds of
teacher satisfaction are lower for teachers teaching in public schools in Hungary. The odds of
teacher satisfaction are also lower when the number of students who speak a language different

183

Table 23: Table of Results for Teacher Satisfaction & Index of Collaboration

Teacher Satisfaction (DV)
Index of Collaboration (IV)
School Controls
Public
Language Difference
Parent Education
School Climate
Index of Curricular Autonomy
Index of Lack of School Personnel
% Time on Classroom Disruption
Teacher Controls
Teacher Gender (Female)
Teacher Experience
Teacher Efficacy
Constant
N
Notes:

DNK

POL

BRA

MEX

HUN

2.193**
(0.663)

5.956***
(1.353)

11.49***
(2.102)

9.810***
(3.459)

2.587***
(0.445)

0.442***
(0.075)
1.021
(0.088)
0.942
(0.135)

1.423
(0.417)
1.024
(0.100)
1.083
(0.281)

0.652*
(0.166)
0.977
(0.113)
1.566
(0.625)

1.286
(0.610)
0.883
(0.137)
2.402**
(0.881)

0.593*
(0.163)
0.577***
(0.116)
1.286**
(0.131)

1.241**
(0.113)
1.038**
(0.018)
0.974***
(0.005)

1.177**
(0.084)
0.968***
(0.007)
0.965***
(0.004)

0.986
(0.063)
1.010
(0.009)
0.990***
(0.003)

1.041
(0.102)
0.992
(0.011)
0.981***
(0.004)

0.951
(0.084)
1.008
(0.008)
0.978***
(0.004)

0.967
(0.090)
0.835
(0.113)
1.529**
(0.263)
10.05***
(2.977)
1,147

1.246**
(0.111)
0.986
(0.083)
1.581***
(0.214)
5.550***
(1.935)
2,725

1.260**
(0.112)
0.718***
(0.045)
2.136***
(0.264)
2.916***
(0.978)
4,014

1.174
(0.122)
1.159
(0.124)
1.876***
(0.270)
5.421***
(2.974)
2,301

1.297***
(0.105)
0.862
(0.096)
4.779***
(0.906)
9.543***
(2.975)
2,532

1. All numbers reported are odds ratios. 2. Numbers in parentheses represent standard errors
3. Significance is given by p<0.01 ***; p<0.05 **; p<0.10*

from the language of instruction increases. Presumably, this is so because communication with
the student becomes more challenging when language is a barrier. In Hungary, as in the other
countries analyzed the odds of teacher satisfaction are lower when more of a teacher’s class time
is given over to maintaining discipline. Female teachers in Hungary have higher odds of job
184

satisfaction and there is a significant positive relationship between job satisfaction and teacher
efficacy in Hungary.
3.8.36 Summary – RQ3 Results. To summarize, we observe a consistently significant
positive relationship between teacher satisfaction and frequent teacher engagement in a breadth
of collaborative activities across all the countries analyzed. We also observe that across all five
countries more class time spent maintaining discipline is significantly associated with lower odds
of teacher job satisfaction. This makes intuitive sense since, time spent in maintaining order is
time taken away from instruction which is an integral part of a teachers work. Demographically,
we see that in three of the five countries (Poland, Brazil, and Hungary) the odds of teacher
satisfaction are higher for females than for males and this association is significant (p<0.05).
This may simply reflect the fact that traditionally teaching has been a female dominated
profession or it may be due to the fact that in these countries over 70 percent of the teachers are
female. Finally, across all five countries we see a statistically significant relationship between
teacher efficacy and teachers’ job satisfaction. This finding is consistent with the literature on
teacher efficacy (See Capara et al, 2006; Lee, Dedrick, & Smith, 1991; Tschannen-Moran et al.
1998; Tschannen-Moran & Hoy, 2007).
The following section provides a detailed look at how recent policy changes and country
context serve as an explanatory framework within which to make sense of the results in Mexico.
3.9 Contextual Discussion of Results in Mexico
3.9.1 The National Context: In 2009 Mexico ranked as the world’s fourteenth largest
economy (OECD 2011). Improving the quality of education is a political and social priority in
Mexico given the high poverty rates and high inequality in the system (OECD, 2010; OECD,
2011). A wide range of reforms over the past twenty years have led to improvements in school
enrolment and school quality. Despite previous efforts to institute systemic improvement
185

however, one in two 15 year olds in Mexico failed to reach even the baseline achievement level
(level 2) in PISA 2006 as compared to the OECD average of 19.2 percent (OECD, 2010).
Similarly, though attainment rates in Mexico have risen sharply since 2000 the proportion of
adults who have attained at least an upper secondary education in Mexico is one of the smallest
among the OECD countries (36 percent) and considerably lower than the OECD average of 75
percent (OECD, 2013).
Location, ethnic background, and poverty are all factors associated with inequality in
school enrolment in Mexico. In Chiapas for example, for every 1,000 students who entered basic
education at the appropriate age in the year 2000, only 476 exited on time versus 747 in
Aguascalientes (OECD, 2011). Issues of educational coverage and equity continue to be
fundamental problems at the forefront of educational service provision in Mexico.
3.9.2 Overview of the Educational System. In 2010, Mexico spent approximately 6.2
percent of its GDP on education (OECD, 2013). This is slightly below the OECD average of 6.3
percent but higher than the proportion of GDP spent on education in some developed countries
like Australia, Spain, and Switzerland (OECD, 2013). Yet, in the Mexican context increasing
expenditure on education does not necessarily translate into more spending per student. Thus,
average annual expenditure per student (from primary to tertiary education) at 20 percent in
Mexico continues to lag the OECD average of 28 percent of GDP per capita (OECD, 2013).
Schools in Mexico operate with scarce resources. These resources are allocated mostly to
staff compensation. Mexico devotes 93.3 percent of its education budget to staff compensation
(OECD, 2013).
The Mexican education system is organized into three major levels. These are basic
education (consisting of pre-school, primary, lower secondary), upper secondary education and

186

higher education (comprising university education, technological education and normal
education) (Guevara et al., 2004). Primary education in Mexico is compulsory and consists of
six grades. In 1993 lower secondary education was made obligatory. Mexico is now changing
coverage of compulsory education to include upper secondary education with the aim of
attaining universal upper secondary education by 2022 (OECD, 2013).
Compared to other OECD countries, Mexico has the highest student-teacher ratios at all
levels of compulsory education (OECD, 2013). In 2011, the average pupil-teacher ratio for
primary education in Mexico was 28 students per teacher (OECD, 2013). For the same year the
average pupil-teacher ratio for secondary education in Mexico was even higher - 30 students per
teacher (OECD, 2013).
Since Mexico’s efforts to decentralize in 1992, it has been the state educational
authorities’ responsibility to provide, organize and supervise basic education and teacher training
in Mexico (Guevara et al, 2004). Nonetheless, the Ministry of Education (SEP) still has
considerable control in setting the curriculum, selecting the textbooks, hiring and firing school
personnel and setting salary schedules for teachers. According to Guevara and Gonzalez (2004),
the Ministry of Education (SEP) and the teachers’ union are two key actors in the education
policy arena within Mexico. The World Bank and the Inter-American Development Bank also
have a major external presence in Mexican education (Guevara et al. 2004).
3.9.3 Profile of Teachers in Mexico. The largest proportion of teachers in Mexico is
concentrated at the level of basic education. Primary school teachers constitute the bulk of this
number (OECD, 2010). Further, the majority of teachers (around 80 percent) are public school
teachers (Guevara et al. 2004). According to Guevara and Gonzalez (2004), primary school
teachers attend to just one group of children for the whole school year. This implies that the

187

same teacher teaches all the subjects for the corresponding grade in the academic year. A
characteristic of secondary school teachers, however, is the specification of the number of hours
because a teacher teaches certain subjects to different groups/grades. Thus, secondary teachers
may attend to different groups of students in different schools. This fact complicates the labor
situation with respect to secondary school teachers and may also interfere with the nature and
extent of teacher collaboration in Mexico (Guevara et al, 2004).
The TALIS 2007 data show that 57 percent of lower secondary teachers in Mexico are
female which is consistent with the fact that in most TALIS countries there is a predominance of
female teachers at the lower secondary level (TALIS, 2008). Further, around half of all lower
secondary teachers in TALIS 2007 have ten or more years of experience. In Mexico, teacher
education is organized by level. Teachers in basic education receive their preparation in higher
education institutions called Teachers’ Colleges (OECD, 2010). Upper secondary school
teachers however, are prepared in universities in their subjects often without any specific training
to develop teaching skills (OECD, 2010).
Almost all teachers, principals, and administrative personnel associated with basic
education, training of teachers, and federal and state services of the Department of Public
Education are members of the National Union of Educational Workers (SNTE) created in 1943.
The SNTE deducts 1 percent of the salary of its members and also has considerable say in the
hiring of new teachers (Guevara et al. 2004). Given its large membership (well in excess of a
million education workers) the SNTE thus enjoys considerable economic and political clout in
Mexico (Guevara et al, 2004).
Teacher promotion in Mexico (at the basic level) is addressed via two national programs.
These are the vertical scale (EV) and the teaching career (CM). The CM is essentially a system

188

of horizontal promotion that allows teachers mobility and access to a higher income on a base
salary. Evaluation criteria include seniority, academic degree, professional training, professional
upgrading, professional performance and student progress. Notably, the CM system seeks to
connect teacher performance to teacher salary (Guevara et al. 2004; OECD, 2010).
The vertical scale (EV) mechanism for promotion on the other hand only takes into
account a teacher’s seniority, educational qualification and professional development not their
performance (Guevara et al., 2004). Promotion in the EV is conditional on the creation of
positions via retirement or resignation of teachers. Promotions can occur within the same level
from teaching to administrative or between teaching levels from primary to secondary. For those
teachers who are in interim teaching positions and competing for permanent positions,
evaluations by principals via teacher observations become quite relevant (Guevara et. al 2004).
Although both career ladders (the CM and the EV) partially help to identify high performing
teachers there are few negative consequences associated with identification of poor performance
in Mexico.
Recent reform efforts in Mexico with respect to school management have included
recommendations for a teacher evaluation system that is both summative and formative in nature.
The OECD Steering group recommendations (2009) promote teacher evaluation that provides
teachers feedback on how to improve their practice. The main purpose of evaluating teachers (as
stated in these recommendations) is to improve their practice with the ultimate goal of improving
student achievement.
Overall, Mexican teachers operate under fairly difficult conditions. Many hold more than
one job at a time in order to supplement their income (OECD, 2010). This second job may or
may not be related to the teaching profession but it is made possible by the fact that schools of

189

basic education work in two shifts (morning/afternoon) that are each five hours long. Mexico
does not have formal induction programs for beginning teachers and teachers do not have
adequate support and mentoring or a probationary period before obtaining a permanent position
(OECD, 2010).
3.9.4 Leadership in Mexico. My analysis shows that only 36 percent of school
principals in Mexico are women. This is consistent with what we know about leadership in
Mexico perhaps indicating a “glass ceiling” for promotion possibilities for women within schools
(TALIS, 2008). Around 62 percent of principals have 10 or more years of experience in Mexico
and about 73 percent of schools in Mexico are public schools.
School administrators have virtually no role in personnel decisions or allocation of
resources (OECD, 2010). It is the SNTE that negotiates directly with the SEP to set teacher
salary schedules and yearly pay increases. By law all principals must belong to the SNTE, which
is the teachers’ union. The placement and hiring of teachers is often negotiated by state SNTE
factions with corresponding state education authorities (OECD, 2010).
The role of the principal extends to evaluating the performance of their teachers,
managing the assigned budget for their school, managing the school’s relationship with the
Parents’ Association, and school maintenance. Parents’ Associations mainly support the school
authorities in the collection of funds and the organization of voluntary work related to school
maintenance. Additionally, principals in Mexico spend a good bit of their time in paperwork
compiling school statistics and completing school documents.
Because schools in Mexico operate in multiple shifts it is challenging to develop the
school as a learning community where students can engage in extended discussions and extra-

190

curricular activities. Similarly, teachers too have little time to interact with parents and students,
evaluate students individually and even prepare their classes (Santibanez et al. 2005). .
Given these sorts of constraints in the school environment it is not surprising that the
aspect of transformational leadership that appears to matter positively for teacher collaboration in
the context of Mexico is the principal’s role in establishing a positive school climate (Factor B).
Collaboration is only possible when school facilities are adequate and the principal is supportive
of such teacher interaction. The index measuring lack of school personnel in my analysis has a
low value of -0.73 for Mexico (on a scale from -1 to 1) thus indicating that schools in Mexico
lack adequate instructional and other support personnel.
The variables that load onto Factor B are the principal’s role in creating an orderly
atmosphere in the school, principal and teacher collaboration on creating a school development
plan, the principals role in fostering a task-oriented atmosphere in the school and the principals
interaction with parents. In particular, one can see the relevance of the variable parent_outreach,
(principal interaction with parents) which gets at the principals efforts at community fundraising.
This is a powerful way in which principals in Mexico try to compensate for poor school
infrastructure. Funds collected in this manner are subsequently utilized for building
improvements and the purchase of essential office equipment such as a fax or copy machine.
The relevance of the principal’s role in creating an orderly atmosphere in the school (the variable
schorder) for teacher collaboration is also somewhat intuitive. Presumably, less time spent on
maintaining discipline helps foster opportunities for professional collaboration. Teachers in
Mexico report that around 12 percent of their class time is spent maintaining discipline.
My analysis also shows specific collaborative activities that matter in the Mexican
context are all school-based (see table 21, activities (3) to (8)) and pertain more to sharing of

191

instructional materials and improving instructional practice via teacher discussion rather than
providing teacher observation and feedback.
Finally, the finding in table 23 that the odds of teacher satisfaction are higher in Mexico
if the teachers engage in a breadth of collaborative activity is also explicable in terms of the poor
support infrastructure in most schools there. Presumably, engagement in a variety of
collaborative activities provide teachers the opportunity to form closer ties with colleagues and
thus establish a sense of teacher community difficult for the most part in a majority of schools in
Mexico due to the challenging work conditions there.
In sum, a key contribution of my work is that it extends knowledge about teacher
collaboration - a less explored aspect of teacher policy in Mexico, as well as school management
using large-scale comparative data for Mexico. This research is particularly relevant in the
Mexican context given that teacher quality and school management are two aspects of education
policy prioritized by the SEP (OECD, 2010; Santibanez et al., 2005).
3.10 Conclusion & Limitations
This analysis finds that different aspects of a principals’ transformational leadership
pertaining to staff development consistent with school goals, developing a positive school
culture, and joint management of classroom issues are significantly associated with the
frequency and breadth of teacher collaboration in the countries analyzed. This index of
collaboration is also significantly and consistently associated with teacher satisfaction across all
five countries. Additionally, teacher efficacy is positively associated with the index of
collaboration as well as teachers’ job satisfaction controlling for a range of teacher, principal and
school factors and these findings appear to be quite robust across contexts.

192

There is a vast body of literature that documents that teachers’ self-efficacy beliefs are
related to the efforts teachers invest in the classroom, their persistence when things are less than
smooth sailing and their resilience in the face of adversity. These behaviors are important
because ultimately they impact students. From a policy perspective, principals would do well to
address issues of teacher efficacy in their schools given the apparently strong and consistent
association that exists between efficacy and teacher job satisfaction and also between efficacy
and the frequency of teacher participation in a breadth of collaborative activities in schools.
Leaders can address issues of teacher self-efficacy by providing institutional support (via
sustained opportunities for formal and informal teacher collaboration that enable shared norms
and values among teachers) and verbal endorsement of the job that teachers do. Research has
shown that self-efficacy is stronger among teachers who perceive a positive school climate with
a strong sense of teacher community. Transformational school leaders can help to create this
strong teacher community within schools. For example, research has shown that “principals play
an important role in allocating time for teachers to meet and for providing increased
opportunities for job-embedded professional development (Wahlstrom & Louis, 2008, p.463).”
This analysis also shows that more class time spent maintaining discipline is negatively
associated with teacher satisfaction and is also negatively associated with the index of
collaboration across diverse national contexts. From a policy perspective the percent of time
spent managing class disruption is reflective of school climate, a factor that is subject to policy
manipulation by school principals.
This study relies primarily on the TALIS 2008 data to analyze the relationships between
transformational leadership, teacher collaboration and teacher satisfaction in a subset of
countries. A methodological limitation of this dataset is that the data are cross-sectional

193

permitting only correlational analyses. We cannot infer causality in the identified associations
between transformational leadership and teacher collaboration or between collaboration and
satisfaction because it is typically difficult to measure and identify all confounding variables and
to include these in the quantitative model (Schneider, Carnoy, Kilpatrick, Schmidt & Shavelson,
2007). Thus, there is considerable potential to replicate similar analyses using longitudinal data
which offer more flexibility for causal modelling. Additional analyses using different data
would provide a valuable opportunity to examine whether and to what extent the results
generated in this analysis hold true across diverse contexts and over time.

194

BIBLIOGRAPHY

195

BIBLIOGRAPHY
Adams, C. M., & Forsyth, P. B. (2013). Revisiting the trust effect in urban elementary
schools. The Elementary School Journal, 114(1), 1-21.
Achinstein, B. (2002). Conflict and community: the micropolitics of teacher collaboration.
Teachers College Record, 104(3), 4212-4455.
Barnett, K., & McCormick, J. (2004). Leadership and individual principal-teacher relationships
in schools, Educational Administration Quarterly, 40(3), 406-434.
Bass, B. M., & Avolio, B. J. (1994). Improving organizational effectiveness through
transformational leadership. Thousand Oaks, CA: Sage Publications.
Bolam, R., McMahon, A., Stoll, L., Thomas, S., & Wallace, M. (2005). Creating and sustaining
professional learning communities. Research Report Number 637. London: General
Teaching Council for England, Department for Education and Skills.
Bray, Adamson & Mason (Eds.) (2007). Comparative education research: Approaches and
methods. Hong Kong: Springer.
Brundrett, M. (1998). What lies behind collegiality, legitimation or control? An analysis of the
purported benefits of collegial management in education. Educational Management
Administration & Leadership, 26(3), 305-316.
Bryk, A. S., & Schneider, B. (2003). Trust in schools: A core resource for school
reform. Educational leadership, 60(6), 40-45.
Burns, J. (1978). Leadership. New York, NY: Harper & Row.
Caprara, G. V., Barbaranelli, C., Steca, P., & Malone, P. S. (2006). Teachers' self-efficacy
beliefs as determinants of job satisfaction and students' academic achievement: A study at the
school level. Journal of school psychology, 44(6), 473-490.
Cha, Y. K., & Ham, S. H. (2012). Constructivist teaching and intra-school collaboration among
teachers in South Korea: an uncertainty management perspective. Asia Pacific Education
Review, 13(4), 635-647.
Coburn, C. E. (2001). Collective sensemaking about reading: How teachers mediate reading
policy in their professional communities. Educational evaluation and policy analysis, 23(2),
145-170.
Cosner, S. (2009). Building organizational capacity through trust. Educational Administration
Quarterly, 45(2), 248-291.

196

Daly, A. J. (2009). Rigid response in an age of accountability. The potential of leadership and
trust. Educational Administration Quarterly, 45(2), 168-216.
Demir, K. (2008). Transformational Leadership and Collective Efficacy: The Moderating Roles
of Collaborative Culture and Teachers' Self-Efficacy. Eurasian Journal of Educational
Research (EJER), (33).
Dumay, X., Boonen, T., & Van Damme, J. (2013). Principal Leadership Long-Term Indirect
Effects on Learning Growth in Mathematics. The Elementary School Journal, 114(2), 225251.
Duyar, I., Gumus, S., & Bellibas, M. S. (2013). Multilevel analysis of teacher work attitudes:
The influence of principal leadership and teacher collaboration. International Journal of
Educational Management, 27(7), 700-719.
Goddard, Y., Goddard, R., & Tschannen-Moran, M. (2007). A theoretical and empirical
investigation of teacher collaboration for school improvement and student achievement in
public elementary schools. The Teachers College Record, 109(4), 877-896.
Goddard, Y. L., Miller, R., Larson, R., & Goddard, R. (2010). Connecting principal leadership,
teacher collaboration, and student achievement. Paper presented at the Annual Meeting of
the American Educational Research Association (AERA), May 3, 2010, Denver, CO.
Guevara, M. R., & Gonzalez, L. E. (2004). Attracting, developing and retaining effective
teachers: Country background report for Mexico. Mexico, DF: Secretaría De Educación
Pública-OECD. www.oecd.org/edu/calidadeductiva
Gumus, S., Bulut, O., & Bellibas, M. S. (2013). The Relationship between Principal Leadership
and Teacher Collaboration in Turkish Primary Schools: A Multilevel Analysis. Education
Research & Perspectives, 40(1).
Hargreaves, D. (1994). The new professionalism: The synthesis of professional and institutional
development. Teaching & Teacher Education, 10(4), 423-438.
Heck, R. H., & Hallinger, P. (2010). Collaborative leadership effects on school improvement:
Integrating unidirectional-and reciprocal-effects models. The Elementary School
Journal, 111(2), 226-252.
Johnson, S. M., Kraft, M. A., & Papay, J.P. (2011). How context matters in high needs schools:
The effects of teachers’ working conditions on their professional satisfaction and their
students’ achievement. Project on the Next Generation of Teachers, Harvard Graduate
School of Education.
Johnson, S. M. (2006). The workplace matters: Teacher quality, retention and effectiveness.
Harvard Graduate School of Education, July 2006.
197

Johnson, S. M., & Birkeland, S. E. (2003). Pursuing a ‘Sense of Success’: New Teachers
Explain their Career Decisions. American Educational Research Journal, 40(3), 581-617.
Judge, T.A., Parkers, S., Colbert, A., Heller, D., & Ilies, R. (2001). Job satisfaction: A crosscultural review in Anderson, N. (ed.), Handbook of Industrial and Organizational
Psychology, SAGE Publications.
Kelchtermans, G. (2006). Teacher collaboration and collegiality as workplace conditions. A
review. Zeitschrift für Pädagogik, 52(2), 220-237.
Ladd, H. (2011). Teachers’ perceptions of their work environments: How predictive of planned
and actual teacher movement? Educational Evaluation and Policy Analysis, 33(2), 235-261.
DOI: 10.3102/0162373711398128.
Lavie, J. M. (2006). Academic discourses on school-based teacher collaboration: Revisiting the
arguments. Educational Administration Quarterly, 42, 773-805.
Lee, V. E., Dedrick, R. F., & Smith, J. B. (1991). The effect of the social organization of schools
on teachers' efficacy and satisfaction. Sociology of Education, 64(3) 190-208.
Leithwood, K., & Jantzi, D. (2005). A review of transformational school leadership research
1996–2005. Leadership and policy in schools, 4(3), 177-199.
Leithwood, K., & Sun, J. (2012). The nature and effects of transformational school leadership:
A meta-analytic review of unpublished research. Educational Administration Quarterly,
48(3), 387-423.
Levine, T. H., & Marcus, A. S. (2010). How the structure and focus of teachers' collaborative
activities facilitate and constrain teacher learning. Teaching and Teacher Education, 26(3),
389-398.
Little, J. W. (1990). The persistence of privacy: Autonomy and initiative in teachers’
professional relations. Teachers College Record, 91(4), 509-534.
Little, J. W. (2003). Inside teacher community: Representations of classroom practice.
Teachers College Record, 105(6), 913-945.
Liu, X. S., & Ramsey, J. (2008). Teachers’ job satisfaction: Analysis of the Teacher Follow-up
Survey in the United States for 2000-2001. Teaching & Teacher Education, 24, 1173-1184.
Lortie, D. C. 2002. Schoolteacher, Chicago: University of Chicago Press.
Louis, K. S., & Marks, H. M. (1998). Does professional learning community affect the
classroom? Teachers’ work and student experiences in restructuring schools. American
Journal of Education, 106(4), 532-575.
198

Marks, H. M., & Printy, S. M. (2003). Principal leadership and school performance: An
integration of transformational and instructional leadership. Educational administration
quarterly, 39(3), 370-397.
McLaughlin, M. W., & Talbert, J.E. (2001). Professional communities and the work of high
school teaching. Chicago: University of Chicago Press.
Moolenaar, N. M., Daly, A. J., & Sleegers, P. J. (2010). Occupying the principal position:
Examining relationships between transformational leadership, social network position, and
schools’ innovative climate. Educational administration quarterly, 46(5), 623-670.
OECD (2013). Education at a glance 2013: OECD indicators, OECD Publishing, Paris,
http://dx.doi.org/10.1787/eag-2013-en.
OECD (2011). Establishing a framework for evaluation and teacher incentives: Considerations
for Mexico, OECD Publishing. doi: 10.1787/9789264094406-3-enOECD. (2010). TALIS 2008
technical report. Paris: OECD Publishing. URL:
http://www.oecd.org/education/school/44978960.pdf
OECD (2010). TALIS user guide for the International database. Paris: OECD Publishing. URL:
http://data.library.utoronto.ca/datapub/codebooks/oecd/talis/round1/talis_idb_user_guide.pdf
OECD Steering Group on School Management and Teacher Policy (2009). Preliminary
Recommendations on School Management and Teacher Policy in Mexico. OECD, Paris.
www.oecd.org/edu/calidadeductiva.
OECD (2008). Creating effective teaching and learning environments: First results from TALIS.
www.oecd.org/edu/calidadeductiva
Papke, L.E., & Wooldridge, J. M. (1996). Econometric methods for fractional response
variables with an application to 401(K) plan participation rates. Journal of Applied
Econometrics, 11, 619-632.
Parise, L. M., & Spillane, J. P. (2010). Teacher learning and instructional change: How formal
and on-the-job learning opportunities predict change in elementary school teachers'
practice. The Elementary School Journal, 110(3), 323-346.
Price, H.E. (2012). Principal-Teacher interactions: How affective relationships shape principal
and teacher attitudes, Educational Administration Quarterly, 48, 39-85.
Printy, S. (2010). Principals' influence on instructional quality: Insights from US schools. School
Leadership and Management, 30(2), 111-126.
Printy, S. M. (2008). Leadership for teacher learning: A community of practice perspective.
Educational Administration Quarterly, 44, 187-225.
199

Robinson, V. M., Lloyd, C. A., & Rowe, K. J. (2008). The impact of leadership on student
outcomes: An analysis of the differential effects of leadership types. Educational
administration quarterly, 44(5), 635-674.
Ross, J. A., & Gray, P. (2006). Transformational leadership and teacher commitment to
organizational values: The mediating effects of collective teacher efficacy. School
Effectiveness and School Improvement, 17(2), 179-199.
Sergiovanni, T. J. (1994). Organizations or communities? Changing the metaphor changes the
theory. Educational Administration Quarterly, 30(2), 214-226.
Schneider, B., Carnoy, M., Kilpatrick, J., Schmidt, W. H., & Shavelson, R. J. (2007).Estimating
causal effects: Using experimental and observational designs. Washington, DC: American
Educational Research Association.
Strahan, D. (2003). Promoting a collaborative professional culture in three elementary schools
that have beaten the odds. The Elementary School Journal, 104(2), 127-146.
Supovitz, J.A. (2002). Developing communities of instructional practice. Teachers College
Record, 104(8), 1591-1626.
Supovitz, J., Sirinides, P., & May, H. (2010). How principals and peers influence teaching and
learning. Educational Administration Quarterly, 46(1), 31-56.
Talbert, J. E., & McLaughlin, M. (1994). Teacher professionalism in local school contexts.
American Journal of Education, 102(2), 123-153.
Tschannen-Moran, M. (2001). Collaboration and the need for trust. Journal of Educational
Administration, 39(4), 308-331.
Tschannen-Moran, M., & Hoy, A. W. (2007). The differential antecedents of self-efficacy beliefs
of novice and experienced teachers. Teaching and teacher Education, 23(6), 944-956.
Tschannen-Moran, M., Hoy, A. W., & Hoy, W. K. (1998). Teacher efficacy: Its meaning and
measure. Review of educational research, 68(2), 202-248.
James, C. R., Dunning, G., Connolly, M., & Elliott, T. (2007). Collaborative practice: a model of
successful working in schools. Journal of Educational Administration, 45(5), 541-555.
Urick, A., & Bowers, A. J. (2013). What are the Different Types of Principals Across the United
States? A Latent Class Analysis of Principal Perception of Leadership. Educational
Administration Quarterly, 0013161X13489019.

200

Vescio, V., Ross, D., & Adams, A. (2008). A review of research on the impact of professional
learning communities on teaching practice and student learning. Teaching and Teacher
Education, 24(1), 80-91.
Waters, T., Marzano, R. J., & McNulty, B. (2003). Balanced leadership: What 30 years of
research tells us about the effect of leadership on student achievement (pp. 1-19). Aurora,
CO: Mid-continent Research for Education and Learning.
Wahlstrom, K.L., & Louis, K.S. (2008). How teachers experience principal leadership: The roles
of professional community, trust, efficacy and shared responsibility. Educational
Administration Quarterly, 44(4), 458-495.
Witziers, B., Bosker, R. J., & Krüger, M. L. (2003). Educational leadership and student
achievement: The elusive search for an association. Educational administration
quarterly, 39(3), 398-425.
Woods, A. M., & Weasmer, J. (2004). Maintaining job satisfaction: Engaging professionals as
active participants. The Clearing House, 77(3), 118-121.
Yasumoto, J. Y., Uekawa, K., & Bidwell, C. E. (2001). The collegial focus and high school
students’ achievement, Sociology of Education, 74(3), 181-209.
Youngs, P., & King, M. B. (2002). Principal leadership for professional development to build
school capacity. Educational Administration Quarterly,38(5), 643-670.
Youngs, P. (2007). How elementary principals’ beliefs and actions influence new teachers’
experiences. Educational Administration Quarterly, 43(1), 101-137.

201

CHAPTER 4
CUMULATIVE SYNTHESIS
The main purpose of this chapter is to synthesize results across the three earlier chapters
in this dissertation. Accordingly, the next section succinctly summarizes key findings followed
by a discussion of how these findings contribute to the literature on teacher satisfaction,
leadership and collaboration. Next, I examine whether these findings are similar to or different
from the extant U.S. based literature. Additionally, I address the limitations of the current study
and discuss the policy implications of my findings. The chapter concludes by suggesting
directions for future research.
4.1 Summary of Findings
The guiding question in chapter 1 of this dissertation was - Is there variation in
contextual factors associated with teacher satisfaction in a multi-country setting? Here, I
explored whether variation in student, teacher, and school backgrounds is associated with teacher
satisfaction cross-nationally. Additionally, I examined the organizational context within schools
to parse out which organizational factors matter for teacher satisfaction. This analytic review
revealed several gaps in the literature on satisfaction. First, it indicated that multi-country work
on teacher satisfaction is limited to a handful of studies leaving ample scope for further analysis.
Second, an emphasis on leadership within the U.S. emerged, revealing the importance of this
organizational factor for teacher satisfaction. Third, the review suggested the potential to
explore teacher collaboration in settings other than the U.S. given the paucity of such work
across multiple countries. Substantively, several of the international studies reviewed indicated
that the wider social context of which the school is a part is an important predictor of teacher

202

satisfaction. Thus, the respect and recognition accorded to teachers by the community mattered
for teacher satisfaction internationally.
Chapter 2 in this study capitalized on the results of the analytic review undertaken in the
first chapter by focusing on principals’ leadership style in a multi-country setting. Specifically,
in this chapter, I examined both principal and teacher perspectives on teacher satisfaction for
several countries in Asia and the U.S. after accounting for an extensive range of aggregated
student and teacher demographic variables and school controls. Results of this analysis
(Analysis III (PRTS – school level)) showed that principals’ transformational leadership, is
significantly associated with principal reported teacher satisfaction across all the Asian countries
analyzed as well as the United States (p<.01)30. This result appears to indicate that a visionary
and supportive leadership style as measured by the transformational leadership index is
important for teacher satisfaction in Asia at least from the perspective of the school principals in
these countries. The finding that two dimensions of transformational leadership - the principal’s
role in ‘setting directions’ for the school and in ‘developing people’ were associated with higher
odds of teacher satisfaction in the countries analyzed is consistent with the U.S. literature on
transformational leadership in schools (Bass & Avolio, 1994; Leithwood & Sun, 2012;
Leithwood & Jantzi, 2005; Silins, 1994).
Interestingly, however, this pattern did not hold across all countries31 when I analyzed
transformational leadership factors and teacher perspectives on satisfaction. Teachers of a
representative sample of students in Mongolia, Taiwan and Thailand seem to attach more
importance to student and school background variables (with respect to their satisfaction levels)
than particular aspects of transformational leadership. In Mongolia at least, a plausible reason

30
31

See table 9 in chapter 2.
The exceptions being Hong Kong, Indonesia, and Malaysia.

203

for this divergence may be a heavy emphasis on policy borrowing (See Steiner-Khamsi, 2004 for
an understanding of policy borrowing in this country). It is possible that a similar phenomenon
is reflected in Taiwan and Thailand. Implicit in the notion of policy borrowing is the possibility
that principals in these countries are applying a predominantly western conceptualization of
leadership considered appropriate by their Ministries but that these conceptualizations are less
well understood or internalized by the teachers in these countries because many of these
countries are still in the process of transitioning from centralized to decentralized systems. This
may explain why principals in all the countries analyzed perceive these dimensions to be highly
significant while teachers in this sample (with the exception of Hong Kong, Indonesia and
Malaysia) do not. An alternative explanation is that the divergence in perspectives simply
mirrors the centrality of the school for the principal versus the centrality of the classroom for the
teacher.
Again, drawing on insights from chapter 1, chapter 3 focused on an empirical exploration
of teacher collaboration in five OECD countries – Denmark, Brazil, Mexico, Poland and
Hungary. Much has been written about the purported benefits of teacher collaboration but the
notion of collaboration particularly with respect to leadership and satisfaction internationally has
not received similar attention. I found that different aspects of a principals’ transformational
leadership pertaining to staff development consistent with school goals, developing a positive
school culture, and joint management of classroom issues were significantly associated with the
index of teacher collaboration (indicative of frequency and breadth of collaboration) in the
countries analyzed. This index of collaboration was also significantly and consistently
associated with teacher satisfaction across all five countries. Additionally, I also found that
teacher efficacy (which I used as a control in these regressions) is positively associated with the

204

index of collaboration as well as teachers’ job satisfaction even after controlling for a range of
teacher, school, and student variables.
4.2 Study Contributions
This dissertation informs the literature on teacher satisfaction, leadership, and teacher
collaboration in a number of ways. First, despite a growing awareness of the importance of
context in leadership, it has not been adequately explored previously. By attempting to parse out
the type of leadership that matters for teacher satisfaction in diverse geographic, social and
cultural contexts and comparing these results to what is known about leadership within the U.S,,
this dissertation empirically tests whether specific leadership phenomena that are U.S. based hold
across diverse contexts. As such, it makes a valuable contribution to the field of Educational
Leadership.
Second, only a few scholars (Sargent & Hannum, 2005 and Michaelowa & Wittman,
2007) have attempted to use large-scale survey data to understand and explore the issue of
teacher satisfaction. Large-scale survey data have the advantage of generalizability of inferences
within and across countries which is not necessarily true of self-collected or administrative data.
By using two large-scale education datasets (TIMSS 2007 and TALIS 2008) that are
international in scope, my dissertation addresses this limitation in previous literature on
satisfaction.
Third, empirical analysis especially focused on six Asian countries in this dissertation
addresses the issue of paucity of multi-country research (as identified in Table 1) that is Asiabased.
Fourth, in chapter 3, I show that a collaborative school climate matters for teacher
satisfaction in international contexts. Additionally, my study contributes to the literature on

205

collaboration by deconstructing how different types of collaboration are associated with different
measures of transformational leadership across countries. To date, large-scale empirical analysis
on collaboration that is international in scope has been somewhat limited.
4.3 Comparative Insights
This analysis provides support for the view that organizational variables explain much of
the variation in teacher satisfaction (see Johnson, Kraft & Papay, 2011; Ladd, 2011). Thus, I
find that two aspects of transformational leadership – the principal’s role in ‘setting directions’
and in ‘developing people’ are positively associated with teacher satisfaction in Asia. The
results for Asia do provide support for what we know in a general sense about transformational
leadership within the United States.
Additionally, my analysis in chapter 2 shows that there is no clear and consistent pattern
across all six Asian countries analyzed in the role of teacher demographic and school context
variables vis-à-vis teacher satisfaction. Findings with respect to teacher age and experience vary
by country. For example, table 9 in chapter 2 indicates that principals in Hong Kong, Taiwan
and Thailand perceive the odds of teacher satisfaction are higher when the teacher is more
experienced. In Indonesia, however, this relationship is negative and significant. The literature
on teacher experience within the U.S. is also somewhat contradictory with Perie and Baker
(1997) indicating that very young (less than 3 years of experience) and very experienced teachers
(more than twenty years) have lower levels of satisfaction and Liu and Ramsey (2008)
contending that satisfaction increases with years of experience.
With respect to teacher gender, principals in Hong Kong and Indonesia perceive that the
odds of satisfaction are lower for female teachers. In the other countries (including the U.S.)
however, this variable is not statistically significant. This finding is consistent with earlier U.S.

206

literature that corroborate that the role of gender is not important in explaining variation in
teacher satisfaction as compared to other school organization factors (Culver, Wolfle, & Cross,
1990, Johnson et al. 2011)
Previous research has found that school infrastructure is important for teacher satisfaction
at least in the context of urban schools within the U.S. Results for the U.S. shown in table 9
appear to corroborate this finding. Table 9 shows that within the U.S. there is a positive
correlation between good school facilities and teacher satisfaction. Results in other Asian work
environments are not consistent with this finding perhaps indicating that in these schools
leadership is more important as an explanatory factor for teacher satisfaction. Since not many
scholars internationally have focused on the role of school infrastructure however, further
research is necessary to support these results.
Finally, results from chapter 3 indicate a positive and significant relationship between the
index of teacher collaboration and teacher satisfaction in the five OECD countries analyzed. This
appears to be consistent with some of the U.S. literature that underscores the importance of
teacher collaboration for teacher outcomes (See Culver et al, 1990; Johnson & Birkeland, 2003;
Ladd, 2011; and Rosenholtz, 1990). I also find (in chapter 3) a negative and statistically
significant relationship between student discipline issues and teacher satisfaction that is mirrored
in research that is U.S. based (Ingersoll, 2001; Kennedy, 2005; Liu & Meyer, 2005; Perie &
Baker, 1997; Stockard & Lehman, 2004).
4.4 Policy Implications & Limitations
What are the implications of these findings for leadership behavior in schools? Research
has shown that “principals play an important role in allocating time for teachers to meet and for
providing increased opportunities for job-embedded professional development (Wahlstrom &

207

Louis, 2008, p.463).” Thus, transformational school leaders can help to create a strong teacher
community within schools.
From a policy perspective, this analysis shows that principals would do well to also
address issues of teacher efficacy in their schools given the apparently strong and consistent
association that exists between efficacy and teacher job satisfaction and also between efficacy
and the frequency of teacher participation in a breadth of collaborative activities in schools.
Transformational Leaders can address issues of teacher self-efficacy by providing institutional
support (via sustained opportunities for formal and informal teacher collaboration that enable
shared norms and values among teachers) and verbal endorsement of the job that teachers do.
Finally, this analysis finds that more class time spent maintaining discipline is negatively
associated with teacher satisfaction and is also negatively associated with the index of
collaboration across diverse national contexts. By altering the disciplinary climate in schools,
principals can help to minimize the percent of time teachers spend managing classroom
disruptions and maintaining order.
While the results mentioned above indicate promising evidence of the relationship
between transformational leadership, collaboration and teacher satisfaction, it should be noted
that the analyses presented here have some limitations which warrant discussion. Notably,
although the literature indicates that aspects of transformational and instructional leadership may
overlap (such as creation of a school vision, professional development of teachers etc) care was
taken in chapter 2 to minimize such overlaps in order to avoid issues of multicollinearity. Thus,
investigating the complexity of interrelationships between the different types of leadership
behaviors32 was not attempted in this dissertation.

32

See for example Marks & Printy (2003) who show that instructional leadership may also be transformational in
some instances.

208

Second, TIMSS 2007 data mainly allow analysis of one or at most two teachers per
principal. This deviates from the literature on teacher satisfaction which ideally references a
larger sample of teachers than was available to me via TIMSS. This limitation is somewhat
offset by the fact that the TIMSS data are the only publicly available large-scale data that have
rich contextual information on students, teachers and schools for several countries in Asia.
Third, all the analyses presented in this dissertation are correlational in nature. Thus, the
findings pertaining to the relationships between transformational leadership and satisfaction or
collaboration and satisfaction should not be interpreted in a causal manner. In other words, we
cannot infer from the analyses presented that transformational leadership causes teacher
satisfaction or that participation in collaborative activities causes teachers to be more satisfied
with their jobs.
A final limitation of the study is that it does not include any measures of teacher salary
although there is some international evidence to indicate that salary may be associated with
teacher satisfaction (Zembylas & Papanastasiou, 2004). Neither TIMSS 2007 nor TALIS 2008
include salary information for the teachers sampled. While, country level salary information
may be obtained this figure is not appropriate to use as a control because it lacks within country
variation and is thus essentially a constant in the within country regressions.
4.5 Directions for Future Research
Some of the limitations mentioned may be addressed in future work in one of two ways.
First, the use of qualitative fieldwork may provide valuable insight into the satisfactionleadership relationship by adding finer grained level of information at the teacher level not
generally available through survey data.

209

Second, the use of sub-scales measuring the satisfaction construct rather than a single
measure of satisfaction as used in this analysis may facilitate a more nuanced understanding of
teacher satisfaction.
Third, this study indicated that teacher efficacy is positively correlated with teacher
satisfaction even after controlling for a range of student, teacher, school, and organizational
variables. Building on this result, research into the processes by which teacher efficacy may
influence teacher and student outcomes may be quite valuable for leaders attempting to improve
school effectiveness.
Fourth, we know from chapter 1 that satisfaction is a known antecedent of teacher
turnover, student achievement and teacher commitment. Future research could also pursue an
exploration of the pathways by which teacher satisfaction may influence these other more distal
outcomes.
Additionally, given the apparent importance of the school infrastructure variable noted in
other studies (Kloep & Tarifah, 1994; Rogers-Jenkinson & Chapman, 1990; Sargent & Hannum,
2005) a fruitful area of future research would be to analyze the role of school infrastructure vis-àvis teacher outcomes in the developing country context.

210