Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Understanding teacher well-being and motivation : measurement, theory, and change over time Collie, Rebecca J 2014

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2014_spring_collie_rebecca.pdf [ 1.24MB ]
Metadata
JSON: 24-1.0165878.json
JSON-LD: 24-1.0165878-ld.json
RDF/XML (Pretty): 24-1.0165878-rdf.xml
RDF/JSON: 24-1.0165878-rdf.json
Turtle: 24-1.0165878-turtle.txt
N-Triples: 24-1.0165878-rdf-ntriples.txt
Original Record: 24-1.0165878-source.json
Full Text
24-1.0165878-fulltext.txt
Citation
24-1.0165878.ris

Full Text

UNDERSTANDING TEACHER WELL-BEING AND MOTIVATION:  MEASUREMENT, THEORY, AND CHANGE OVER TIME  by  Rebecca J. Collie  B.Ed. (Hons.), University of Melbourne, 2006 M.A., University of British Columbia, 2010   A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF  THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY  in  The Faculty of Graduate and Postdoctoral Studies (Human Development, Learning and Culture)   THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  February 2014  ? Rebecca J. Collie, 2014     ABSTRACT Teacher well-being and motivation play important roles in teacher and student experiences at school. When teachers are faring well and feeling motivated to teach, they are more effective in their teaching, leave the profession less often, and promote motivation and achievement among their students. In this dissertation, three studies that investigated teacher well-being and motivation were conducted with the aim of advancing our understanding of the two constructs, as well as how they can be promoted among teachers. Study 1 involved conceptualising, developing, and testing the Teacher Well-Being Scale, which measures three factors of teacher well-being: workload well-being, organisational well-being, and student interaction well-being. Among a sample of 603 practicing teachers, results revealed that the new measure functioned similarly across the different demographic groups in the sample and that the three factors of well-being related as expected with other constructs (stress, job satisfaction, and flourishing). Study 2 involved elaborating and testing an explanatory model of teacher well-being, motivation, job satisfaction, and affective organisational commitment that was based in self-determination theory (Deci & Ryan, 1985, 2002). Using the same sample as Study 1, structural equation modelling provided support for the model?s main relationships. In addition, there were some unexpected findings that provide directions for future research (e.g., a double-sided view of autonomy revealing that it can be associated with positive and negative types of motivation). Study 3 involved examining growth curve models of change in teacher well-being and self-efficacy for teaching over two to three months. Among a sample of 71 practicing teachers, the findings showed that teacher well-being was stable over time, whereas self-efficacy for classroom management increased (the other two types of self-efficacy that were examined, self-efficacy ii  for student engagement and instructional strategies, did not change over time). Findings also revealed the significance of the basic psychological needs (autonomy, competence, and relatedness) in predicting teacher well-being and self-efficacy. Taken together, the three studies help to improve our understanding of the highly important variables of teacher well-being and motivation. Implications of the findings for both research and practice are discussed.    iii  PREFACE This dissertation is original, unpublished, independent work by the author, R. Collie. The three studies were approved by The University of British Columbia?s Behavioral Research Ethics Board. Certificate number for Studies 1 and 2: H11-03417. Certificate number for Study 3: H10-02457.  iv  TABLE OF CONTENTS  Abstract ..................................................................................................................................... ii Preface ..................................................................................................................................... iv Table of Contents .......................................................................................................................v List of Tables ......................................................................................................................... viii List of Figures ............................................................................................................................x Acknowledgements.................................................................................................................. xi Dedication ............................................................................................................................... xii Chapter 1. Introduction ..............................................................................................................1 1.1. Well-being ...................................................................................................................... 1 1.2. Motivation ...................................................................................................................... 5 1.3. The Dissertation Research ............................................................................................. 7 Chapter 2. Study 1: Conceptualising and Measuring Teacher Well-being ...............................9 2.1. Literature Review ........................................................................................................... 9 2.1.1. Measurement Method .............................................................................................13 2.1.2. Overview of the Current Study ..............................................................................16 2.2. Methods ........................................................................................................................ 23 2.2.1. Participants .............................................................................................................23 2.2.2. Procedures ..............................................................................................................29 2.2.3. Measures .................................................................................................................34 2.3. Results .......................................................................................................................... 39 2.3.1. Factor Analyses ......................................................................................................39 2.3.2. Demographic Comparisons ....................................................................................49 2.3.3. Preliminary Evidence of Validity ...........................................................................53 2.4. Discussion .................................................................................................................... 56 2.4.1. Research Question 1: What is the Nature of Teacher Well-being? ........................57 2.4.2. Research Question 2: How do Scores on the New Measure Differ Across Demographic Groups? ......................................................................................................63 2.4.3. Research Question 3: Are Interpretations of Scores from the New Measure Supported by Evidence of Validity? ................................................................................65 2.5. Limitations and Future Directions ............................................................................... 67 2.6. Conclusions .................................................................................................................. 71 Chapter 3. Study 2: Developing and Testing a Framework .....................................................73 v  3.1. Literature Review ......................................................................................................... 73 3.1.1. Self-Determination Theory ....................................................................................76 3.1.2. SDT, Job Satisfaction, and Organisational Commitment ......................................82 3.1.3. The Explanatory Model ..........................................................................................88 3.1.4. Overview of the Current Study ..............................................................................97 3.2. Methods ........................................................................................................................ 99 3.2.1. Participants and Procedures ...................................................................................99 3.2.2. Measures .................................................................................................................99 3.3. Results ........................................................................................................................ 104 3.3.1. Data Analyses .......................................................................................................104 3.3.2. Factor Analyses ....................................................................................................106 3.3.3. Structural Modelling ............................................................................................113 3.4. Discussion .................................................................................................................. 121 3.4.1. Basic Psychological Needs, Well-being, and Motivation ....................................122 3.4.2. Job Satisfaction ....................................................................................................126 3.4.3. Affective Organisational Commitment ................................................................128 3.5. Limitations and Future Directions ............................................................................. 130 3.6. Conclusions ................................................................................................................ 132 Chapter 4. Study 3: Teacher Well-Being and Motivation Over Time ...................................134 4.1. Literature Review ....................................................................................................... 134 4.1.1. Teachers? Experiences Over Time .......................................................................135 4.1.2. Overview of the Current Study ............................................................................144 4.2. Methods ...................................................................................................................... 148 4.2.1. Participants ...........................................................................................................148 4.2.2. Procedures ............................................................................................................149 4.2.3. Measures ...............................................................................................................150 4.3. Results ........................................................................................................................ 158 4.3.1. Overview of Data Analysis ..................................................................................163 4.3.2. Results for Multilevel Findings ............................................................................165 4.4. Discussion .................................................................................................................. 188 4.4.1. Self-efficacy for Classroom Management ............................................................188 4.4.2. Teacher Well-Being .............................................................................................190 4.4.3. Need Satisfaction ..................................................................................................192 vi  4.5. Limitations and Future Directions ............................................................................. 195 4.6. Conclusions ................................................................................................................ 196 Chapter 5. General Discussion ..............................................................................................198 5.1. Conclusions and Implications .................................................................................... 202 5.1.1. Contributions for Scientific Understanding .........................................................202 5.1.2. Contributions for Practice ....................................................................................206 5.1.3. Final Conclusions .................................................................................................207 References..............................................................................................................................209 Appendices ............................................................................................................................226 Appendix A: Invitation Letter for Studies 1 and 2 ............................................................ 226 Appendix B: Questionnaire for Studies 1 and 2 ............................................................... 228 Appendix C: Invitation Letter for Study 3 ........................................................................ 237 Appendix D: Questionnaire for Study 3 ........................................................................... 238     vii   LIST OF TABLES Table 2.1. Personal Demographic Characteristics .................................................................. 25 Table 2.2. Job-Related Demographic Characteristics............................................................. 27 Table 2.3. Perceived Socio-Economic Status and Academic Achievement Levels ............... 28 Table 2.4. Demographics of Study Sample and Population ................................................... 29 Table 2.5. Constructs Under Examination in Each of the Three Studies ............................... 36 Table 2.6. Descriptive Information for the Additional Measures ........................................... 39 Table 2.7. Factor Loadings for the Final Exploratory Factor Analysis .................................. 45 Table 2.8. Standardised Factor Loadings from the Confirmatory Factor Analysis ................ 46 Table 2.9. Latent Variable Correlations of the Well-being Factors ....................................... 47 Table 2.10. Descriptive Statistics for Items in the Teacher Well-being Scale ....................... 48 Table 2.11. Descriptive Statistics by Demographic Characteristics ....................................... 51 Table 2.12. Factor Loadings for the Flourishing and Job Satisfaction Items ......................... 54 Table 2.13. Correlations of the Well-being Factors with the Additional Measures ............... 55 Table 2.14. Partial Correlations Controlling for Social Desirability ...................................... 55 Table 3.1. Reliability Indexes, Means, and Standard Deviations of All Variables .............. 103 Table 3.2. Guidelines for Fit Indices .................................................................................... 106 Table 3.3. Factor Loadings from the Factor Analyses .......................................................... 108 Table 3.4. Correlations of Latent Variables .......................................................................... 112 Table 3.5. Standardised Effects on Each Outcome Variable ................................................ 115 Table 4.1. Frequency of Participants? Ethnic Origins .......................................................... 149 Table 4.2. Actual and Potential Sample Size at Each Time Point ........................................ 150 viii  Table 4.3. Items Used to Assess the Constructs ................................................................... 153 Table 4.4. Mean, Standard Deviation, and Range for all Variables ..................................... 156 Table 4.5. Cronbach?s Alphas at the Three Time Points ...................................................... 157 Table 4.6. Differences in Participants Remaining In or Dropping Out of Study ................. 159 Table 4.7. Correlations Between Variables at Time 1 .......................................................... 160 Table 4.8. Correlations Between Variables at Time 2 .......................................................... 161 Table 4.9. Correlations Between Variables at Time 3 .......................................................... 162 Table 4.10. Multilevel Models for Workload Well-Being ................................................... 168 Table 4.11. Multilevel Models for Organisational Well-Being ............................................ 172 Table 4.12. Multilevel Models for Student Interaction Well-Being ..................................... 175 Table 4.13. Multilevel Models for Self-Efficacy for Student Engagement .......................... 178 Table 4.14. Multilevel Models for in Self-Efficacy for Classroom Management ................ 182 Table 4.15. Multilevel Models for Self-Efficacy for Instructional Strategies ...................... 185 Table 4.16. Hypothesised and Observed Need Satisfaction Effects ..................................... 187     ix  LIST OF FIGURES Figure 3.1. The explanatory model of teacher well-being, motivation, job satisfaction, and affective organisational commitment. .................................................................................... 76 Figure 3.2. Summary of the hypotheses tested with predicted relationship polarity. ............. 98 Figure 3.3. Structural equation model of perceived autonomy support, need satisfaction, well-being, motivation, job satisfaction, and affective organisational commitment.. .......... 114 Figure 4.1. Predicted values of workload well-being for different levels of autonomy. ...... 169 Figure 4.2. Predicted values of organisational well-being for different levels of relatedness with colleagues and autonomy. ............................................................................................ 173 Figure 4.3. Predicted values of student interaction well-being for different levels of autonomy. ............................................................................................................................. 176 Figure 4.4. Predicted values of self-efficacy for student engagement for different levels of relatedness with students. ..................................................................................................... 179 Figure 4.5. Predicted values of self-efficacy for classroom management for different levels of relatedness with students and autonomy............................................................................... 183 Figure 4.6. Predicted values of self-efficacy for instructional strategies for different levels of relatedness with students. ..................................................................................................... 186    x   ACKNOWLEDGEMENTS I would like to express my gratitude for the help and support provided by my supervisor, Dr. Jennifer Shapka, and my committee members, Dr. Nancy Perry and Dr. Kimberly Schonert-Reichl. Thank you for providing me with the perfect combination of autonomy and support. I have learnt so much from you about the research process and writing academically.  I would also like to gratefully acknowledge the funding support provided by the Social Sciences and Humanities Council of Canada that enabled me to complete my doctoral research. In addition, I must thank my fellow graduate students who shared their ideas and feedback with me and helped this research take form and reach fruition. Also important are the many teachers who participated in this research. Thank you. Thank you Mum and Dad for a lifetime?s worth of support, and more recently, Dad for reading over the manuscript so thoroughly and finding many of those little errors that were all but invisible to me. Finally, thank you Hugues and L?o for supporting me throughout this endeavour and providing me with not only ample time to work, but also lots of enjoyable distractions.    xi   DEDICATION       For my family and friends, and for teachers everywhere       xii   CHAPTER 1. INTRODUCTION Effective teaching and learning are central goals of our educational system. Although seeking to achieve these goals involves many complexities, there are certain factors that have been shown to support them, two of which are teacher well-being and teacher motivation. Research has shown that these two factors are linked to teaching effectiveness (Duckworth, Quinn, & Seligman, 2009), resiliency in teaching (Klusmann, Kunter, Trautwein, L?dtke, & Baumert, 2008), quality instructional practices (Retelsdorf, Butler, Streblow, & Schiefele, 2010), and students? motivation (Pakarinen et al., 2010; Taylor & Ntoumanis, 2007) and achievement (e.g., Caprara, Barbaranelli, Steca, & Malone, 2006).  Despite knowing that the two variables are critical for effective teaching and learning, there are several gaps in our understanding of teacher well-being and motivation including (a) how teacher well-being can be measured, (b) how teacher well-being and motivation interrelate with one another and other relevant constructs, and (c) how teacher well-being and motivation change over time. The purpose of this dissertation research was to address each of these gaps through three related studies to help improve our understanding of the two variables and to foster optimal well-being and motivation among teachers, which will likely lead to better learning and social-emotional outcomes among students.  1.1. Well-being The construct of well-being is a complex one with several accepted definitions existing in the literature. One of the most well-known definitions is subjective well-being, which is defined as happiness, or more specifically, satisfaction with life and the experience of positive emotions (Diener, 1984). Subjective well-being is a global measure of well-being (i.e., it measures general life well-being and is not domain specific) and is also known as 1   hedonic well-being (e.g., Ryan & Deci, 2001; Vivoll Straume & Vitters?, 2012). It has been examined extensively over the past three decades (e.g., Bowling, Eschleman, & Wang, 2010; Diener, Suh, Lucas, & Smith, 1999). While a great deal of interest in subjective well-being remains, some researchers have begun to question the limitations of subjective well-being. Seligman (2012) has even argued that life satisfaction essentially measures an individual?s cheerful mood at the time of taking the questionnaire. As a result, in recent times researchers have begun to consider a broader definition of well-being: human flourishing (e.g., Deci & Ryan, 2008c; Huppert & So, 2013; Ryan & Deci, 2001; Ryff, 1989; Seligman, 2012; Vivoll Straume & Vitters?, 2012), which is related to eudaimonic well-being (Huppert & So, 2013) and psychological well-being (Ryff, 1989). Like subjective well-being, human flourishing is a global measure of well-being; however, it is defined more broadly.  Human flourishing is defined as ?open, engaged, and healthy functioning? (Ryan & Deci, 2011, p.47). It has also been defined as the ?combination of feeling good and functioning effectively? (Huppert & So, 2013, para. 1). Flourishing is broader than subjective well-being in that it focuses on the wellness or way-of-living that underlies day-to-day emotional experiences while also encompassing life satisfaction and positive affect, which are measured in subjective well-being. For example, an individual may be unhappy about a certain circumstance, while functioning well in their broader life (Ryan & Deci, 2011). According to Ryan and Deci (2011), human flourishing ?conduces to happiness, but does not guarantee it? whereas happiness may be evident when people either are or are not living well? (p. 48). As an example related to teaching, on any given day a teacher may be very unhappy with her or his teaching, but use this as motivation for changing the method of their teaching. After implementing the new strategy, the teacher may be quite happy. The 2   teacher?s use of the negative experience to spur better teaching methods suggests positive functioning, even though the experience involved a negative emotion (unhappiness).  In the interest of supporting robust research efforts, an operational definition of flourishing is needed. One such definition has been provided by Diener and colleagues (2010) that includes eight factors which are considered central to flourishing: purpose and meaning, supportive relationships, engagement/interest, contributing to others, competence, being a good person, optimism, and feeling respected. The current research utilised this definition of human flourishing for the purpose of understanding teacher well-being. More specifically, it examined flourishing in relation to teaching work as the definition for teacher well-being. The importance of well-being among the general population has been well documented. Indeed, philosophers and researchers have been contemplating the issue of well-being for many centuries (Linley, Maltby, Wood, Osborne, & Hurling, 2009). In modern times, this research has generally involved consideration of a global measure of well-being in reference to an individual?s life (including both subjective well-being and human flourishing; e.g., Diener, Emmons, Larsen, & Griffin, 1985). In a recent review, however, Diener (2009) made a call for the advancement of research on well-being in different domains, including the domain of work. This type of contextual well-being is known as work-related well-being and refers to an individual?s positive evaluation of and healthy functioning in their work environment (Van Horn, Taris, Schaufeli, & Schreurs, 2004). Domain specific well-being can be defined using either subjective well-being or human flourishing where these definitions are altered to tap into the context rather than global well-being. As indicated above, the current research was concerned with human flourishing. In 3   particular, it predominantly involved examination of work-related flourishing of teachers?that is, teacher well-being. In addition, global assessments of well-being were also included to further develop our understanding of well-being in relation to teachers.   The issue of teacher well-being has received much attention over the past century (e.g., Borg & Riding, 1991; Kyriacou & Sutcliffe, 1977; Skaalvik & Skaalvik, 2011). This likely stems from the belief that teachers who are functioning well make better teachers. Research supports this belief?teacher well-being, measured through related constructs such as stress and burnout, has been linked with teaching effectiveness (e.g., Duckworth et al., 2009) and quality instructional practices (e.g., Retelsdorf et al., 2010).  Despite knowing that teacher well-being is critical for teachers and students, there are still several gaps in the literature. To begin with, although we have global measures of well-being, there is a need for measures that are dedicated to assessing teachers? work-related well-being based upon their experiences at work. This is a necessary next step so as to gain a better understanding of what aspects of teaching work are aiding or thwarting teacher well-being. Second, despite knowing that teacher well-being is significantly related to teacher job satisfaction and organisational commitment (i.e., commitment to their school of employment; e.g., Collie, Shapka, & Perry, 2012; Jackson, Rothmann, & van de Vijver, 2006; Klassen & Chiu, 2011), we do not know how these variables interact when considered simultaneously and we do not have explanatory frameworks for understanding their interrelationships. This type of examination is important given that teachers? experiences do not occur in isolation (Collie et al., 2012). For example, when teachers feel a sense of job satisfaction, it not only relates to their experiences of well-being (e.g., Klassen & Chiu, 2010), but also their experiences of organisational commitment (e.g., Canrinus, Helms-Lorenz, Beijaard, Buitink, 4   & Hofman, 2012). Finally, for the purpose of developing a comprehensive and more nuanced understanding of teacher well-being, we need to employ diverse methodologies in our research. More specifically, research that involves multiple time points is needed. This dissertation research aimed to address these gaps through a systematic examination of teacher well-being. 1.2. Motivation A second major focus of this dissertation research is teacher motivation. Motivation is the drive to act, think, and develop (Deci & Ryan, 2008a). Effective motivation requires energy or drive, as well as affective and cognitive guidance (Deci & Ryan, 2012). Individuals can be motivated to undertake an activity because they value it or because there is a compelling external reason (Ryan & Deci, 2000). In this dissertation, reference is made to motivational constructs based upon two different theories: self-determination theory (Deci & Ryan, 1985, 2002) and social cognitive theory (Bandura, 1982, 1997). These constructs are more thoroughly defined in the relevant studies throughout this dissertation.  When motivation results from an individual valuing an activity, the individual experiences greater excitement, interest, and confidence, and, in turn, greater performance, persistence, and creativity in relation to the activity (Ryan & Deci, 2000). Motivation, therefore, is highly important in students? learning (e.g., Fortier, Vallerand, & Guay, 1995; Vallerand, Fortier, & Guay, 1997). Motivation is also important in teaching; however, as Richardson and Watt (2010), along with Butler (2007) have established, research into teacher motivation has been largely overlooked in favour of research on students? motivation. In spite of this, it is clear that teacher motivation is important for both teachers and students. More specifically, research has shown that teacher motivation is associated with effective 5   instructional practices (e.g., Butler & Shibaz, 2008), greater occupational commitment, greater retention (e.g., Klassen & Chui, 2011), and reduced emotional exhaustion (e.g., Klassen, Perry, & Frenzel, 2012) among teachers. In addition, teachers who are motivated encourage greater motivation (e.g., Patrick, Hisley, & Kemper, 2000) and achievement (e.g., Caprara et al., 2006) among their students.  Given that teacher motivation is so important, more research is needed to extend our understanding of what influence teachers? experiences of motivation at work. In particular, more systematic and theory-driven research is needed to develop our conceptualisations of teacher motivation (Richardson & Watt, 2010). The aim of the dissertation, therefore, was to address three main gaps in the literature through systematic and theory-driven research. First, although self-determination theory (Deci & Ryan, 1985, 2002) has been used extensively to understand students? motivation, only limited research has examined this theory among teachers. In order to further our understanding of teacher motivation, there is a need for research that examines the relevance of theories like self-determination theory for examining teacher motivation. Second, although research has linked motivation with well-being, job satisfaction, and organisational commitment among teachers, we do not have research or explanatory frameworks for understanding how these variables interact simultaneously. As indicated above, understanding this is necessary to obtain a full picture of teachers? experiences of these variables. Finally, like well-being, there is a need for more research that examines teacher motivation at multiple time points in order to understand how it functions over time.  6   1.3. The Dissertation Research Although research has established that teacher well-being and motivation are critical for teachers and students, further research is needed. The objective of this dissertation research was to address the gaps in the literature, as described above, through three related studies. The first study investigated the measurement of teacher well-being. A well-being instrument based upon teachers? perceptions of core aspects of their work was developed and tested. This instrument is distinct from existing instruments that target global appraisals of well-being (e.g., ?I am satisfied with my life?; Diener et al., 1985), because it asks teachers to rate the degree to which different aspects of their teaching work (e.g., managing the classroom, marking assignments) influence their well-being. Although existing measures are useful for providing understanding about individual functioning overall, we still do not understand what aspects of teaching work are central to teacher well-being. The teacher well-being instrument that I developed in Study 1 addresses this gap. It also builds understandings about underlying dimensions of teacher well-being and how these are driven by aspects of teachers? work.  The second study examined reports of teacher well-being, motivation, job satisfaction, and organisational commitment through an explanatory framework based in self-determination theory (Deci & Ryan, 1985, 2002). As established above, research is needed to ascertain the relevance of self-determination theory for understanding teacher well-being and motivation, and to examine how these two variables interact simultaneously with teacher job satisfaction and organisational commitment. Thus, an explanatory model was described and tested in Study 2 to advance our understanding of teachers? experiences at work and how they interrelate. In addition, the model will help to provide critical information 7   for understanding how efforts to improve teachers? experiences of one outcome may affect their experiences of the others. For example, if schools implement an initiative to improve teacher well-being, the explanatory model will provide a better idea of how this will simultaneously affect teachers? experiences of motivation, job satisfaction, and organisational commitment. The final study in this dissertation examined how teachers? perceptions of well-being and motivation function over time. This type of examination is important for the purpose of gaining understanding of the development or stability of constructs over time. Moreover, given that teachers? experiences at work are constantly changing within (and across) school days, months, and years, this type of research will provide a better understanding of teachers? experiences at work. In Study 3, therefore, I conducted analyses of data collected at three time points with the aim of advancing our knowledge of teacher well-being and motivation, including whether and how they change over time.  Taken together, the three studies address several gaps in the literature and provide a comprehensive and systematic examination of teacher well-being and motivation. Moreover, the three studies help to improve our understanding of the highly important variables of teacher well-being and motivation not only independently, but also as they interact with one another. Although the research focuses on teachers, the implications extend to students and schools. After all, flourishing and motivated teachers encourage greater learning, achievement, and positive social-emotional development among their students (e.g., Caprara et al., 2006; Pakarinen et al., 2010).  8   CHAPTER 2. STUDY 1: CONCEPTUALISING AND MEASURING TEACHER WELL-BEING 2.1. Literature Review Teaching is a rewarding profession that involves meaningful work fostering growth and development among students, as well as the experience of joy and inspiration from watching those individuals grow. At the same time, however, teaching can be challenging due to the complex nature of the job and the myriad demands placed upon teachers. The impact of teaching work on teachers has been a topic of research for many decades and, in some cases, the findings have been quite concerning. Namely, research has shown that up to one-third of teachers are stressed or extremely stressed (Geving, 2007; Collie, 2010; Thomas, Clarke, & Lavery, 2003), and that between 25 and 30% of teachers leave the profession within the first five years of teaching (Darling-Hammond, 2000; Ingersoll, 2002; Ingersoll & May, 2012). As a result of these concerning trends, there has been increasing interest in the well-being of teachers and how their working lives can be improved to promote optimal well-being. This area of research is important not only for teachers, but also for students. After all, well-being among teachers has been associated with effective teaching (Duckworth et al., 2009; Klusmann et al., 2008), organisational commitment (Jackson et al., 2006), and perceived teaching efficacy (Lent et al., 2011). It has also been associated with students? attachment to school (Wei & Chen, 2010) and motivation for learning (e.g., Pakarinen et al., 2010). Given that these outcomes are associated with significant financial, social-emotional, and academic ramifications, improving our understanding of the well-being of teachers is highly important not only for teachers, but for all members of our school communities. 9   Over the past 50 years, the well-being of teachers has been examined in research through several related constructs. A prevalent construct featured in the literature is teacher stress. As indicated above, teaching is considered a highly stressful career (Kyriacou, 2001). Research has shown that high levels of stress are associated with negative outcomes for teachers and students, whereas low levels of stress are associated with positive outcomes. Collie, Shapka, and Perry (2012) found that higher levels of teacher stress due to student behavioural issues were associated with lower self-efficacy for teaching. Moreover, Pakarinen and colleagues (2010) showed that lower levels of teacher stress predicted greater learning motivation among kindergarten students.  Another commonly examined construct is teacher burnout. Burnout is a multidimensional construct composed of emotional exhaustion, depersonalisation, and personal accomplishment (Maslach & Jackson, 1981). The literature is replete with evidence of the negative effects of burnout on teachers (e.g., Schaufeli & Bakker, 2004; Steinhardt, Smith Jaggers, Faulk, & Gloria, 2011; Wolpin, Burke, & Greenglass, 2002). In a model looking at job demands and resources, Schaufeli and Bakker (2004) examined teachers? perceptions of burnout, work engagement, and intentions to leave their job. They found that burnout was negatively associated with work engagement and that it positively predicted intention to leave. In other research, Steinhardt, Smith Jaggers, Faulk, and Gloria (2001) showed that burnout had a mediating effect in the relationship between chronic teacher stress and depressive symptoms. Teachers who felt stressed, reported higher burnout, and higher depressive symptoms as a result. Finally, in a longitudinal study, Wolpin, Burke, and Greenglass (2002) found that perceived burnout resulted in decreased in job satisfaction over time among teachers. 10   Another variable that is more removed, but that has also been used as a proxy for well-being is teacher job satisfaction. Klusmann and colleagues (2008) defined teacher well-being as emotional exhaustion and job satisfaction. They found that teachers reporting the lowest emotional exhaustion and highest job satisfaction had students who felt the most competent and autonomous. In other research, Van Horn, Taris, Schaufeli, and Schreurs (2004) used job satisfaction as part of their definition for affective well-being in a five dimensional model of occupational well-being. Along with job satisfaction, their model included measures of stress, burnout, and several other variables (e.g., organisational commitment) to assess affective, professional, social, cognitive, and psychosomatic well-being. Taken together, the findings from these studies have provided us with important information about teachers? experiences at work. However, we still do not know much about the construct of teacher well-being itself. That is to say, these variables relate highly to well-being, but they are different from the construct of well-being, which in this dissertation is defined as flourishing (i.e., ?open, engaged, and healthy functioning;? Ryan & Deci, 2011, p. 47). In fact, very little research has been conducted on the construct of teacher well-being itself (i.e., without defining it by using related constructs). This is surprising given the large body of literature that has examined well-being among individuals more generally and provided evidence highlighting its importance for individual and societal functioning (e.g., Tov & Diener, 2009).  In order to extend our understanding of the well-being of teachers, there is a need for research that directly examines the construct of teacher well-being. This type of direct examination will not only broaden our knowledge of well-being, but also our knowledge of 11   related variables like teacher stress, burnout, and job satisfaction. Furthermore, directly examining the construct of teacher well-being has the potential to provide a different picture of well-being than related constructs. For example, job satisfaction looks at how satisfied a teacher is with his or her work, but this does not necessarily include an aspect of personal flourishing that is associated with well-being as defined here. Moreover, by using stress as a proxy for well-being, we may inadvertently disregard the positive aspects of teachers? work experiences. For example, Brenner, Perry, and Collie (2012) and Collie, Perry, and Brenner (2013) found that student teachers reported moderately high levels of stress while still reporting high levels of engagement in and commitment to the profession of teaching. This suggests that these student teachers were faring well (as evidenced by the high levels of engagement and commitment) in spite of experiencing relatively high stress levels. Furthermore, it underscores the need for further research to better understand well-being and how it differs from stress, which echoes Huppert and So?s (2013) suggestion that we need to study well-being on its own and not simply as the absence of negative affect. More recently, researchers have begun to examine the well-being of teachers directly by using subjective well-being measures, such as life satisfaction, positive affect, and negative affect (e.g., Albuquerque, Pedroso de Lima, Figueiredo, & Matos, 2011; Duckworth et al., 2009). These studies have shown that teachers report average (Albuquerque et al., 2011) to above average life satisfaction (Duckworth et al., 2009), but that experienced teachers are less satisfied with life than beginning teachers (Duckworth et al., 2009). Taken together, these studies reveal information about the subjective well-being of teachers. We still do not have a thorough understanding, however, of well-being defined as flourishing, nor of the well-being of teachers as it relates specifically to their work. In other words, global 12   well-being?whether defined as life satisfaction or flourishing?tells us how a group of individuals who happen to be teachers are faring in their life. However, it does not tell us how the well-being that teachers experience is related to aspects of their work. This is the focus of Study 1.  There has been an increasing call for the examination of well-being in different contexts such as work (Diener, 2009). In addition, worldwide research conducted by Gallup (Rath & Harter, 2010) revealed five domains of well-being, of which work-related well-being was the most powerful at determining overall life well-being. This highlights the importance of work-related well-being?not only is it important for individuals? outcomes at work, but it also relates strongly to their broader life experience. Therefore, research that specifically focuses on the work context of teachers is needed to extend our understanding of teacher well-being. Moreover, measures that tap directly into teachers? experiences at work will be helpful in understanding the different dimensions of teacher well-being and the aspects of teachers? working lives that affect their well-being, both positively and negatively. 2.1.1. Measurement Method One method for measuring teacher well-being is by creating an instrument that taps into the construct itself. This method focuses on individuals? global appraisals of well-being and is the format used in the well-known Satisfaction with Life scale (e.g., ?In most ways my life is close to my ideal;? Diener et al., 1985), and the more recently developed Flourishing scale (e.g., ?I am a good person and live a good life;? Diener et al., 2010). These instruments provide researchers with an overall score of participants? well-being that can be used to understand how an individual is faring generally. However, they do not provide information 13   about what specific aspects of the individual?s life influenced his or her score of well-being, nor do they provide understanding about well-being specifically related to work.  A second method for measuring teacher well-being is by examining the impact of teachers? experiences at work on their well-being. This involves asking teachers to rate the extent to which different aspects of their teaching work (e.g., attending meetings, student behaviour) affect their well-being. In teacher research, this type of questioning has commonly been used to examine teacher job satisfaction (e.g., Dinham & Scott, 1998) and stress (e.g., Clunies-Ross, Little, & Kienhuis, 2008; Klassen & Chiu, 2011), and provides a relevant method for gaining insight into the core aspects of teaching work that affect teacher well-being. An important feature of this type of measure is that it highlights tangible factors that administrators and schools can address to better support teacher well-being.  In Study 1, the second method for assessing well-being was used to create an instrument to measure teacher well-being. The reason for using this method as opposed to the first method is that it provides information about the aspects of teaching work that support or thwart teacher well-being, while at the same time it has the ability to provide an overall well-being score like the first method when the items are combined to form a composite variable.  A limitation of measures that have used the second method, however, is that they do not often consider whether aspects of teaching work impact teachers negatively and/or positively. Generally, these scales only allow teachers to indicate one polarity. For example, a commonly used stress measure, the Teacher Stress Inventory (Boyle, Borg, Falzon, & Baglioni, 1995), asks teachers to rate the level of stress that they experience from different aspects of their work on a scale from no stress to extremely stressful. The total score from all items provides an overall stress score for each teacher. Although this scale allows teachers to 14   indicate that certain aspects of their work have no stress, it does not take into account the fact that stress can be a motivating force (e.g., seen as a challenge rather than a threat; Hobfall, 1989) or that some of the items may actively reduce teacher stress. For example, by only allowing teachers to report whether relationships with colleagues create no stress, the scale does not allow us to determine if these relationships actually help to reduce teacher stress by providing teachers an opportunity to express or share their experiences/feelings. Similarly, by asking teachers to what extent various aspects of their work promote burnout, we neglect to consider whether any of these aspects are energising and do the opposite. Although one-dimensional views of stress and burnout have provided us with important information, the consideration of positive and negative views is needed to extend our understanding. Furthermore, although teacher well-being is defined differently from stress, it is also affected positively and negatively by teachers? experiences. Clearly, instruments that consider the full range of influences on teacher well-being are necessary for the advancement of the field.  Emerging research from northern Europe is beginning to examine teacher well-being through the second method while also allowing teachers to assess the different aspects of their work as negatively or positively affecting their well-being. More specifically, Konu, Viitanen, and Lintonen (2010) asked Finnish teachers about four categories of school welfare: school conditions, social relationships, means of self-fulfillment, and health. Teachers were asked to respond to the items (e.g., ?I have friends in this school?) on a scale that considered positive and negative influences?from totally agree to I completely disagree. Konu and colleagues found that teachers rated means for self-fulfillment (e.g., teaching effectively, support from colleagues) most positively and school conditions (e.g., size of classrooms, safety of facilities) most negatively in relation to their welfare.  15   Although this research provides some understanding of teacher well-being, there are several additional factors not included in Konu and Lintonen?s (2004) research that need to be addressed. For example, attention needs to be paid to the impact of classroom management on teacher well-being given that it has been shown to be a key cause of stress among teachers (e.g., Collie et al., 2012; Klassen & Chiu, 2010). In addition, attention needs to be paid to issues related to teachers? workloads such as staying late after hours or the overcrowded curriculum that are significant concerns for teachers (e.g., Naylor, 2001). These are just a couple of examples that highlight the need for further research in this area so as to develop our understanding of teacher well-being.  This type of examination will help to extend our understanding of what factors influence the well-being of teachers and it will help place teacher education programs, schools, school boards, and policy-makers in a better position to promote optimal well-being among their teachers. Study 1, therefore, involved developing and testing a measure of teacher well-being based upon teachers? experiences at work. The measure was designed to give teachers the option to report whether different aspects of their teaching work influence their well-being positively or negatively so as to address the limitations discussed above. 2.1.2. Overview of the Current Study Study 1 involved the development and testing of the Teacher Well-being Scale (TWBS), which aims to provide teachers with the opportunity to report how key aspects of their working lives impact their well-being in positive or negative ways. The study involved measure development, testing, and refinement, as well as comparisons of scores on the new measure across demographic groups and with related measures. Three research questions were used to guide the investigations: 16   1. What is the nature of teacher well-being? 2. How do scores on the new measure differ across demographic groups? 3. Are interpretations of scores from the new measure supported by evidence of validity? 2.1.2.1. Measure development, testing, and refinement. Following, Miller, McIntire, and Lovler?s (2011) steps for measure development (further details are provided below), a large pool of items that covered key aspects of teaching work were collated to form the TWBS, which was then administered via an online questionnaire to a sample of practicing teachers. Items were chosen to cover a broad range of teachers? experiences at work. In addition, they were chosen based upon two theories to help determine if existing theory is relevant for understanding teacher well-being.  The first theory used to inform item development was self-determination theory (SDT; Deci & Ryan, 1985, 2002). SDT posits that optimal motivation and well-being within individuals is promoted through the satisfaction of three basic psychological needs: autonomy, competence, and relatedness (Deci & Ryan, 1985, 2002). Autonomy refers to people?s perception that they are the origin or source of their behaviour and have some choice or input into the activities they undertake (Deci & Ryan, 2002; Filak & Sheldon, 2008). Competence refers to people?s sense that they are effective in undertaking and mastering challenges and able to exercise and express their capacities (Deci & Ryan, 2002; Filak & Sheldon, 2008). Finally, relatedness refers to people?s need to feel connected to others and the community, and to care for and be cared for by those others (Deci & Ryan, 2002). In work-place research, relatedness generally refers to colleagues; however, for teachers it also refers to relatedness with students (Klassen, Perry, et al., 2012). Research 17   suggests that the three basic psychological needs do, in fact, relate to well-being and motivation among students (e.g., Standage, Duda, & Ntoumanis, 2005) and employees in business settings (e.g., Deci et al., 2001). Emerging research suggests that the same is true for teachers (e.g., Klassen, Perry, et al., 2012). Given the relevance of SDT for examining teacher well-being, items that tap into the three basic psychological needs were included in the pool to see if they are relevant for teacher well-being as measured by teachers? experiences at work.  The second theory was Bronfenbrenner?s (1978, 1994) socio-ecological theory, which posits that individuals reciprocally influence and are influenced by the different environments or systems of which they are a direct or indirect part (Bronfenbrenner, 1994). These include micro-, meso-, exo-, macro-, and chronosystems (Bronfenbrenner, 1994). Microsystems are the different environments in which the individual is an immediate part (e.g., home, school). Mesosystems involve the processes or relations linking two or more microsystems (e.g., relations between home and school). Exosystems involve the processes or relations linking two or more microsystems where the individual only participates in one of the systems?the other systems affect the individual indirectly (e.g., relations between school and the district-level school board). Branching further out, macrosystems concern the cultural characteristics that influence each of the smaller systems (e.g., belief systems, values). Finally, chronosystems concern the changes in the individual and the environment over time. As individuals go about their daily lives, there is reciprocal influence between them and the different systems. In the current study, items that relate to relevant microsystems (e.g., classroom, school, home), mesosystems (e.g., relations between school and home), exosystems (e.g., students? online social networking), and macrosystems (e.g., 18   the greater community) were included. The chronosystem was not considered given that this measure examines teachers? experiences at one point in time (however, this would be an interesting consideration when attempting to understand how and why teachers? reports may change over time). Following data collection, factor analyses were used to refine the measure and ascertain its dimensionality (Research Question 1). Descriptive statistics for the final instrument as a whole and as individual items were calculated, including the mean, standard deviation, observed range, skewness, and kurtosis. In addition, Cronbach?s alphas were calculated as an indicator of internal consistency. 2.1.2.2. Demographics comparisons. Group differences in scores on the TWBS were examined based on several demographic variables (Research Question 2). These variables were sex, years of teaching experience, school level (elementary, secondary), school setting (urban, suburban, small town, rural), and teachers? perceptions of students? socio-economic status and academic achievement. The demographic comparisons were conducted to provide preliminary understanding of how teacher well-being functions among different groups. The presence of group differences would provide information about how different groups of teachers experience well-being with the potential to facilitate targeted attempts to improve teacher well-being. 2.1.2.3. Preliminary evidence of validity. The Standards for Educational and Psychological Testing (Standards; American Psychological Association [APA], American Educational Research Association [AERA], and National Council on Measurement in Education [NCME], 1999) establish that ?validity refers to the degree to which evidence and theory support the interpretations of test scores? (p. 9). They also establish that validation is a 19   continuing process involving the accumulation of evidence based on several sources to provide a sound basis for the interpretation of scores. Study 1 begins the process of accumulation of validity evidence for the new measure by providing evidence based on two sources included in the Standards. The first is evidence of validity based on internal structure, which indicates the extent to which the relationships among the instrument?s items conform to the construct under examination (APA, AERA, & NCME, 1999). The second is evidence of validity based on relations to other variables, which indicates the extent to which relationships with other similar or different variables are consistent with the construct under examination (APA, AERA, & NCME, 1999). For this second source, convergent evidence of validity was provided by comparing scores on the new measure with other similar constructs that were assumed to be related (both positively and negatively).  Evidence based on internal structure was provided through Cronbach?s alphas and factor analyses. Cronbach?s alpha provides an understanding of the internal consistency of the measure by ascertaining the extent to which items in a test are interrelated (Cortina, 1993). A higher coefficient alpha reflects a higher expectation of reliability of the measure. In the case of internal consistency, this suggests that there is consistency in the results across items in the test. Factor analyses provide an understanding of the underlying structure in a measure by revealing subsets of items that are internally related, but relatively independent of other subsets (Tabachnick & Fidell, 2007). These subsets, or factors, reflect underlying processes (latent variables) that have prompted the correlations among the items within each factor (Tabachnick & Fidell, 2007). Convergent evidence was provided by comparing scores on the TWBS with three previously published measures: the Teacher Stress Inventory (Boyle et al., 1995), an 20   unnamed 4-item job satisfaction scale (Caprara, Barbaranelli, Borgogni, & Steca, 2003), and the Flourishing Scale (Diener et al., 2010; Research Question 3). In Study 1, stress was defined as teachers? experiences of negative emotions as a result of their teaching work (Kyriacou, 2001). Job satisfaction was defined as the fulfilment gained from working as a teacher (Locke, 1969). Finally, as previously described, human flourishing was defined as healthy human functioning (Diener et al., 2010; Ryan & Deci, 2011). This was a global measure of well-being (i.e., about the individual?s life as a whole and not specifically related to teachers? work). In order to obtain convergent evidence of validity, several different hypotheses were elaborated. First, it was expected that teacher well-being would have a moderately negative relationship with teacher stress (Hypothesis 1). The rationale for this is that when employing a one-dimensional definition as I have here, stress by its very nature of causing people to experience negative emotions (e.g., Kyriacou, 2001) affects well-being negatively. The rationale for the moderate strength of this relationship is that aspects of teaching work may influence teacher stress and well-being differently. In other words, it is possible that some aspects may associate positively with teacher stress, while at the same time, also associate positively with teacher well-being. For example, teacher-student relationships can be stressful for teachers, but also positive at the same time given the varying relationships that teachers have with different students, at different times, and with respect to different parts of the curriculum.  The second hypothesis was that teacher well-being would have a moderately positive relationship with job satisfaction (Hypothesis 2a). The rationale for this is that teachers who are experiencing well-being at work will tend to feel that their work-related needs are being 21   met and have experiences with which they are happier about. This, in turn, will relate positively to their sense of job satisfaction. This hypothesis is supported by previous research. Bowling, Eschleman, and Wang (2010) conducted a meta-analysis of research involving job satisfaction and subjective well-being among employees in a variety of settings (e.g., university instructors, customer service employees, physicians) and found a moderately positive relationship (r = .36) between the two variables. Given that subjective well-being concerns an individual?s life in general rather than being specific to his or her work, it is hypothesised that the relationship between job satisfaction and teacher well-being in the current study will be slightly higher than Bowling and colleagues? reported correlation because job satisfaction and teacher well-being both refer to the working context (Hypothesis 2b). The third hypothesis was that teacher well-being would have a moderately positive relationship with flourishing (Hypothesis 3). The rationale for this hypothesis is that both teacher well-being and flourishing are well-being constructs; however, teacher well-being is a contextual construct specifically related to work (i.e., work-related well-being), whereas flourishing is related to an individual?s life in general (i.e., it is a global measure). There will not be a perfect relationship, therefore, between the two constructs. For example, an individual may feel he or she leads a meaningful life (flourishing), while at the same time his or her teacher well-being may be negatively impacted by workload.  In addition to the constructs measured to provide convergent evidence of validity, social desirability was also measured. In psychological measurement, social desirability refers to the tendency of individuals to endorse questionnaire items they perceive to be socially acceptable and not endorse items they perceive to be socially unacceptable (Spector 22   & Brannick, 2009). Social desirability can mean that responses tend to be more positive than they are in actuality. Measuring this construct provides an understanding of the extent to which participants? responses are biased by the desire to appear in a positive light (Uziel, 2010). Put another way, controlling for this bias provides an understanding of the results while excluding the influence of social desirability. Although controlling for social desirability is not commonly done in teacher research, it is regularly mentioned as a necessary inclusion in future research in the literature to add to the robustness of self-report data collection (e.g., Retelsdorf et al., 2010). The handful of studies that have controlled for social desirability have found small, but significant effects of social desirability on teachers? responses (e.g., Brackett, Palomera, Mojsa-Kaja, Reyes, & Salovey, 2010; Roth, Assor, Kanat-Maymon, & Kaplan, 2007). Other research, however, has suggested that social desirability is less of an issue in anonymous online questionnaires than in paper questionnaire (Joinson, 1999). Study 1 included a measure of social desirability to see whether it played a role in teacher well-being. In addition, this inclusion helps to temper the limitation of single-source bias that occurs in self-report research (Podsakoff, MacKenzie, Lee, & Podsakoff, 2003). 2.2. Methods 2.2.1. Participants Participants included 603 teachers from four different district-level teachers? associations in British Columbia, Canada. The sample sizes for the four districts were 135, 138, 146, and 184 and ranged from 12% to 23% of the district populations (support for the representativeness of these samples is provided below). Participants were from 218 different schools (19% of participants chose not to identify their school) and had an average teaching 23   experience of 15.8 (SD = 9.93) years and an average age of 44.2 (SD = 10.83) years. Table 2.1 shows the frequencies of personal demographic characteristics among the participants including sex, birth country, ethnic origins, and previous education. It reveals that most of the sample was female (76%), born in Canada (86%), had northern and western European ethnic origins, and had completed a bachelor?s degree. Participants also reported having one (86%), two (10%), or three (1%) ethnic backgrounds (3% chose not to answer this question). Table 2.2 shows the frequencies of job-related demographic characteristics including teaching level, role, time spent teaching, and school setting. Overall, it reveals that most of the participants taught at the elementary level (53%), were classroom teachers (81%), spent the majority of their working time teaching (89%), and worked in a suburban setting (48%). Participants were also asked to rate the average socio-economic status (SES) of students and their families and the average academic achievement of students at their schools compared to most other people/schools in the province. Teachers indicated that they taught students from the full range of SES and academic achievement levels. Table 2.3 shows the frequencies of the different levels reported.   24    Table 2.1 Frequencies of Personal Demographic Characteristics of the Participants   Frequency Sex  Female 76% Male 21% Birth country  Canada 86% United Kingdom 4% Mainland Europe 2% United States 2% Asia 1% Othera 2% Ethnic origins  Northern and western European 82% Eastern and southern European 16% Aboriginal 3% East Asian 3% South Asian 2% Other 4%   25   Table 2.1 (Continued)    Frequency Previous education  Bachelor 51% Master 30% Post-graduate diploma or extra credits 12% Undertaking a master?s degree 4% Doctorate 1% Note. Where percentages do not add up to 100, the remaining participants chose not to respond. Northern and Western European origins refers to British, Scottish, German, Swedish, Danish, Norwegian, Dutch, etc. Eastern and Southern European origins refers to Polish, Russian, Ukrainian, Italian, Greek, Spanish, etc. Aboriginal origins refers to First Nations, Inuit, Metis, etc. East Asian origins refers to Chinese, Japanese, Korean, etc. South Asian origins refers to East Indian, Punjabi, Pakistani, etc. a The other category included Africa, South and Central America, the Middle East, and Oceania.   26    Table 2.2 Frequencies of Job-Related Demographic Characteristics of the Participants   Frequency Teaching level  Elementary school 53% Middle school 29% Secondary school 10% Multiple levels 8% Teaching rolea  Classroom teacher 81% Support teacherb 12% Substitute teacher 3% Teacher librarians 3% Other rolesc 1% Time spent teachingd  0-50% of working time  8% 51-75% of working time  13% 76-100% of working time  76%   27   Table 2.2 (Continued)   Frequency School setting  Urban 34% Suburban 48% Small town 10% Rural 7% Note. Where percentages do not add up to 100, the remaining participants chose not to respond.  a All participants were teachers or undertook teaching roles in addition to their other positions and were, therefore, classified as teachers in the results. b Support teachers refers to resource teachers, special education teachers, and counsellors.  c Other roles included administrators (e.g., principals) and distance educators. d This refers to how much of the teachers? work was spent teaching (as opposed to doing administrative tasks for example)    Table 2.3 Frequencies of Perceived Socio-Economic Status and Academic Achievement Levels   Low Low-average Average High-average High SES of students 18% 21% 39% 15% 5% Academic achievement 9% 20% 45% 19% 4% Note. SES = socio-economic status.    28   In order to gain a sense of the representativeness of the sample, the sample characteristics were compared to the district populations from which they were drawn on three variables: gender, age, and years of teaching experience. Table 2.4 provides the data comparing individuals participating in Study 1 with the population (i.e., the four districts involved in the study). As can be seen from Table 2.4, the sample data were very similar to the population parameters, which offers support for the representativeness of the sample. Details about the procedures for participant recruitment appear in the following section after the description of measure development.  Table 2.4 Demographics of Study Sample and Population  n Female Average age Average years of experience Study 603 78% 44.2 15.8 Population 3700 71% 44.8 12.7 Note. Population columns represent combined data/averages for the four districts involved in the study (Ministry of Education, 2012)   2.2.2. Procedures 2.2.2.1. Measure development. In order to develop psychological instruments, there are several recommended steps that should be followed. In the current study, I followed steps that Miller et al. (2011) have outlined. The first step involves defining the test universe (i.e., the content that the researcher wants to measure), target audience, and test purpose. The test universe is a broad range of teachers? daily work experiences, the target audience is 29   practicing teachers, and the purpose of the test is to gain an understanding of how individual factors affect teacher well-being as well as an overall score of well-being.  The second step involves creating an operational definition of the construct, the format for the questions, and how the test will be scored (Miller et al., 2011). The operational definition of teacher well-being is the definition noted in the Introduction (i.e., flourishing at work involving the factors of purpose and meaning, supportive relationships, engagement, contributing to others, competence, being a good person, optimism, and feeling respected). Because I was interested in teachers? own interpretations of how aspects of their work affect their well-being, a simplified definition of this was included in the test instructions (details below). The format for the questions was determined as a scale ranging from negatively to positively such that teachers could indicate the degree to which different aspects of their work affected their well-being in positive and negative ways. A 7-point Likert-type scale was chosen to score the test. The third step involves composing the test items (Miller et al., 2011). At this point, a pool of items based on the many different aspects of teaching work (e.g., interpersonal relationships, workload, leadership support) was created. The items included in the initial pool were adapted from related instruments including the Teacher Stress Inventory (Boyle et al., 1995) and the Job Satisfaction Scale (Spector, 1997), as well as additional experiences not covered by existing measures (e.g., interacting with the parents of students, lesson planning). These items were chosen to give a broad representation of teachers? work experiences. In addition, item selection was guided by SDT (Deci & Ryan, 1985, 2002) and socio-ecological theory (Bronfenbrenner, 1978) to provide an understanding of whether these 30   two theories are relevant for the consideration of teacher well-being as measured by the new instrument.  Because self-report questionnaires rely on participants? accurate interpretations of the items, several steps were taken to build the strength of the questionnaire. First, language that is familiar to teachers in British Columbia was used. Karabenick and colleagues (2007) identified this as an important consideration in the development of self-report items and explained that unfamiliar language can negatively affect the participants? understanding of the item and, thus, reduce the ability of the item to accurately measure the phenomenon of interest. Similarly, examples were provided in parentheses for items that may have been ambiguous (e.g., ?Relations with students not in my class [e.g., in hallways, etc.]?). This consideration in item design can help to improve participant comprehension and reduce measurement error (Podsakoff et al., 2003). Furthermore, items were described in straight-forward, non-pejorative language to increase the likelihood of truthful responses (Simms, 2008).  The fourth step in measure development involves writing the test instructions (Miller et al., 2011). For this, teachers were asked, ?Currently, how do the following aspects of being a teacher affect your well-being as a teacher?? As you will have noticed, the item stem asked participants about their current experiences, rather than their experiences over a previous time period. This is an important step in reducing the cognitive load required for remembering (Winne, Jamieson-Noel, & Muis, 2002) and aids in the accuracy of responses. However, there was an exception to this. At the time of data collection, teachers were involved in industrial action (called job action in British Columbia) due to stalled negotiations with the provincial government about a new work contract. This industrial 31   action meant that there was a reduction in some administrative tasks normally completed by teachers (further details are provided below). In this circumstance, if teachers were not engaging in certain activities due to industrial action, they were asked about their previous experiences.1  As described above, the instructions also included a brief definition of well-being: ?Well-being refers to open, engaged, and healthy functioning as a teacher.? In order to aid understanding, an example relating to the first item about relations with teaching colleagues was also provided: ?For example, in your current working environment how do your relations with fellow teachers influence your well-being? Do your relations have a positive or negative influence on your well-being as a teacher??  The fifth step of measurement development involves piloting the instrument (Miller et al., 2011). Using the large pool of items, a small sample of practicing teachers (n = 5) was invited to examine the items and suggest additional aspects of teaching work not covered. This led to the sixth step, which involves analysing the items and making changes based on results from pilot testing (Miller et al., 2011). Two additional items were suggested leading to a final pool of 60 items that was used in data collection for the study. The seventh step involves collecting evidence of validity for the test (Miller et al., 2011). This reflects the investigations conducted for Research Question 3. 2.2.2.2. Measure testing. The TWBS was tested among practicing teachers who were recruited from four district-level teachers? associations in British Columbia. The four districts enabled the collection of data from a variety of school settings and locations. 1 Although this contradicts the aim to focus on current experiences in the measure, it was necessary to refer to prior experiences in data collection given the constraints placed on teachers by the industrial action (i.e., some tasks were not being undertaken at the time of data collection).  32                                                    Presidents of the four teachers? associations were emailed details of the study including the invitation letter for teachers (Appendix A shows the invitation letter). The research instrument involved an online questionnaire. Presidents forwarded the study invitation letter?including the URL to the online questionnaire?to teachers who then decided whether or not they wanted to participate Teachers were given three weeks to complete the questionnaire and were sent an email reminder after two weeks. Data was collected during the 2011/2012 school year. Although the nature of the recruitment process did not allow for the calculation of accurate response rates (i.e., it is not possible to ascertain how many teachers actually viewed the study?s invitation email), there are several factors, all of which have been used successfully in the past (e.g., Collie et al., 2012; Mertler, 2002), that provide confidence in the representativeness of the findings. First, the demographics of the study?s sample were very similar to the population (see Table 2.4). Second, average scores found in the current study are comparable to those found using similar measures in other studies (see below). Finally, Mertler (2003) conducted research showing that teachers? responses to paper and online questionnaires were not significantly different, revealing that both methods obtain similar data and offering support for online questionnaires as an appropriate method for data collection among teachers.  As noted earlier, during the time of data collection teachers were involved in industrial action over contract negotiations with the provincial government. This industrial action restricted their work to varying degrees (i.e., it involved work slowdown as opposed to a full strike). All teachers were teaching their regular classes; however, most teachers were not undertaking administrative duties such as writing report cards and attending meetings throughout the data collection timeline (discussed in further detail in the limitations section). 33   Given this situation, teachers were provided with the following statement in the instructions for the TWBS: ?If the current job action means that you are not currently engaging in some of the different items below, please answer based on your experiences prior to job action, or if you are a beginning teacher, please leave these items blank.?  In addition, comparative measures were used to help ascertain the representativeness of the data and these suggest that the results are similar to findings in previous studies despite the industrial action that was taking place. For example, the mean for stress in the current study was 3.53 (see Table 2.6). This falls within the range of average stress scores among Canadian teachers in other studies. Teachers in Klassen and Chiu?s study (2011) reported stress levels that corresponded to means ranging from 3.45 to 3.78 out of 5. Furthermore, Collie (2010) found a similar mean (M = 3.24, not significantly different) among a sample from the same population as the current study (i.e., British Columbia teachers) in data collected when industrial action was not taking place. Similarities also exist for teacher job satisfaction (Klassen, Usher, & Bong, 2010), and flourishing (Diener et al., 2010). These comparisons provide support for the fact that the industrial action did not significantly affect teachers? responses. 2.2.3. Measures Participants completed an online, self-report questionnaire that included the new measure, along with the measures of stress, job satisfaction, flourishing, and social desirability to provide evidence for representativeness and validity. In addition, there were questions pertaining to demographics. Table 2.5 shows the constructs that were examined in Study 1, along with those examined in Studies 2 and 3. It also shows the location in the Appendices in which the items for the different scales can be found.  34   2.2.3.1. Teacher well-being. The TWBS consists of items relating to teachers? experiences at work. Originally, 60 items were included in the questionnaire (see Appendix B, Section 6, Questions 1 through 60). Teachers were asked to indicate the degree to which different aspects of teaching work affect their well-being as a teacher. A brief definition of the construct of teacher well-being was included in the opening question: ?Currently, how do the following aspects of being a teacher affect your well-being as a teacher? Well-being refers to open, engaged, and healthy functioning as a teacher.? The items were scored on a seven-point Likert-type scale ranging from Negatively (1), Mostly negatively (2), More negatively than positively (3), Neither positively not negatively (4), More positively than negatively (5), Mostly positively (6), to Positively (7). These anchor points were chosen to allow teachers to report the positive or negative influence of work experiences on their well-being. Given that the scale was being developed, no evidence of validity or reliability previously existed for this scale. Evidence for reliability and validity are provided below in the results section.  2.2.3.2. Additional constructs. For the purpose of providing some support for the representativeness of the data and preliminary convergent evidence of validity, three additional constructs were assessed: teacher stress, job satisfaction, and flourishing. In addition, a social desirability measure was administered to ascertain whether teachers? responses to the well-being items were subject to biases. Table 2.6 shows the reliability indexes, means, standard deviations, and ranges for these variables. Composite variables were created by taking the mean of all the items for each construct.  35   Table 2.5 Constructs Under Examination in Each of the Three Studies and Item Location in the Appendices  Appendix B  Appendix D  Study 1 Study 2  Study 3 Teacher well-being Yes (see Section 6) Yes (see Table 2.7)  Yes (see Section 2) Stress Yes (see Section 5, Q. 5) ?  ? Job satisfaction Yes (see Section 5, Q. 1-4) Yes (see Section 5, Q. 1-4)  ? Flourishing Yes (see Section 2, Q. 7-14) Yes (see Section 2, Q. 7-14)  ? Social desirability Yes (see Section 5, Q. 6-12) ?  ? Perceived autonomy support ? Yes (see Section 2, Q. 1-6)  ? Need satisfaction ? Yes (see Section 3)  Yes (see Section 3) Autonomous and controlled motivation ? Yes (see Section 4)  ? Organisational commitment ? Yes (see Section 2, Q. 15-20)  ? Self-efficacy for teaching ? ?  Yes (see Section 4)   36   The first construct to provide convergent evidence of validity was a single-item teacher stress measure: ?In general, how stressful do you find being a teacher?? (Boyle et al., 1995; see Appendix B, Section 5, Question 5). Teachers responded to this on a scale ranging from Not Stressful (1) to Extremely Stressful (7). This approach to measuring stress has been used in previous research involving teachers (e.g., Chaplain, 2008). The second construct was job satisfaction and this was assessed using four items that Caprara, Barbaranelli, Borgogni, and Steca (2003) developed (see Appendix B, Section 5, Questions 1 through 4). The items (e.g., ?I am satisfied with my job?) were measured on a scale ranging from Never (1) to Almost Always (7). Previous studies have found evidence for reliability (with Cronbach?s alphas between .83 and .85) and validity (Caprara et al., 2003; Klassen et al., 2010) of scores from this scale. In the current study, a similar level of reliability was found (? = .87). The third construct was the Flourishing Scale (Diener et al., 2010; see Appendix B, Section 2, Questions 7 through 14). The scale consists of eight items covering purpose and meaning, supportive relationships, engagement/interest, contributing to others, competence, being a good person, optimism, and feeling respected (e.g., ?I am engaged and interested in my daily activities?). The items are measured on a scale ranging from Strongly Disagree (1) to Strongly Agree (7). In several samples of college students (Diener et al., 2010), high convergence with conceptually similar scales and a robust factor structure were found, as was evidence of reliability through internal consistency (? = .87) and temporally (? = .71). In the current study, a Cronbach?s alpha of .86 was found. The fourth construct was the Revised Marlowe-Crowne Social Desirability Scale (Form X1; Fischer & Fick, 1993; see Appendix B, Section 5, Questions 6 through 12). This 37   scale consists of seven items used to assess the degree to which participants? answers are influenced by social desirability (e.g., ?I sometimes try to get even rather than forgive and forget?). Participants indicated whether the items were true or false for them. In the current study, the Cronbach?s alpha for this scale was low (? = .52). Unfortunately, this low alpha is common for this scale (e.g., Beretvas, Meyers, & Leite, 2002; Brackett, et al., 2010) and may reflect several minor dimensions (cf. Loo & Thorpe, 2000). The presence of multiple, albeit minor, dimensions can result in a lower Cronbach?a alpha because there is less interrelatedness between the items (Cortina, 1993). Despite the low internal consistency, support for the use of this scale was provided by a confirmatory factor analysis that revealed good fit (details below). Nevertheless, the results involving this scale should be interpreted with caution. 2.2.3.3. Demographic information. Teachers were asked to supply demographic information (see Appendix B, Section 1, Questions 1 to 14) including age, sex, subjects taught, highest level of education, school name, ethnic/cultural heritage, country of birth, average SES of students in their school, average academic achievement of students in their school, years of teaching experience, roles and responsibilities, including school level and teaching position (e.g., general classroom teachers, special education teachers), time spent teaching per week, and school setting (urban, suburban, small town, or rural).      38   Table 2.6 Reliability Indexes, Means, and Standard Deviations of the Additional Measures     Range Variable ? M SD Potential Observed Stress  ? 3.53 0.834 1.00 ? 5.00  1.00 ? 5.00  Job Satisfaction .87 5.35 0.838 1.00 ? 7.00 1.00 ? 7.00 Flourishing .86 6.09 0.686 1.00 ? 7.00 1.75 ? 7.00 Social desirability .52 0.63 0.211 0.00 ? 1.00 0.14 ? 1.00   2.3. Results  There was very little missing data for this study, ranging from 0 to 8% for all the variables. The original pool of items contained 60 items. Before conducting factor analyses, two items were removed because they were highly similar to other items in the pool. Specifically, questions 39 and 60 were mistakenly duplicates of one another (see Section 6, Appendix B) and questions 4 and 5 were highly similar (r = .88; see Section 6, Appendix B), resulting in the removal of questions 4 and 60.  2.3.1. Factor Analyses Factor analyses were used to ascertain the dimensionality of the TWBS (Research Question 1). Exploratory factor analysis (EFA) was first conducted, followed by confirmation factor analysis (CFA). EFA allows researchers to examine patterns of interrelatedness among items in a measure without being constrained by a priori hypotheses about how the items should load on different factors or the number of factors that might be present in the data (Kline, 2011). In other words, it allows the researcher to discover whether 39   there are any coherent subsets of items and the nature of these subsets (Tabachnick & Fidell, 2007). EFA does this by summarising patterns of correlations and shared variance between items (Tabachnick & Fidell, 2007). When items correlate and share variance, they load on a factor (or latent variable) that gives an idea of the underlying dimension that ?causes? these items (i.e., the latent variable that produces scores on the items; Tabachnick & Fidell, 2007). The decision was made to utilise EFA instead of principal components analysis, another tool suitable for preliminary investigations of patterns of interrelatedness, due to the presumed underlying dimensions. Principal components analysis is appropriate when the relationships between items are merely empirical, whereas EFA is appropriate when some underlying dimension is driving the relationship between items (Tabachnick & Fidell, 2007). Given that I believe one or more underlying dimensions of teacher well-being are driving the relationships between items, factor analysis was chosen as the more appropriate method.  Once several EFAs were run and an understanding of the factors was obtained, CFA was conducted. Where EFA aims to identify the underlying processes that produced correlations between items, CFA aims to test whether the correlations among items are consistent with the hypothesised factor structure provided by the exploration conducted through the EFA (Tabachnick & Fidell, 2007). CFA does this by restricting items such that they only load on the factors specified by the researcher (Brown, 2006). For both EFA and CFA, the acceptability of the model is determined by assessing goodness of fit, interpretability of the items, and the strength of the parameter estimates (Brown, 2006).  Three indices were used to assess the goodness of fit of the models: Root Mean Square Error of Approximation (RMSEA), Comparative Fit Index (CFI), and Standardised Root Mean Square Residual (SRMR). RMSEA measures goodness-of-fit by assessing fit of 40   the model compared to a perfect model, where a lower score represents better fit (Tabachnick & Fidell, 2007). CFI measures relative improvements to the fit of the final model compared to an independence model (i.e., a model involving completely unrelated variables; Kline, 2011; Tabachnick & Fidell, 2007). A higher CFI corresponds with better fit. Finally, SRMR examines differences between the observed and predicted correlations in the data and model, respectively. For SRMR, a lower statistic represents better fit. These three indices of fit were chosen because they offer three different methods for calculating fit and combined they provide a more comprehensive argument for the fit of the models than one index may provide. In regard to the statistical power involved in these factor analyses, ?rules of thumb? suggested by Tabachnick and Fidell (2007) indicate that the sample size was adequate. In order to determine an acceptable degree of fit for these indices, cut-offs established in Hu and Bentler?s (1999) work were used. RMSEA values less than .10 were assessed as evidence of adequate fit and values less than  .06 were assessed as evidence of a good fit. CFI values greater than .90 and .95 were assessed as evidence of adequate and good fit, respectively. SRMR values less than .10 and .08 were assessed as evidence of adequate and good fit, respectively. The chi-square model fit test is also reported.  When using factor analysis, it is important to take into account the hierarchical nature of the data. Given that the data came from teachers who were nested within the schools at which they worked, the data were hierarchically nested. Accordingly, steps were taken to reduce artificial inflation of the parameter estimates, which can occur because teachers within the same school tend to be more similar to one another in their responses than teachers from different schools. The issue of hierarchically nested data was addressed using several relevant features of Mplus version 7.1 (Muth?n & Muth?n, 1998-2012). These included 41   using robust (sandwich estimator) standard errors, a robust chi-square test, and the ?TYPE=COMPLEX? option (with school as a cluster variable) as per Asparouhov and Muth?n?s (2005) recommendations. In order to perform the factor analyses, the robust weighted least squares (WLSMV) estimator and geomin oblique rotation were used. The WLSMV estimator is appropriate for categorical data and is robust to missing and non-normal data. Oblique rather than orthogonal rotation was chosen as it allows the factors to be correlated (correlation of the factors was expected given that they were all related to well-being); however, this was only used for the EFA as CFA does not require rotation (the factors are defined a priori and do not require rotation to find best fit).  2.3.1.1. Exploratory factor analyses. As described, data from the TWBS items were subjected to several EFAs. The EFAs were conducted on a randomly selected half of the sample to promote robustness of the results and so that the CFA could be conducted (and the factors confirmed) on a different set of responses. Because EFA is an exploratory procedure (Brown, 2006), steps must be taken to determine the most appropriate number of factors to be retained. In order to do this, parallel analysis was utilised. Parallel analysis is considered a highly accurate method of determining the number of factors to retain (Velicer, Eaton, & Fava, 2000). It estimates eigenvalues based on a random dataset that contains the same number of variables and cases as the real data. For factors to be retained from the real data, their eigenvalues must be greater than those created from the random data (Hayton, Allen, & Scarpello, 2004). This method was used to determine the best number of factors for each of the EFA models that were run. For the initial EFA, there were five factors. By the final EFA, there were three factors. 42   Through the running of several EFAs, 35 items were removed based on the strength of the factor loading (i.e., low loading or cross-loading with other factors) and/or interpretability (i.e., they did not make conceptual sense in relation to the factor on which they loaded; Brown, 2006). Through this process, items that loaded under .400 on all factors, items that cross-loaded as well as loaded under .500 on all factors, and items that cross-loaded within .100 were removed. This removal left 23 items in the measure organised under four factors. At this point, the items were examined to ascertain the underlying factors. Three of the items fell under the fourth factor: (a) ?Using technology for administrative work? (see Section 6, Appendix B, Question 14), (b) ?Using technology for teaching? (Question 15), and (c) ?The challenge of teaching work? (Question 16). Conceptually, grouping these items together under a fourth factor did not make sense. Furthermore, the factor only explained 8% of the variance and so for these two reasons it was removed. This step reduced the scale down to three factors. In the final step, an additional four items were removed because they did not fit well conceptually and/or load well.  The eigenvalues for the final three-factor model were 5.871, 2.254, and 1.693. Together they explained 61% of the total variance. Fit indices for the final EFA involving 16 items showed the following: ?2 (75, N = 239) = 160.511, p < .001, RMSEA = .069, CFI = .972, and SRMR = .037. These indices suggest that the three-factor solution has adequate to good fit. Table 2.7 shows the factors, the corresponding items, and the factor loadings. The items in each factor were examined to identify names that accurately describe the latent variable underlying the items. Factor 1 was given the name, workload well-being, as it relates to issues associated with workload and associated pressure. Factor 2 was given the name, organisational well-being, as it relates to teachers? perceptions of the school-as-an-43   organisation including perceptions of school leadership and the culture towards teachers and teaching at the school. Factor 3 was given the name, student interaction well-being, as it relates to teachers? interactions with students. 2.3.1.2. Confirmatory factor analyses. A confirmatory factor analysis (CFA) was conducted on the other half of the sample to confirm the factor structure revealed in the EFA. Fit indices revealed adequate to good fit of the confirmatory measurement model: ?2 (101, N = 237) = 222.763, p < .001, RMSEA = .071, and CFI = .97 (SRMR was not available for this estimation method). Standardised factor loadings from the CFA are shown in Table 2.8. Using the whole sample, the latent variable correlations between the three factors were calculated. These are shown in Table 2.9 and suggest that these factors are tapping into related, albeit different dimensions of teacher well-being. The moderate strength of these correlations also suggests that the three factors can be combined to obtain an overall score of teacher well-being.    44   Table 2.7 Factor Loadings for the Final Exploratory Factor Analysis Item Loading Factor 1: Workload well-being  1. Marking work .638 2. Fitting everything in to the allotted time .653 3. Administrative work related to teaching .790 4. Work I complete outside of school hours for teaching .748 5. Working to finish my teaching tasks .685 6. Staying late after work for meetings and activities .702 Factor 2: Organisational well-being  7. Relations with administrators at my school .775 8. Support offered by school leadership .809 9. Recognition for my teaching .526 10. School rules and procedures that are in place .700 11. Communication between members of the school .706 12. Participation in school-level decision making .724 Factor 3: Student interaction well-being  13. Relations with students in my class .620 14. Student behaviour .933 15. Student motivation .823 16. Classroom management .657    45   Table 2.8 Standardised Factor Loadings from the Well-being Confirmatory Factor Analysis  Item Loading Factor 1: Workload well-being  1. Marking work .638 2. Fitting everything in to the allotted time .748 3. Administrative work related to teaching .715 4. Work I complete outside of school hours for teaching .766 5. Working to finish my teaching tasks .773 6. Staying late after work for meetings and activities .742 Factor 2: Organisational well-being  7. Relations with administrators at my school .674 8. Support offered by school leadership .790 9. Recognition for my teaching .671 10. School rules and procedures that are in place .773 11. Communication between members of the school .668 12. Participation in school-level decision making .681 Factor 3: Student interaction well-being  13. Relations with students in my class .493 14. Student behaviour .927 15. Student motivation .832 16. Classroom management .792    46   Table 2.9 Latent Variable Correlations of the Well-being Factors  Workload well-being Organisational well-being Organisational well-being .478  Student interaction well-being .555 .483   2.3.1.3. Descriptive statistics. Descriptive statistics for the final 16 items and the three factors (created by taking the average of all items for each factor) are shown in Table 2.10. The observed range among items matched the potential range in all bar one item, suggesting a substantial amount of dispersion of scores. For workload well-being, the means of the items and the composite variable fell around or below the midpoint of the scale, whereas for organisational and student interaction well-being, the means of all items and the composite variables fell above the midpoint. The item that teachers reported affected their well-being most positively was item 13, ?Relations with students in my class.? In contrast, the item that affected teacher well-being most negatively was item 6, ?Staying late after work for meetings and activities.? The items associated with workload well-being generally showed positive skewness, revealing a concentration of scores below the mean. In contrast, the items associated with organisational and student interaction well-being all showed negative skewness, revealing a concentration of scores above the mean. Kurtosis was varied with some items showing leptokurtic distributions (positive kurtosis), others showing platykurtic distributions (negative kurtosis), and some showing almost no kurtosis. Cronbach?s alphas for the three factors were .85, .89, and, .82, respectively.  47    Table 2.10 Descriptive Statistics for Items in the Teacher Well-being Scale    Skewness  Kurtosis  Item M SD Statistic Std. error  Statistic Std. error Observed range Workload 3.43 1.058 0.31 0.102  0.02 0.203 1.00 ? 6.83 1. 4.06 1.372 -0.03 0.102  -0.11 0.203 1.00 ? 7.00 2. 3.11 1.375 0.40 0.102  -0.03 0.203 1.00 ? 7.00 3. 3.17 1.285 0.42 0.102  0.18 0.204 1.00 ? 7.00 4. 3.27 1.523 0.32 0.102  -0.65 0.204 1.00 ? 7.00 5. 3.91 1.367 -0.02 0.103  -0.25 0.205 1.00 ? 7.00 6. 3.06 1.460 0.45 0.102  -0.32 0.204 1.00 ? 7.00 Organisational 4.52 1.050 -0.45 0.102  0.18 0.203 1.00 ? 7.00 7. 5.20 1.433 -1.03 0.101  0.71 0.202 1.00 ? 7.00 8. 4.50 1.385 -0.55 0.103  -0.05 0.205 1.00 ? 7.00 9. 4.05 1.579 -0.21 0.102  -0.61 0.204 1.00 ? 7.00 10. 4.35 1.332 -0.48 0.102  0.03 0.204 1.00 ? 7.00 11. 4.52 1.384 -0.47 0.102  -0.33 0.204 1.00 ? 7.00 12. 4.48 1.312 -0.47 0.103  0.11 0.205 1.00 ? 7.00 Student interaction 4.72 1.092 -0.27 0.101  -0.42 0.202 1.00 ? 7.00 13. 5.90 0.923 -0.97 0.101  1.38 0.202 2.00 ? 7.00 14. 4.30 1.505 -0.40 0.101  -0.57 0.202 1.00 ? 7.00 15. 4.27 1.473 -0.28 0.101  -0.70 0.202 1.00 ? 7.00 16. 4.42 1.418 -0.21 0.102  -0.58 0.203 1.00 ? 7.00  48   Teachers? responses on the three factors were examined in relation to the variable of social desirability. First, the social desirability measure was subjected to a CFA to ensure the items loaded as expected. The CFA revealed good fit: ?2 (14, N = 485) = 20.86, p = .12, RMSEA = .032, and CFI = .95. Social desirability correlated with workload well-being at .103, with organisational well-being at .135, and with student interaction well-being at .098. These three correlations were significant (p < .05); however, their effect sizes were small (i.e., R2 < 2% of variance). This suggests that social desirability played a significant, but small role in teachers? ratings of teacher well-being. 2.3.2. Demographic Comparisons Comparisons of TWBS score differences across groups were conducted based on the following demographic variables: sex, teaching experience, school level, school setting, and teachers? perceptions of students? SES and academic achievement (Research Question 2).2 Table 2.11 shows the descriptive statistics for each group on the different demographic variables for each of the three factors.  2.3.2.1. Sex and teaching experience. A t-test revealed no significant differences between females and males on any of the three well-being factors (i.e., female and male teachers reported similar levels of well-being). In addition, Analysis of Variance (ANOVA) was run to ascertain if there were any significant differences in well-being between different career stages (see Table 2.11 for career stages examined). Once again, there were no significant differences between groups on any of the factors. 2 Comparisons involving different positions were not conducted given that the vast majority of participants were classroom teachers (81%).  49                                                    2.3.2.2. School level. ANOVA revealed significant differences between teachers at different school levels in their reports of workload well-being, F (3, 572) = 4.826, p = .005, ?2 = .02, and organisational well-being, F (3, 571) = 5.571, p = .002, ?2 = .03. Participants whose job involved teaching at more than one school level concurrently (e.g., middle and secondary as a learning support teacher) reported significantly higher workload well-being scores than teachers who taught at only one school level (e.g., elementary only). For organisational well-being, elementary teachers reported significantly higher scores than secondary teachers. In both cases, the effect sizes were small suggesting that minimal variance in scores was accounted for by school level (only 2-3% of the variance). 2.3.2.3. School setting. ANOVA revealed a significant difference in reports of workload well-being for school setting, F (3, 563) = 5.808, p = .001, ?2 = .03. Participants who taught in rural school settings reported significantly higher scores than teachers in all other settings. Again, the effect size was small suggesting little practical difference. There were no significant differences in organisational well-being and student interactional well-being for school setting. 2.3.2.4. Socio-economic status and academic achievement. Teachers? reports of workload, organisational, and student interaction well-being did not differ as a function of teachers? perceptions of their students? SES. For teachers? perceptions of students? academic achievement, there were no significant differences in reports of workload and organisational well-being; however, there were significant differences in reports of student interaction well-being, F (4, 565) = 3.226, p = .025, ?2 = .02. Teachers who reported low-average academic achievement among their students had significantly lower student interaction well-being than teachers who reported high-average academic achievement. Once again, the effect size was 50   small suggesting that minimal variance in student interaction well-being was accounted for by academic achievement of students.  Table 2.11 Descriptive Statistics for the Three Well-being Factors by Demographic Characteristics  Workload well-being M (SD) Organisational well-being M (SD) Student interaction well-being M (SD) Sex    Female 3.39 (1.080) 4.54 (1.047) 4.70 (1.102) Male 3.55 (0.980) 4.41 (1.015) 4.85 (1.001) Significant difference ns ns ns Years? experience    < 6 years 3.44 (0.944) 4.55 (0.884) 4.53 (1.038) 6-10 years 3.27 (0.963) 4.69 (0.843) 4.59 (1.056) 11-20 years 3.39 (1.068) 4.40 (1.074) 4.72 (1.055) 20-30 years 3.52 (1.104) 4.53 (1.086) 4.90 (1.138) > 30 years 3.77 (1.181) 4.51 (1.302) 4.89 (1.244) Significant difference ns ns ns School level    Elementary 3.38 (1.044) 4.65 (1.025) 4.67 (1.124) Middle 3.27 (1.015) 4.39 (0.981) 4.63 (1.087) Secondary 3.43 (1.092) 4.27 (1.124) 4.83 (1.063) All levels 3.96 (0.967) 4.63 (0.855) 4.74 (0.975) Significant difference p = .005 p = .002 ns 51   Table 2.11 (Continued)     Workload well-being M (SD) Organisational well-being M (SD) Student interaction well-being M (SD) School setting    Urban 3.41 (1.042) 4.53 (1.111) 4.70 (1.100) Suburban 3.36 (1.049) 4.52 (0.993) 4.71 (1.071) Small Town 3.42 (0.919) 4.26 (1.087) 4.66 (1.060) Rural 4.11 (1.108) 4.82 (0.895) 5.06 (1.120) Significant difference p = .001 ns ns SES of students? families    Low 3.42 (1.004) 4.53 (1.141) 4.70 (1.021) Low-average 3.35 (0.956) 4.49 (0.988) 4.59 (1.093) Average 3.53 (1.102) 4.56 (0.988) 4.72 (1.129) High-average 3.33 (1.125) 4.50 (1.152) 4.87 (1.054) High 3.38 (1.000) 4.30 (0.997) 4.93 (0.998) Significant difference ns ns ns Academic achievement of students   Low 3.43 (0.979) 4.49 (1.028) 4.70 (1.037) Low-average 3.45 (0.925) 4.70 (0.937) 4.56 (1.052) Average 3.48 (1.088) 4.48 (1.040) 4.68 (1.122) High-average 3.34 (1.119) 4.50 (1.086) 5.01 (1.012) High 3.31 (1.141) 4.21 (1.282) 4.88 (0.984) Significant difference ns ns p = .025 Note. ns = not significant.   52   2.3.3. Preliminary Evidence of Validity Evidence of validity based on internal structure was reported through the Cronbach?s alphas and factor analyses detailed previously. Namely, the Cronbach?s alphas provided adequate support for internal consistency, and the EFAs and CFA showed adequate to good fit for the factor structure of the three-factor solution. Taken together, these findings provide evidence for the validity of TWBS score interpretations based on internal structure.  Below, convergent evidence of validity is presented by comparing scores on the new measure with three related constructs. Before doing that, however, a CFA using the whole sample was conducted for the two multiple item comparative measures (i.e., the flourishing and job satisfaction scales) to ensure that they loaded as expected. The CFA revealed adequate to good fit, ?2 (53, N = 485) = 264.44, p < .001, RMSEA = .091, CFI = .98, with items loadings on two factors as expected: flourishing and job satisfaction (Table 2.12 shows the items and factor loadings). 2.3.3.1. Preliminary convergent evidence of validity. Table 2.13 shows the correlations of the three factors in the new measure with the constructs of stress, job satisfaction, and flourishing (Research Question 3). Given that the three factors of well-being correlated significantly with social desirability, partial correlations were also calculated to assess whether the inclusion of social desirability diminished the findings. Table 2.14 shows the partial correlations that controlled for social desirability. In all cases, the correlations between the well-being factors and the comparison measures remained significant and changed very little as a consequence of controlling for social desirability, indicating that social desirability did not significantly affect the findings. Taken together, both Table 2.13 and Table 2.14 show that reports of stress had moderately negative relationships with the 53   three factors of well-being, and that reports of both job satisfaction and flourishing had moderately positive relationships with the three factors of well-being.  Table 2.12 Standardised Factor Loadings from the Flourishing and Job Satisfaction Confirmatory Factor Analysis Item Loading Factor 1: Flourishing  1. I lead a purposeful and meaningful life .862 2. My social relationships are supportive and rewarding .761 3. I am engaged and interested in my daily activities .829 4. I actively contribute to the happiness and well-being of others .741 5. I am competent and capable in the activities that are important to me .772 6. I am a good person and live a good life .789 7. I am optimistic about my future .629 8. People respect me .715 Factor 2: Job satisfaction  1. I am satisfied  with my job .881 2. I am happy with the way my colleagues and superiors treat me .639 3. I am satisfied with what I achieve at work .896 4. I feel good at work .932    54   Table 2.13 Correlations of the Three Well-being Factors with the Additional Measures  Workload well-being Organisational well-being Student interaction well-being Stress -.393 -.264 -.312 Job satisfaction .339 .515 .515 Flourishing .271 .347 .375 Note. All correlations are significant at p < .01.   Table 2.14 Partial Correlations of the Three Well-being Factors with the Additional Measures Controlling for Social Desirability  Workload well-being Organisational well-being Student interaction well-being Stress -.402 -.273 -.321 Job satisfaction .335 .520 .518 Flourishing .267 .345 .373 Note. All correlations are significant at p < .01.    55   2.4. Discussion Teacher well-being is an issue that is receiving increasing attention given its link to important outcomes for teachers, students, and schools (e.g., Duckworth et al., 2009, Retelsdorf et al., 2010). Study 1 aimed to extend our understanding in this area by developing a dedicated measure for assessing teacher well-being. The majority of extant research in this area has assessed teacher well-being through related constructs (e.g., stress, Van Horn, Schaufeli, & Taris, 2001; burnout, McCarthy, Lambert, O?Donnell, & Melendres, 2009) or through measures of global well-being (e.g., life satisfaction; Albuquerque et al., 2011). More recently, researchers are beginning to consider how teacher well-being is shaped by experiences at work (e.g., Konu et al., 2010). In order to further extend our knowledge, more examinations like this are needed so that we can develop our understanding of what and how work-related factors influence teacher well-being. This type of research also has the potential to inform our understanding of other important ramifications like teacher attrition and retention. Study 1 addressed these issues by developing a measure of teacher well-being based on teachers? experiences at work.  In order to develop the new measure, a pool of items was administered to teachers. Following this, exploratory and confirmatory factor analyses were conducted to identify and confirm three factors of teacher well-being that tap into three distinct areas of teachers? work experiences (Research Question 1): workload well-being, organisational well-being, and student interaction well-being. Following this, demographic comparisons were conducted to ascertain how the new measure functions among different groups of teachers (Research Question 2). Finally, scores on the new measure, the TWBS, were correlated with other 56   measures to provide preliminary evidence of validity (Research Question 3). Key findings are discussed below. 2.4.1. Research Question 1: What is the Nature of Teacher Well-being? The factor analyses revealed three factors of teacher well-being: workload, organisational, and student interaction well-being. Descriptive statistics and factor analyses revealed a psychometrically acceptable instrument for measuring these factors. The scores on the items retained for the three factors fell only slightly above or below the mid-point and the observed ranges matched the potential ranges in all but one item. These properties suggest that the items adequately differentiate between teachers. Furthermore, the Cronbach?s alphas were high for the three factors indicating adequate internal consistency.  2.4.1.1. Factor one: Workload well-being. The first factor of the new measure is workload well-being. The six items associated with this factor relate to tasks that teachers are required to do as part of their teaching duties such as marking assignments, attending meetings, and working after hours. All of the items associated with this factor had means that were below the mid-point of the scale. In other words, the aspects of teaching that relate to workload influenced teacher well-being negatively on average. This finding confirms previous studies showing that these types of teaching tasks are associated with stress (Collie et al., 2012; Klassen & Chiu, 2010) and burnout (Schaufeli & Bakker, 2004). The finding is important for administrators and policy-makers as it underscores the importance of providing balanced workloads for teachers. Ample research has shown that teachers? workloads are associated with stress (e.g., Collie et al., 2012; Klassen & Chiu, 2010). Furthermore, Milkie and Warner (2011) found that teachers who indicated that administrative tasks interfered with their teaching were more likely to report greater 57   externalising behavioural problems among their students (e.g., fighting, disturbing ongoing activities). Administrators and policy-makers, therefore, may want to pay more attention to the increasing workloads that are being placed upon teachers. Although the aim of increasing workloads is to encourage better learning and achievement among students, this may not occur if teacher well-being is threatened. In fact, the opposite may occur given that teacher well-being is associated with important outcomes such as effective teaching (Duckworth et al., 2009). In other words, teaching effectiveness and students? learning and achievement may suffer from the very workloads that are attempting to increase teaching effectiveness and student achievement.  Future research that examines why some teachers reported that these items positively influence their well-being has the potential to shed more light on what school leadership and policy-makers can do to minimise the negative effects of workloads on teachers. Perhaps it is the case that those teachers who reported positive influences worked in environments where workloads were appropriately balanced or more support was offered for meeting workload demands. 2.4.1.2. Factor two: Organisational well-being. The second factor of the new measure is organisational well-being. The six items associated with this factor refer to organisational-level issues relating to teachers and teaching such as relations and communications between teachers and administrators, support and recognition offered by administrators, participation in decision-making by teachers, and the school rules and procedures in place. All of the items associated with this factor had means that were above the mid-point of the scale. In other words, teachers generally felt that the organisational-level aspects of their work influenced their well-being positively. This is a positive finding for 58   school administrators as it suggests they are succeeding in creating working environments that are healthy for teachers. In particular, ?relations with administrators in my school? was rated most highly among these items. This finding is supported by research on school climate that has linked positive perceptions of these types of items with well-being related constructs (e.g., lower burnout, Grayson & Alvarez, 2008; lower stress, Collie et al., 2012).  This finding is important as it highlights the significance of organisational-level issues for teacher well-being. It extends our understanding of well-being by showing how and what organisational-level issues play a role in teachers? reports of their work-related well-being. It also underscores the critical role that administrators play in teacher well-being?all of the items tap into organisational-level issues that are overseen or governed by administrators. Furthermore, this finding identifies several different ways administrators can improve teacher well-being outcomes. Namely, they can help to foster teacher well-being by promoting positive relationships and dialogue across the school, ensuring teachers have input in decision-making, ensuring appropriate and constructive rules and procedures are in place, and recognising teachers? efforts.  The ramifications of this factor of well-being, however, extend further. These organisational-level issues are all related to school climate. Given there are many positive outcomes associated with positive school climates for teachers, students, and schools (e.g., student learning, Bryk & Schneider, 2003; student achievement, MacNeil, Prater, & Busch, 2009), targeting this factor for improvement has the potential for far-reaching and positive implications for all members of the school community, not just teachers.   59   2.4.1.3. Factor three: Student interaction well-being. The third factor of the new measure is student interaction well-being. The four items associated with this factor refer to teachers? relations with their students, student behaviour, student motivation, and classroom management. Like factor two, all of the items associated with this factor had means that were above the mid-point of the scale. Teachers generally felt that their interactions with students influenced their well-being positively. This finding corroborates Klassen, Perry, and Frenzel?s (2012) research showing that relatedness with students was positively associated with enjoyment for teaching and negatively associated with anxiety, anger, and emotional exhaustion among teachers. Furthermore, given that working with students is a primary reason that many teachers cite for entering the profession (e.g., Watt & Richardson, 2007; Sinclair, 2008) and finding it satisfying (e.g., Shann, 1998), it is not surprising that student interactions influenced teacher well-being positively. The fact that these items formed their own factor underscores the importance of teacher-student relations. Despite findings from the current study highlighting the importance of student interactions for teacher well-being, teacher-student interactions can also be stressful (e.g., Collie et al., 2012; Klassen & Chiu, 2010). In other words, it appears that student interactions can be stressful for teachers, while also influencing teacher well-being positively. Taken together, these links between teachers? reports of well-being/stress and teacher-student relationships further underscore the importance that teachers place on their relationships with students. This finding also highlights the complex nature of teaching and has important implications for researchers. In particular, it highlights the importance of considering a broad picture when attempting to understand how teachers are influenced by their experiences at work. Focusing on one variable may only provide a partial understanding of teachers? 60   experiences at work. This finding also extends our understanding by suggesting that stress and well-being can both be present in relation to the same aspect of teaching. This may be because of the multifaceted nature of stress (Selye, 1974)?at moderate levels, stress can be positive for teachers, whereas at high levels it is often detrimental (Collie, Perry, & Brenner, 2013). Alternatively, it may relate to teachers? relationships with different students in the class (e.g., relationship quality may vary with different students) at different times (e.g., relationship quality may change over the course of the school year). Future research should attempt to untangle when and how teachers experience stress and well-being at the same time and what this means for teacher effectiveness.  2.4.1.4. Summary. Taken together, the three factors highlight that teacher well-being as measured by the TWBS is a multi-dimensional construct. This has important implications for research and practice. In particular, given that workload well-being was rated quite differently from organisational and student interaction well-being, utilising a one-dimensional view of teacher well-being in research may not provide a full picture of how teachers experience their work. Instead, researchers should consider a multi-dimensional view so as to obtain a more complete understanding of this construct. For practice, a multi-dimensional view of teacher well-being promises to give administrators and policy-makers a deeper understanding of how teachers are faring in three critical areas, along with an indication of what specific issues should be targeted to improve teachers? experiences at work and their well-being.  Although not directly examined, the results also provide an understanding of how existing theory is relevant to understanding teacher well-being. As described earlier, items derived from components of SDT (Deci & Ryan, 1985, 2002) and socio-ecological theory 61   (Bronfenbrenner, 1978) were included in the pool of items. The final list of items provides insights into how these theories are relevant for understanding teacher well-being. Many of the items appeared to relate to the basic psychological needs of SDT (autonomy, competence, and relatedness). For workload well-being, if teachers are struggling with their marking workload, they may lack a sense of competence (if they find it hard to mark) or autonomy (if they were not responsible for designing the assignment). For student interaction well-being, the items appear to connect with the need for relatedness with students (e.g., ?Student motivation?). This supports Klassen, Perry, and colleagues? (2012) finding that relatedness with students was an important predictor of teachers? outcomes.  Socio-ecological theory (Bronfenbrenner, 1978) also appears to provide an understanding of the new measure. For organisational and student interaction well-being, all the items appear to relate to macrosystems?that is, the beliefs and values of a particular group (Bronfenbrenner, 1994). For example, the item, ?recognition offered for teaching work,? relates to belief systems that dictate how teachers are viewed and treated at the school-level, as well as at the district and broader societal levels. In contrast, for workload well-being, the items mostly relate to microsystems (e.g., the classroom microsystem for completing marking work). In addition, there were two items that related to mesosystems. ?Work I complete outside of school hours for teaching? and ?staying late after work for meetings and activities? appear to relate to the linkages between school and teachers? homes. It is interesting to note that staying late after work was rated the most negatively of all the items in the instrument, suggesting that the linkage between home and school may not be positive for teacher well-being. This is understandable given that staying late after school takes away teachers? time with their families. 62   In combination, these links to theory suggest a need for research that directly examines relationships that the three factors of well-being have with components from SDT (Deci & Ryan, 1985, 2002) and socio-ecological theory (Bronfenbrenner, 1978). Such research promises to help us understand how and why certain aspects of teaching work are implicated in teacher well-being. Studies 2 and 3 take first steps in addressing this in relation to SDT by examining empirical relations between the three factors of well-being and the three basic psychological needs from two different perspectives. Study 2 examines this cross-sectionally as part of an explanatory model based in SDT that also investigates relationships between and among well-being, motivation, job satisfaction, and organisational commitment. Study 3 examines relationships between the basic psychological needs and teacher well-being over time.  2.4.2. Research Question 2: How do Scores on the New Measure Differ Across Demographic Groups? For the second research question, an examination of differences in reports of teacher well-being was conducted across demographic groups that were formed based on sex, years of teaching experience, school level, school setting, and teachers? perceptions of students? SES and academic achievement. These comparisons were undertaken to examine whether or not these demographic characteristics play a substantial role in teacher well-being. There were no significant effects obtained for teachers? reports of sex, years of teaching experience, and SES of students, indicating these three demographic variables did not play a statistically significant role in differences observed in scores on the TWBS for participants in this study.  Significant effects were obtained for reports of workload well-being and organisational well-being in relation to school level, for workload well-being in relation to 63   school setting, and for student interaction well-being in relation to teachers? perceptions of students? academic achievement. In all cases, the effect sizes were small and explained less than 3% of the variance. Although statistically significant, these small effect sizes suggest that the examined demographic variables do not play a substantial role in differences in teacher well-being. For the personal demographic characteristics (e.g., sex, ethnicity), it is possible that this finding reflects the relative homogeneity of teachers (e.g., Caucasian, female). However, the fact that these job-related demographic characteristics (e.g., teaching experience, school level) also did not play a major role provides support for the instrument?s relevance among teachers in different school settings and teaching situations. This finding has several implications for research and practice. First, it extends our understanding of teacher well-being by suggesting that we should focus our attention towards other demographic variables. For instance, given that special education teachers experience particularly high levels of stress, burnout, and attrition (e.g., Lavian, 2012; Thornton, Peltier, & Medina, 2007), comparing differences across teaching positions may reveal more differences than the examined demographic variables. The findings also highlight the importance of examining differences in well-being based on other factors such as school psychosocial factors (e.g., school climate) and individual beliefs.  Second, this finding reduces concerns about variance in scores being due to demographic characteristics given that teachers from the different groups reported similar scores. It is important to note, however, that in more diverse samples this may not be the case (e.g., with more male teachers, greater ethnic diversity). There is a need, therefore, for research among more diverse teaching samples on whether and how these and other demographic variables play a role in teacher well-being. Notwithstanding this, the finding 64   suggests that efforts to improve teacher well-being are likely to affect teachers among demographically comparable samples in similar ways.   Finally, and more broadly, the fact that the measure functioned similarly across these demographic groups poses larger questions about the types of individuals that enter teaching in the first place and why they are so similar demographically. Research that endeavours to examine why individuals from broader demographic groups are not choosing teaching will have implications for understanding how to make teacher education programs more accessible to and representative of the larger population.  2.4.3. Research Question 3: Are Interpretations of Scores from the New Measure Supported by Evidence of Validity? The final research question involved conducting analyses to provide preliminary evidence of validity. Factor analyses and reliability analyses provided preliminary evidence of validity based on internal structure. In addition, correlations provided preliminary evidence of validity based on relations to three other variables: stress, job satisfaction, and flourishing (i.e., the measure of well-being relating to an individual?s life as a whole).  Three of the four hypotheses about convergent evidence of validity were supported. Namely, the moderately negative correlations between stress and the well-being factors (Hypothesis 1), along with the moderately positive correlations between job satisfaction and the well-being factors (Hypothesis 2a) and between flourishing and the well-being factors (Hypothesis 3), were supported and provide evidence of validity. Taken together these correlations conceptually situate the well-being factors where they were expected to exist and provide support for the interpretations of scores on the TWBS. Future research should 65   examine scores on the TWBS in relation to additional variables so as to provide further evidence of validity and greater understanding about the three factors of teacher well-being. The fourth and final hypothesis (Hypothesis 2b) concerned a comparison with Bowling and colleagues? (2010) research. It was hypothesised that the relationship between reports of job satisfaction and teacher well-being in the current study would be stronger than that found between reports of job satisfaction and subjective well-being by Bowling and colleagues (i.e., r = .36). Support for this hypothesis was only found for two factors: organisational (r = .52) and student interaction well-being (r = .52). The correlation for workload well-being (r = .34) was similar to Bowling and colleagues? correlation. Although the hypothesis was not supported for workload well-being, upon reflection the finding makes sense given that teachers rated the workload well-being factor as negatively affecting their well-being. In other words, negatively-rated workload well-being should relate less strongly with job satisfaction than positively-rated subjective well-being. Although unexpected, this finding does, in fact, provide evidence of validity for the meaning of the latent variable. Taken together, the relationships with other variables extend our understanding of teacher well-being by situating it in relation to other relevant constructs. In addition, the findings indicate that teacher well-being is distinct from the other constructs. The findings warrant concluding, therefore, that teacher well-being should be examined separately. Finally, the findings provide important evidence of validity to support interpretations of the test scores of the TWBS. In order to further support the robustness of the measure, future research is needed to test the measure with different samples to see whether the three factors remain relevant and/or whether other factors are important.  66   2.5. Limitations and Future Directions There are several limitations to this study that must be highlighted. First, due to the nature of the online data collection it was not possible to ascertain exact response rates. More specifically, because details of the study were sent to presidents of the teachers? association who then forwarded them onto teachers, I could not determine how many teachers received and/or read the study email and chose not to participate. It is possible that some teachers did not check their email, ignored the email, or had filters that deleted the email about the study (e.g., the email was marked as spam). Given this, I was only able to ascertain the number of teachers who read the details and chose to participate.  Despite this limitation, online data collection does have several strengths including its broad reach (data can be collected from anywhere there is an Internet connection) and lower cost (it does not require expensive mail outs). Furthermore, given the increasing time individuals are spending online (on laptops, tablets, and mobile phones), online questionnaires are a method of data collection that are readily accessible by participants. Taking all this into account, online data collection is a method that will likely increase in popularity. Steps are needed, therefore, to support the robustness of this data collection method.  In the current study, several steps were taken to temper the limitation of the unknown response rates and to support the representativeness of the data. In particular, the demographic characteristics of the sample were compared with the population and found to be very similar. In addition, average scores on the constructs were compared with other studies using related measures among similar populations and were found to be comparable. In the future, efforts that attempt to collect response rates will be important in providing 67   support for the robustness of the current study?s findings and also to provide further support for the TWBS as an appropriate measure of teacher well-being. One method for doing this would involve sending personalised email invitations to participants. By emailing individual teachers directly, researchers would gain a clearer idea of how many teachers read the email (e.g., by a follow-up email or phone call). Another method would involve face-to-face recruiting in schools. Collecting consent forms at the time of introducing the study to teachers would provide researchers with an idea of response rates. Given that online survey research is becoming more prevalent, these types of steps are needed more globally to support the robustness of this data collection method.  A second limitation of the study concerns the sample size. Rules of thumb in the literature (e.g., Tabachnick & Fidell, 2007) suggest that the number of participants was adequate for the analyses undertaken. However, rules of thumb have been criticised for being overly simplistic (Klein, 2011). Power analyses conducted prior to data collection in future research will help to temper this limitation and provide a more robust understanding of the appropriateness of the sample size. The third limitation of Study 1 is that the teachers were involved in industrial action at the time of data collection. It is possible, therefore, that teachers? responses were affected by the industrial action. In order to mitigate this limitation, comparisons with teachers not involved in such action from previous studies were provided and they suggested that teachers? responses were not significantly different. Of particular note, average stress levels in the current study were not significantly different from those measured among the same population when there was no industrial action taking place (Collie, 2010). Nevertheless, 68   future research among different samples of teachers is important to ascertain whether the measure behaves similarly among teachers who are not involved in industrial action.  At the same time, it is also important to note that at any given time, individual teachers or groups of teachers may be experiencing other specific, time-limited stressors that may influence their well-being (family issues, students? personal issues, etc.). Collecting data about contextual variables (e.g., leadership stability in the school) and/or personal experiences (e.g., health issues of teachers or their families) will help to provide researchers with an understanding of what factors teachers consider when selecting their responses on the TWBS and what other issues affect their responses. In addition, research involving other samples will help to provide support for the relevance of these three factors of teacher well-being among different teachers.  The limitations of self-report data have been argued elsewhere (e.g., Perry & Winne, 2006; Winne, Jamieson-Noel, & Muis, 2002). A well-documented limitation of self-report data is that participants? responses may not be accurate representations of reality. Given that I am interested in how teachers experience their workplace, self-reports are a relevant method of obtaining their perceptions. However, it is important when reporting this type of data that representative language is used (e.g., teachers perceived, reported, experienced) to clarify that the data refer to self-reports rather than other means of measurement (e.g., biological measures).  Another limitation of self-report data is that it is assumed participants interpret questions in the way the researcher intended, but this may not be the case. There are several methods that can help to address this limitation. The use of factor analyses provided evidence that teachers? interpretations were as expected. That is, the factors made theoretical sense and 69   the items interrelated in expected ways, thus providing support for the participants? interpretations of the items? meanings. Furthermore, the correlations with other variables provided further evidence of the construct meaning. Notwithstanding this, investigations of participants? thought processes as they respond to each question are a highly important area of future research and will provide understanding about teachers? interpretations and decision-making processes for score selection. Specifically, it would be helpful to ask participants to indicate their thoughts as they come to a decision about each item and to ask about their reasoning when choosing between the anchor points on the scale. This would also provide evidence of validity of TWBS score interpretations based on response processes (APA et al., 1999).  The final limitation to be discussed is that the data were collected from one administration of a self-report questionnaire. This means that there was only one source of data and that the findings are threatened by single source bias. There are several factors that influence single source bias, one of which is social desirability. Specifically, items that are more (or less) socially desirable may relate more (or less) strongly with each other in spite of differences in the underlying construct (i.e., social desirability causes some of the relatedness; Podsakoff et al., 2003). Controlling for social desirability, therefore, is one method of reducing single source bias and this method was employed in the current study. Given that the social desirability measure had low internal consistency, future research utilising more robust measures is important. Moreover, investigations involving multiple measures are an important next step as these will reduce the biases that are associated with single-source methods (e.g., higher measurement error) and allow for greater robustness though triangulation. Examples of this include comparing results on the TWBS with other 70   indicators of teacher well-being (e.g., physical well-being, behaviour at work) through multiple measure investigations. The limitations notwithstanding, the findings reported here suggest that the TWBS is an appropriate measure for ascertaining teacher well-being that is based on teachers? own experiences.  2.6. Conclusions In summary, Study 1 has involved developing and testing a measure of teacher work-related well-being that is based on teachers? own experiences. As such, the TWBS is a practical measure that holds promise for helping teachers, school administrators, and policy makers to understand and, potentially, guide efforts to improve teacher well-being. In addition, it is hoped that the findings of Study 1 will help to spur greater attention to the construct of teacher well-being in the literature. Indeed, the need for more attention to be paid to work-related well-being has been highlighted by several experts on the matter (Diener 2009; Rath & Harter, 2010).  EFA and CFA revealed three factors of teacher well-being?well-being related to workload, organisational-level issues, and student interactions. These findings indicate that teacher well-being, as measured here, has a multidimensional nature. Other findings provided evidence of validity for interpretations of scores on the new measure, along with evidence that variance in scores was minimally impacted by the examined demographic variables.  For the purpose of further developing our understanding of teacher well-being, two key types of examinations are needed. First, there is a need for examinations that attempt to situate the well-being construct in relation to other relevant constructs, such as motivation and job satisfaction. This will help build a greater understanding of teacher well-being, how it interacts with other constructs, and whether it actually measures what we want it to 71   measure. Examinations that attempt to understand how the construct functions over time are also needed. This will promote our understanding of teachers? experiences of well-being, how well-being develops over time and, ultimately, guide research that aims to promote teacher well-being. Studies 2 and 3 of this dissertation begin the process of addressing these needs. Specifically, Study 2 examined teacher well-being as measured by the TWBS alongside other important constructs (motivation, job satisfaction, etc.) within the framework of SDT, and Study 3 examined reports of teacher well-being and self-efficacy for teaching at three time points across several months.  72   CHAPTER 3. STUDY 2: DEVELOPING AND TESTING A FRAMEWORK 3.1. Literature Review What helps teachers do their best teaching? How can teachers be supported to teach effectively? How can teachers inspire learning and achievement among their students? The answers to these questions have long been the focus of educational theorists and researchers. Although definitive answers are complex, we do know that there are certain factors that can help to promote effective teaching and learning, two of which are a teacher?s well-being and motivation. As described in Study 1, research has shown that these two factors have been positively associated with teaching effectiveness (e.g., Duckworth et al., 2009), resiliency in teaching (e.g., Klusmann et al., 2008), quality instructional practices (e.g., Retelsdorf et al., 2010), and students? achievement (e.g., Caprara et al., 2006). Also relevant for effective teaching and learning are teachers? job satisfaction and organisational commitment (e.g., Bogler & Somech, 2004; Weiqi, 2007). In fact, research has established that teachers? experiences of these two work-related variables are not only interrelated (e.g., Feather & Rauter, 2004), but they are also associated with well-being and motivation among teachers (e.g., Collie et al., 2012; Hakanen, Bakker, & Schaufeli, 2006; Jackson et al., 2006; Klassen & Chiu, 2011; Mo?, Pazzaglia, Ronconi, 2010; Weiqi, 2007). These teacher-related variables, therefore, have far-reaching ramifications for teachers, students, and schools. Despite knowing that teachers? experiences of these four variables (well-being, motivation, job satisfaction, and organisational commitment) are interrelated, we still do not know how they interact when considered simultaneously. To date, no research in this area has examined these four variables in combination. In order to advance our understanding of teachers? experiences at work, therefore, it is necessary to examine how these four variables 73   interact with one another in concert. This is important given that these variables are associated with key outcomes for teachers and students (e.g., teaching effectiveness, Duckworth et al., 2009; students? motivation, Pakarinen et al., 2010). It is also important given that the variables may function differently when considered in combination. Furthermore, it will help to provide critical information in understanding how efforts to improve one of the variables may affect teachers? experiences of the other three variables. For example, if schools implement an initiative to increase teachers? motivation, it would be very helpful to understand how this may simultaneously affect teachers? well-being, job satisfaction, and organisational commitment.  A relevant framework through which to examine the interrelationships between these variables is self-determination theory (SDT; Deci & Ryan, 1985, 2002). Briefly, SDT establishes that an individual?s well-being and motivation in a certain context, such as work, are influenced by the autonomy support provided/experienced within that context, as well as the satisfaction of three basic psychological needs?autonomy, competence, and relatedness?in that context. For teachers, this means that their motivation and well-being are affected by the autonomy support provided in their working environment, along with their need satisfaction for autonomy, competence, and relatedness.  SDT research has been applied in a variety of settings, including classroom settings (e.g., Filak & Sheldon, 2008; Roth et al., 2007); however, most of this research has involved students, with only a limited amount of research examining teachers? self-determination and factors associated with it (e.g., Klassen, Perry, et al., 2012; Taylor, Ntoumanis & Standage, 2008). Roth, Assor, Kanat-Maymon, and Kaplan (2007) have called this surprising and I agree given that SDT provides a relevant and empirically supported framework for 74   understanding well-being and motivation among teachers. Moreover, SDT provides a strong foundation for considering how well-being and motivation are related to teachers? job satisfaction and organisational commitment. The aim of Study 2 was to present and test an explanatory model of teacher well-being, motivation, job satisfaction, and affective organisational commitment that is based in SDT. The explanatory model is shown in Figure 3.1. The first part of the model utilises SDT to show that autonomy-supportive work climates (shown as perceived autonomy support) predict need satisfaction and, in turn, need satisfaction predicts well-being and motivation. The second part of the model shows how well-being and motivation are associated with job satisfaction and organisational commitment (shown as affective organisation commitment). Although some conceptual support exists for these relationships (e.g., Caprara et al., 2006), more research is needed to extend our understanding of how these constructs interact simultaneously. Study 2, therefore, involved testing a model of these relationships. Before describing the explanatory model, SDT and the core variables are briefly outlined.     75     Figure 3.1. The explanatory model of teacher well-being, motivation, job satisfaction, and affective organisational commitment.  3.1.1. Self-Determination Theory SDT posits that an autonomy-supportive work climate has a positive influence on employees? well-being and motivation (Deci & Ryan, 2002). One way to examine autonomy support is to look at the support provided by people who supervise in each context (e.g., managers, principals). In business settings, autonomy-supportive work climates involve managers providing employees with a meaningful rationale for the work tasks that they are required to do, acknowledging employee?s feelings regarding uninteresting tasks, and providing employees with choices regarding what work tasks they do, when they do them, and how they complete them (Deci, Eghrari, Patrick, & Leone, 1994). In educational settings, many studies have shown that teachers who provide an autonomy-supportive 76   learning environment for their students (i.e., through taking the students? perspective, acknowledging the students? feelings, providing choice, and minimising demands; Black & Deci, 2000) tend to have students who experience positive outcomes (e.g., Deci, Schwarz, Sheinman, & Ryan, 1981; Filak & Sheldon, 2008; Jang, Reeve, & Deci, 2010; Reeve, Jang, Carrell, Jeon, & Barch, 2004; Roth et al., 2007; Taylor et al., 2008).  Emerging research is showing that a similar relationship may occur between teachers and principals. Namely, when teachers perceive that their principal is supporting their autonomy, they are more likely to report positive outcomes (e.g., Klassen, Perry, et al., 2012).  One relevant premise of SDT is that autonomy-supportive work climates allow individuals to experience satisfaction of three basic psychological needs: autonomy, competence, and relatedness (see Path 1 in Figure 3.1; Deci & Ryan, 1985, 2002). As defined in Study 1, autonomy involves the perception that we are the origin or source of our behaviour and that we have a choice in the types of actions we undertake (Deci & Ryan, 1985, 2002; Skinner & Edge, 2002). Competence involves the perception that we are effective in our interactions (Deci & Ryan, 2002). Relatedness refers to our need to feel connected to important others and the community (Deci & Ryan, 2002; Taylor & Ntoumanis, 2007).  One notable piece of research examining this premise of SDT among teachers is Klassen, Perry, et al.?s (2012) work. In two related studies conducted among Canadian teachers, they utilised structural equation modelling to examine the impact of teachers? perceptions of principal?s autonomy support on need satisfaction, and the subsequent relationships that need satisfaction has with work engagement, emotional exhaustion, anger, anxiety, and enjoyment. The findings revealed that teachers? perceptions of principal?s 77   autonomy support positively predicted their reports of basic psychological need satisfaction, which in turn predicted positive outcomes. Of particular importance, was their finding of a two-component model of the basic psychological need of relatedness: relatedness with colleagues and relatedness with students. Like research in business settings, they found that relatedness with colleagues was an important predictor of teachers? outcomes; however, they found that relatedness with students was more important. Thus, the findings of the Klassen, Perry, et al. study extend our understanding of relatedness among teachers by showing that it is different from the more typical conceptualisations of this need in business and other settings.  A second premise of SDT is that when an individual experiences need satisfaction of autonomy, competence, and relatedness, his or her well-being (see Path 2 in Figure 3.1) and motivation (see Path 3 and 4 in Figure 3.1) are positively affected (Deci & Ryan, 1985, 2002). In the current study, well-being is defined as flourishing in relation to teaching work. As defined in the Introduction, it includes eight factors: purpose and meaning, supportive relationships, engagement, contributing to others, competence, being a good person, optimism, and feeling respected (Diener et al., 2010). Research has supported the premise that need satisfaction positively influences well-being (defined in different ways; e.g., emotional exhaustion, vitality) in a variety of settings. Klassen, Perry, et al.?s (2012) study introduced above found that teachers? need satisfaction of relatedness with students was negatively associated with emotional exhaustion?a construct negatively related to well-being. Among college students, Reis, Sheldon, Gable, Roscoe, and Ryan (2000) found that the degree of need satisfaction reported in daily activities predicted fluctuations in individual well-being (i.e., positive and negative affect, vitality). In another study, Gagn?, Ryan, and 78   Bargmann (2003) found that athletes? perceptions of basic psychological need support from coaches during a training session predicted their reports of well-being (i.e., positive and negative affect, vitality) after the training.  Motivation (see Path 3 and 4 in Figure 3.1), as explained in the Introduction, is defined as the drive to act, think, and develop (Deci & Ryan, 2008a) in relation to teaching work. Although motivation is both an individual characteristic (i.e., a result of our innate personality) and a contextual characteristic (i.e., a result of our surroundings and experiences), this study focuses on how the experiences within a context can influence motivation. SDT highlights two types of motivation: autonomous and controlled motivations, each of which are influenced by intrinsic and/or extrinsic regulatory mechanisms (see Deci & Ryan, 2012 for a review).  Autonomous motivation is highly self-determined and involves identifying with behaviours and activities, integrating them into one?s sense of self, and undertaking them with full volition and choice (Deci & Ryan, 2008a; Deci & Ryan, 2008b). There are three types of behaviour regulation that are associated with autonomous motivation. The first is intrinsic regulation, which is internally or ?self? motivated and refers to engaging in a behaviour because it is inherently interesting and satisfying (e.g., a teacher reads new topic material because it is enjoyable; Deci & Ryan, 2008a; Ryan & Connell, 1989). The second is integrated regulation, which is extrinsically motivated but involves a behaviour that is brought into line with the individual?s overarching beliefs, values, and goals (e.g., a teacher reads new topic material because it means she will be a better teacher; Deci & Ryan, 2002). The third type of regulation, identified regulation, is also extrinsically motivated and involves consciously valuing the behaviour, but it may not represent an individual?s 79   overarching beliefs (e.g., a teacher reads new topic material because he wants to understand the topic or recognises that the topic is important for teaching; Deci & Ryan, 2002; Ryan & Connell, 1989).  In contrast, controlled motivation is not self-determined and involves undertaking behaviours and activities due to pressure or demands that are perceived to be externally controlled (Deci & Ryan, 2008a). Controlled motivation encompasses two types of behaviour regulation that are extrinsically motivated. The first is external regulation, where an individual engages in a behaviour to satisfy an external demand, to gain a reward, or to avoid a punishment (e.g., a teacher reads new topic materials because that is what she is supposed to do; Deci & Ryan, 2002; Ryan & Connell, 1989). This involves no self-determination and no internalisation?it is fully externally determined. The second type of regulation, introjected regulation, reflects behaviour that is partially internalised (i.e., incorporated with the individual?s sense of self), but undertaken to avoid guilt/shame (e.g., a teacher reads new topic material because he will feel guilty if he does not know it well enough) or obtain contingent self-worth (e.g., a teacher reads new topic material so that others think she is a good teacher; Deci & Ryan, 2002; Ryan & Connell, 1989). Although this is partially internalised, it is not self-determined, but self-controlled (Pelletier, Fortier, Vallerand, & Bri?re, 2001). Another notable study involving teachers is Roth and colleagues (2007) research, which examined whether the types of motivation described by SDT are relevant for understanding teachers? experiences. They investigated whether teachers discriminate between the types of regulation (e.g., identified, introjected) as expected and whether their experiences of these types of regulation fall along the continuum of self-determination as 80   described by SDT. The study, conducted among Israeli teachers, revealed that four types of behaviour regulation?external, introjected, identified, and intrinsic (integrated regulation was excluded because it is difficult to differentiate it from intrinsic regulation psychometrically; Roth et al., 2007)?were distinguished by teachers and fell along the expected continuum of self-determination. They also examined whether reports of autonomous motivation for teaching were associated with positive outcomes for teachers and students. Their results revealed that autonomous motivation for teaching was negatively associated with teachers? emotional exhaustion and positively associated with teachers? sense of personal accomplishment, autonomy-supportive teaching, and students? own autonomous motivation for learning. Combined, the findings of this study provide important foundational understanding of how the types of regulation associated with SDT are relevant for explaining teacher motivation. In order to further advance our understanding, there is a need for research that examines whether need satisfaction predicts these different types of regulation.  According to SDT, when individuals? basic psychological needs are fulfilled, they are more likely to internalise or identify more closely with certain behaviours (i.e., act with self-determination), which results in more autonomous forms of motivation (Deci, Vallerand, Pelletier, & Ryan, 1991). Research has shown that autonomous motivation is associated with positive outcomes at work (e.g., Fernet, Guay, & Sen?cal, 2004). This also occurs among teachers.  A third notable piece of research involving teachers is Taylor, Ntoumanis, and Standage?s (2008) study, which tested a model linking perceptions of job pressure with need satisfaction (i.e., autonomy, competence, relatedness with colleagues), self-determined motivation (calculated using a weighted average of responses to the different types of 81   regulation), and teachers? use of autonomy-supportive teaching strategies (e.g., providing a meaningful rationale to students) among physical education teachers in the UK. They found that perceived job pressure negatively predicted need satisfaction, which positively predicted self-determined motivation for teaching, and in turn the use of autonomy-supportive teaching strategies. Although this study provides support for the relevance of several key components of SDT among teachers, the use of the weighted average for self-determination does not allow us to understand the relationships between need satisfaction and each different type of regulation. Research examining this will provide a more nuanced understanding of the three needs and how they interact with the different types of regulation. Taken together, these constructs and the premises of SDT establish that autonomy-supportive work climates predict need satisfaction, which in turn predicts well-being and motivation. These relationships are shown in the explanatory model in Figure 3.1 (see paths 1 through 6). The next section elaborates the relationships that well-being and motivation have with job satisfaction and organisational commitment among teachers (see paths 7 through 12 in Figure 3.1). 3.1.2. SDT, Job Satisfaction, and Organisational Commitment SDT provides an ideal framework through which to explore the relationships between teachers? experiences of well-being, motivation, job satisfaction, and organisational commitment. As noted earlier, research has linked job satisfaction and organisational commitment with well-being and motivation (e.g., Collie et al., 2012; Hakanen et al., 2006; Lent et al., 2011). In addition, the use of SDT to theorise about organisational commitment and job satisfaction among employees in general is not unprecedented (e.g., Gagn?, Chemolli, Forest, & Koestner, 2008; Gagn? & Deci, 2005; Meyer & Maltin, 2010). 82   However, the consideration of teacher well-being, motivation, job satisfaction, and organisational commitment simultaneously with the SDT variables of perceived autonomy support and need satisfaction has not been elaborated.  SDT provides a relevant and conceptually-sound framework for examining this given that the components of need satisfaction and self-determination form the underlying mechanisms for the relationships between well-being, motivation, job satisfaction, and organisational commitment. For example, when teachers experience self-determined forms of motivation, it is the accompanying internalisation of the value of work tasks that promotes a sense of job satisfaction and organisational commitment (Gagn? et al., 2008; Gagn? & Deci, 2005). For well-being, flourishing reflects positive feelings about one?s activities, relationships, and self-perceptions (e.g., purpose and meaning, supportive relationships, competence; Diener et al., 2010). In the context of work, I suggest that these positive feelings are associated with one?s work tasks and the place of work, leading to greater job satisfaction and organisational commitment. I believe a model that considers all these constructs simultaneously is needed given that SDT establishes relationships between several of these constructs and empirical evidence suggests relationships between the remaining constructs. In addition, previous models have been concerned with employees more generally. A model that specifically focuses on teachers is important given that teachers? work is unique in many ways (e.g., their ?clients? are students and they are physically isolated from their colleagues for much of the work day; Barnab? & Burns, 1994). Furthermore, this type of examination is necessary in order for us to extend our understanding of SDT in relation to teachers and to obtain a more complete picture of teachers? experiences. As described above, several important studies 83   have examined the components of SDT among teachers (e.g., Klassen, Perry, et al., 2012; Roth et al., 2007; Taylor et al., 2008). In order to advance our understanding of teachers? experiences and the relevance of SDT, research is needed to examine the components of SDT simultaneously and to consider how these components relate to other important outcomes like job satisfaction and organisational commitment. The current study, therefore, aims to extend the literature by exploring these variables simultaneously and in direct relation to teachers. 3.1.2.1. Job satisfaction. Job satisfaction can be defined as a sense of fulfilment that is gained from working in a job (Locke, 1969). Teachers? sense of job satisfaction is important for teachers, students, and schools. For teachers themselves, research has shown that job satisfaction is positively associated with their experiences of positive affect (Mo? et al., 2010), life satisfaction (Lent et al., 2011), and emotion regulation ability (Brackett et al., 2010). In outcomes that extend to students? learning and achievement, and school effectiveness, teachers? job satisfaction is positively associated with teachers? work enthusiasm (Weiqi, 2007), teaching efficacy (Collie et al., 2012), and organisational commitment (Feather & Rauter, 2004).  A significant amount of the research on teachers? job satisfaction is framed within Herzberg, Mausner, and Snyderman?s (1959) two-factor theory, which establishes that job satisfaction is influenced by two different factors. Intrinsic factors, which relate to work tasks (e.g., work performance, professional growth), are responsible for job satisfaction, whereas extrinsic factors of job satisfaction, which relate to peripheral issues surrounding work tasks (e.g., administrative practices, job security), are responsible for job dissatisfaction. Research on the two-factor theory has shown that teachers are generally satisfied with the work-related 84   aspects of their job (e.g., professional growth), but dissatisfied with the working conditions (e.g., physical environment; Butt et al., 2005; Dinham & Scott, 1998). The two-factor theory has been criticised on several grounds, however, one being that it only considers what promotes motivated action, rather than also how motivated action is directed (Gagn? & Deci, 2005). Furthermore, the two-factor theory does not consider the potentially positive impact on employees? job satisfaction of interpersonal relations with supervisors?in the two-factor theory, relations with supervisors are considered a dissatisfying extrinsic factor. Given that research has highlighted the importance of supervisors on individual functioning (e.g., Deci et al., 2001), this is an important factor to consider.  In order to extend our understanding, we need to examine teacher job satisfaction utilising robust theoretical frameworks. SDT provides a more robust framework than the two-factor theory. It addresses the two limitations of the two-factor theory. Namely, it considers both what promotes motivated action (e.g., need satisfaction) and how it is directed (e.g., autonomous motivation; Gagn? & Deci, 2005). In addition, it considers that supervisors may have a positive or negative influence on employee functioning. Perhaps most importantly, support for SDT has been established among a variety of samples and in multiple contexts.  Another important consideration about SDT is that it allows us to examine how job satisfaction simultaneously relates to teacher well-being, motivation, and organisational commitment. This is important given that these four variables are empirically linked. In other words, examining them simultaneously will allow us to extend our understanding of how they influence and are influenced by one another. The explanatory model, therefore, utilises 85   SDT to examine job satisfaction and how it is related to teacher well-being, motivation, and organisational commitment.  3.1.2.2. Organisational commitment. Another aspect of teachers? working lives that can be examined as part of SDT is organisational commitment. Although organisational commitment is related to job satisfaction, research has shown that these two variables are distinct constructs: Job satisfaction refers to feelings of contentment, while work commitment refers to a perceived link or attachment to the organisation (Schleicher, Hansen, & Fox, 2011). In Study 2, organisational commitment is defined more specifically as ?identification with, involvement in, and emotional attachment to the organisation? (Allen & Meyer, 1996, p. 253). This is known as affective organisational commitment (Allen & Meyer, 1996) because it includes an emotional component and is considered a robust measure of organisational commitment (Solinger, van Olffen, & Roe, 2008).  Interest in teachers? organisational commitment has been largely fuelled by concern regarding high rates of turnover and attrition, especially among beginning teachers (Ingersoll, 2002). Affective organisational commitment has been associated with teaching efficacy, and behaviours that are supportive to students, colleagues, and the school (Bogler & Somech, 2004). In addition, organisational commitment more broadly defined has been associated with job satisfaction (Feather & Rauter, 2004), work engagement, and burnout (Hakanen et al., 2006). These outcomes affect teaching performance and, as a result, students? learning and achievement (e.g., Caprara et al., 2006).  SDT provides a framework through which to extend our understanding of teacher organisational commitment. In fact, the use of SDT to examine organisational commitment has been established for employees in general (e.g., Meyer, Becker, & Vandenberghe, 2004; 86   Meyer & Maltin, 2010). The literature has not examined this construct, however, within an SDT framework among teachers. Moreover, the literature has not examined teacher organisational commitment alongside well-being, motivation, and job satisfaction. This is important given that organisational commitment does not occur in isolation?it affects and is affected by teachers? experiences of well-being, motivation, and job satisfaction (e.g., Hakanen et al., 2006; Jackson et al., 2006; Feather & Rauter, 2004). Furthermore, it may function differently when examined concurrently with the other three variables. Thus, the explanatory model will extend the literature by considering all four variables together in relation to teachers and by using the robust definition of affective organisational commitment. At this point, I should mention another type of work commitment that has been commonly examined in the literature: professional or occupational commitment. Professional commitment refers to attachment to the profession of teaching, rather than the school of employment (Klassen & Chiu, 2011). Although professional commitment is likely implicated in the variables I am considering, it is beyond the scope of the model. I have chosen to focus on organisational commitment because it is contextually based and refers directly to a teachers? experiences at his or her current place of work. Professional commitment refers more broadly to experiences in the profession of teaching and may relate to experiences teachers have had at previous schools. Given that the model is focused on current experiences, organisational commitment was deemed more relevant.   87   3.1.3. The Explanatory Model The model elaborated in this study aims to provide understanding of how four important teacher constructs interrelate using the empirically-supported framework of SDT. In addition, it aims to support a program of research that will enable studies of teachers? experiences at work and how these experiences relate to other outcomes relevant for teachers, students, and schools. The hypothesised relationships in the model are described below in relation to theoretical and empirical support.  3.1.3.1. Perceived autonomy support and need satisfaction. As noted earlier, SDT establishes that autonomy-supportive work climates predict satisfaction of the three basic psychological needs among employees. Research in a variety of settings supports this relationship. Of most relevance, Klassen, Perry, et al.?s (2012) study described earlier found that teachers? perceptions of the autonomy support offered by their principals was a positive predictor of their reports of need satisfaction of autonomy, competence, relatedness with colleagues, and relatedness with students. The model incorporates this central view of SDT showing that teachers who perceive autonomy support, in this case from their principals, will report greater need satisfaction for autonomy, competence, and relatedness with colleagues and students as a result.  ? Hypothesis 1. Perceived autonomy support will positively predict need satisfaction (Path 1). 3.1.3.2. Need satisfaction, well-being, and motivation. As previously established, SDT refers to both well-being and motivation indicating that these two constructs are influenced by need satisfaction. Among teachers, two studies that were introduced above have shown support for this relationship. Klassen, Perry, and colleagues (2012) showed that 88   when teachers? needs for autonomy, competence, and relatedness with students were met, they experienced greater work engagement and lower emotional exhaustion. In addition, Taylor and colleagues (2008) found that teachers? reports of a composite need satisfaction variable positively predicted self-determined motivation (a weighted average of the different types of regulation). Given that work engagement is considered an indicator of intrinsic motivation (Klassen, Aldhafri et al., 2012; Leiter & Bakker, 2010) and emotional exhaustion is inversely related to well-being, coupled with Taylor et al.?s findings about teachers? self-determined motivation, these previous studies provide a starting point for understanding how we might expect these relationships to function using the definitions of well-being and motivation used here. Specifically, the model uses the results from these previous studies to provide the basis for the paths from basic psychological needs to both well-being defined as flourishing and autonomous/controlled motivation.  ? Hypothesis 2. Satisfaction of autonomy, competence, and relatedness will positively predict well-being (Path 2). ? Hypothesis 3. Satisfaction of autonomy, competence, and relatedness will positively predict autonomous motivation (Path 3). ? Hypothesis 4. Satisfaction of autonomy, competence, and relatedness will negatively predict controlled motivation (Path 4). 3.1.3.3. Well-being and motivation. Although not a main focus of the model, it is important to consider how well-being and motivation interrelate. Research has shown that a relationship exists between well-being and motivation among teachers. Lent and colleagues (2011) found that self-efficacy for teaching (i.e., motivation) was associated with positive affectivity and life satisfaction (i.e., well-being). Furthermore, Jackson et al. (2006) found 89   that teachers? work engagement (motivation) was negatively associated with emotional exhaustion and stress-induced psychological ill-health (well-being).  Up to this point in time, research examining the relationship between autonomous/controlled motivation and well-being has not been conducted among teachers. However, research in other settings indicates that such relationships are relevant. For example, Lynch, Plant, and Ryan (2005) showed that well-being (vitality and stress) was related to autonomous motivation among hospital employees. While this study provides support for similar relationships among teachers, the work environments at hospitals are very different from schools highlighting the need for research to examine this relationship specifically among teachers.  Although there is evidence to support the relationship between motivation and well-being, the causality remains unknown (e.g., Brouwers & Tomic, 2000; Hakanen, Schaufeli, & Ahola, 2008). Hence, a bi-directional relationship between these two variables is shown in the model. This incorporates the view that motivation and well-being influence one another reciprocally. Teachers who are feeling autonomously motivated by their work may experience greater well-being because they identify with and value their work tasks leading to greater flourishing at work. At the same time, teachers who experience well-being will tend to experience autonomous motivation because their flourishing or healthy functioning will provide fertile ground for internalisation of work tasks. For controlled motivation, it is expected that the non-existent (i.e., external regulation) or limited (i.e., introjected regulation) levels of internalisation that occur will mean a weak, but positive, relationship between it and well-being (i.e., when individuals are not self-determined regarding their work tasks this will relate to less flourishing at work).  90   ? Hypothesis 5. Autonomous motivation and well-being will be positively associated (Path 5). ? Hypothesis 6. Controlled motivation will be positively, but weakly associated with well-being (Path 6). 3.1.3.4. Well-being and job satisfaction. The relationship between well-being and job satisfaction is the first relationship in the explanatory model that goes beyond the main components of SDT. As yet, research has not examined this relationship considering well-being as flourishing, highlighting the need for more research in this area. Using other definitions, research has shown that well-being (e.g., positive affect) is positively associated with job satisfaction among teachers (e.g., Brackett et al., 2010). In terms of the directional nature of this relationship, research suggests that well-being predicts job satisfaction. For example, Wolpin and colleagues (2002) used longitudinal research methods to show that perceptions of burnout predicted lower job satisfaction among teachers, not the other way around.  Research among employees and individuals in other settings also supports this with subjective well-being. Specifically, Bowling and colleagues (2010) conducted a meta-analysis and found overall support for a causal link from subjective well-being to job satisfaction. Although burnout and subjective well-being are defined differently, a similar relationship between flourishing and job satisfaction may occur. This is based on the idea that when teachers are flourishing at work, among other positive outcomes, they experience a sense of meaning and purpose, supportive relationships, and interest in their work activities (Diener et al., 2010). In turn, these experiences mean that teachers are likely to place more value on their work (because they feel a sense of purpose) and enjoy it more (because they 91   have supportive relationships and are interested in what they do), leading to greater job satisfaction. In the model, this reasoning is used to show that well-being positively predicts job satisfaction. ? Hypothesis 7. Well-being will positively predict job satisfaction (Path 7).  3.1.3.5. Well-being and affective organisational commitment. The second relationship involving well-being is that with affective organisational commitment. Empirical support for this relationship among teachers has generally involved other assessments of organisational commitment (i.e., not affective organisational commitment; e.g., Hakanen et al., 2006). One exception, however, is Van Horn, Taris, Schaufeli, and Schreurs? (2004) study conducted among Dutch teachers. Through factor analyses, they found a relationship between self-reported positive affect and affective organisational commitment. However, details about the polarity were not provided. Although we assume this was a positive relationship, it highlights the need for more research involving these constructs.  For further evidence, we turn to research in other settings. Thoresen, Kaplan, Barsky, Warren, and de Chermont (2003) conducted a review of research among employees in a variety of work places (e.g., hospitals, businesses) and showed that reports of positive affect were positively associated, whereas reports of negative affect were negatively associated with affective organisational commitment. These other settings suggest that a similar relationship may occur among teachers, but given the aforementioned differences in work environments between teachers and hospital or office workers, there is a need for research specifically involving teachers.  In terms of the direction of this relationship, the causal nature has not been established (e.g., Jamal, 1990; Jepson & Forrest, 2006). Consequently, longitudinal research 92   is needed in this area. In the model, well-being is shown to predict affective organisational commitment. This is derived from research that has shown stress is a determinant of employee commitment (e.g., Klassen & Chiu, 2011). I believe a similar relationship occurs with well-being. Specifically, when an individual is flourishing he or she experiences meaning and purpose, supportive relationships, and interest in the activities undertaken (Diener et al., 2010). When this occurs at work, I believe it reflects and promotes positive emotions and associations in the individual about the work tasks and place of work. This, in turn, positively influences the emotional connection and affective commitment that an individual has to his or her work (e.g., emotional attachment, identification; Solinger et al., 2008). Study 2 aims to provide an understanding of whether well-being predicts affective organisational commitment; however, future research is needed to examine the direction of this relationship longitudinally in order to test the causality. ? Hypothesis 8. Well-being will positively predict affective organisational commitment (Path 8). 3.1.3.6. Motivation and job satisfaction. The next relationship in the model involves motivation and job satisfaction. Empirical research has provided support for a relationship between these two variables, albeit with narrower definitions of motivation. For example, Barnab? and Burns (1994) showed that self-reported intrinsic work motivation among teachers was related to general satisfaction with work. Research in other settings has corroborated this finding by showing that perceptions of intrinsic motivation were positively associated with reports of job satisfaction among factory employees (Ilardi, Leone, Kasser, & Ryan, 1993) and that perceptions of external regulation were negatively related to reports of job satisfaction among hospital employees (Lynch et al., 2005). Taken together, these studies 93   provide support for a relationship between motivation and job satisfaction. Given that these previous studies used narrower definitions of motivation and/or were conducted among different populations, however, there is a need for research using autonomous/controlled motivation and involving teachers. In order to understand the direction of this relationship, we must turn to research using different definitions of motivation. Caprara and colleagues (2003, 2006) examined the relationship between motivation (defined as self-efficacy for teaching) and job satisfaction and found that motivation was a determinant of job satisfaction as reported by Italian teachers. This provides some support for situating autonomous/controlled motivation as a predictor of job satisfaction (and also highlights the need for more research in this area). Further support is derived from SDT?s premise of internalisation.  Specifically, SDT states that autonomous motivation involves internalising behaviours such that they become more self-determined (Deci & Ryan, 1985, 2002). If behaviours or tasks are internalised then it is more likely that the individual will identify with and value those behaviours/tasks and experience greater job satisfaction as a result (i.e., because the individual perceives he or she is doing meaningful tasks; Gagn? & Deci, 2005). As a result of this understanding, the model shows that autonomous motivation positively predicts job satisfaction. In addition, controlled motivation is shown to negatively predict job satisfaction. This is based on the reverse reasoning. If individuals are acting with controlled motivation, they will have less identification with and place less value on their work tasks (Deci & Ryan, 1985, 2002), likely predicting lower job satisfaction as a result (i.e., if an individual does not value what he or she is doing, then it is less likely to leave that individual feeling satisfied). 94   ? Hypothesis  9. Autonomous motivation will positively predict job satisfaction (Path 9). ? Hypothesis 10. Controlled motivation will negatively predict job satisfaction (Path 10). 3.1.3.7. Motivation and affective organisational commitment. The second relationship involving motivation includes affective organisational commitment. Although research has examined this relationship with other definitions of commitment among teachers (e.g., occupational commitment; Canrinus et al., 2012), little attention has been paid to this relationship using affective organisational commitment. One exception is Bogler and Somech?s (2004) study. They examined motivation (defined as self-efficacy) reported by Israeli teachers and found that it positively predicted affective organisational commitment. In order to further develop our understanding, there is a need for research that examines this relationship using autonomous/controlled motivation.  In order to provide understanding of the directional nature of this relationship, research conducted in other work settings is more forthcoming. Gagn?, Chemolli, Forest, and Koestner (2008) used cross-lagged analysis to examine the causality of the relationship between reports of autonomous/controlled motivation and affective organisational commitment among employees at a large Canadian telecommunications company. Their results showed that autonomous motivation positively predicted affective organisational commitment, and not the other way around. They explained that it is the internalisation that occurs with autonomous motivation that promotes affective organisational commitment?when individuals identify with and value their work tasks, they become attached to their work place as a result. Their results also showed that the two forms of controlled motivation?introjected and external regulation?had weaker positive relationships with 95   affective organisational commitment. This was likely due to the lower levels of internalisation that occurred among these individuals.  Using this previous research as a starting point, I suggest that a similar relationship occurs among teachers. Although office workers? and teachers? working environments are different, internalisation is an underlying mechanism that is relevant in both working contexts (i.e., internalisation can occur in any context depending on the supports for self-determination). Consequently, the model shows that autonomous and controlled motivation predict affective organisational commitment. The nature of these two relationships differ though. Autonomous motivation is shown to positively predict affective organisational commitment. This is based on reasoning described above by Gagn? et al. (2008), which suggests that individuals (in this case, teachers) who are autonomously motivated will experience greater affective organisational commitment due to the internalisation that occurs as part of autonomous motivation. In contrast, controlled motivation is shown to negatively predict affective organisational commitment. This is different from Gagn? et al.?s findings and is based on the idea that teachers who are acting with controlled motivation will experience lower affective organisational commitment due to the non-existent or limited levels of internalisation that occur regarding work tasks and the organisation. ? Hypothesis 11. Autonomous motivation will positively predict affective organisational commitment (Path 11). ? Hypothesis 12. Controlled motivation will negatively predict affective organisational commitment (Path 12).  96   3.1.3.8. Job satisfaction and affective organisational commitment. The final relationship in the model is that between teacher job satisfaction and affective organisational commitment. Very little research has examined this relationship among teachers using the definitions chosen in this dissertation?research has examined this relationship, however, using other definitions of commitment (e.g., occupational commitment; Canrinus et al., 2012). A notable exception is Feather and Rauter?s (2004) study conducted among Australian teachers. They found that perceptions of job satisfaction were positively related to affective organisational commitment. Given that the directional nature between these two variables remains unclear (e.g., Billingsley & Cross, 1992), the model incorporates a bi-directional view. Job satisfaction is shown to be positively associated with affective organisational commitment. This is based on the idea that teachers who are satisfied with their work will enjoy their work more, resulting in positive emotions and associations and higher affective organisational commitment as a result. At the same time, teachers who are highly committed may experience greater job satisfaction because they feel their time is spent on a worthwhile cause. This interpretation is supported by Billingsley and Cross (1992) who proposed that the two variables evolve simultaneously. ? Hypothesis 13. Job satisfaction and affective organisational commitment will be positively associated (Path 13). 3.1.4. Overview of the Current Study Research has shown that teachers? perceptions of principal autonomy support predict need satisfaction (Klassen, Perry, et al., 2012), which in turn, predicts emotional exhaustion (negatively) and engagement (positively; Klassen, Perry, et al., 2012), as well as self-determined motivation (Taylor, et al., 2008). Research has also suggested that the different 97   types of regulation proposed by SDT are appropriate for examining teacher motivation (Roth et al., 2007). There remain, however, several aspects of SDT untested among teachers including the relationships that need satisfaction has with well-being defined as flourishing and the different types of regulation (e.g., identified, introjected). Additionally, relationships that well-being (defined as flourishing) and autonomous/controlled motivation have with job satisfaction and affective organisational commitment have been largely untested among teachers. In order to address these gaps, Study 2 involved hypothesising and testing the explanatory model of teacher well-being, motivation, job satisfaction, and affective organisational commitment shown in Figure 3.1, and more specifically, the hypotheses detailed above. Figure 3.2 shows a summary of the hypotheses tested in the current study.    Figure 3.2. Summary of the hypotheses tested with predicted relationship polarity.   98   3.2. Methods Data for this study were drawn from the same data that were collected for Study 1.  Specifically, data were examined from the Teacher Well-Being Scale (TWBS) developed in Study 1, as well as from the more established flourishing and job satisfaction measures used in Study 1. Additional measures included in Study 2 are described below. These measures were included to examine relationships theorised in the explanatory model.  3.2.1. Participants and Procedures The participants and procedures for data collection in Study 2 are identical to Study 1. To recap, participants included 603 teachers recruited through four different district-level teachers? associations in British Columbia, Canada. The sample statistics were identical to Study 1. In an attempt to further extend our understanding of well-being among teachers, two types of well-being were examined in Study 2. The first type, teacher well-being, was used to measure teachers? work-related well-being. The second type, a measure of flourishing was used to assess global (i.e., general life) well-being of the participants. Together, these two types of well-being provide an idea of how work-related and global well-being function in relation to the other variables under examination in the model.  3.2.2. Measures Participants completed an online, self-report questionnaire that included measures of perceived autonomy support, need satisfaction, flourishing, teacher well-being, motivation at work, job satisfaction, and organisational commitment. Table 3.1 shows the descriptive statistics and Cronbach?s alphas for the measures used in the current study. As introduced in Study 1, Table 2.5 shows the constructs (and item location in Appendices) that were examined in Study 2, along with those examined in the other two studies.   99   3.2.2.1. Perceived autonomy support. An adapted version of the Work Climate Questionnaire  (WCQ; Baard et al., 2004) was used to measure perceived autonomy support from principals (see Appendix B, Section 2, Questions 1 through 6). The original 6-item WCQ contains statements about the employee?s manager (e.g., ?I feel understood by my manager?) and the participants are asked to respond on a scale ranging from Strongly Disagree (1) to Strongly Agree (7). In the adapted version, Klassen, Perry, and colleagues (2012) replaced the word manager with principal and found high internal consistency for the adapted measure (? = .95). Using this adapted measure, the Cronbach?s alpha in the current study was .96. 3.2.2.2. Need satisfaction. The Work-related Basic Need Satisfaction scale (Van den Broeck, Vansteenkiste, De Witte, Soenens, & Lens, 2010) was used to measure teachers? sense of need satisfaction for autonomy, competence, and relatedness with colleagues (see Appendix B, Section 3, Questions 1 through 18). The scale consists of 18 statements (e.g., ?I really master my tasks at my job?) and participants respond to these on a scale ranging from Strongly Disagree (1) to Strongly Agree (7). The measure has been used with a large sample of Dutch-speaking workers in the Netherlands (Van den Broeck et al., 2010), as well as with a large sample of Canadian teachers (Klassen, Perry, et al., 2012). In both examples, support was found for the factor structure, validity, and reliability of the scores from the measure (Cronbach?s alphas in the two studies ranged from .77 to .85 for the three subscales). In the current study, Cronbach?s alphas ranged from .84 to .88 (see Table 3.1). In educational settings, teachers not only have relationships with colleagues, but also students. Four additional items were included, therefore, to measure relatedness with students (e.g., ?I am very committed to my students?; see Appendix B, Section 3, Questions 19 100   through 22). Klassen, Perry, et al. (2012) adapted these items from Deci and colleagues? (2001) relatedness scale and they were measured on the same 7-point scale as the other need satisfaction items. High internal consistency was found for the items in Klassen, Perry, et al.?s study (? = .80), as well as in the current study (? = .88). 3.2.2.3. Well-being and job satisfaction. Two measures were used to measure well-being and one was used to measure job satisfaction. These three measures?the Flourishing Scale (Diener et al., 2010), the TWBS (developed in Study 1), and the job satisfaction items Caprara and colleagues (2003) developed?were described in Study 1. The Flourishing Scale was used to measure global well-being and the TWBS was used to measure work-related well-being. 3.2.2.4. Motivation. The Motivation at Work Scale (Gagn? et al., 2010) was used to assess teachers? levels of autonomous and controlled motivation (see Appendix B, Section 4, Questions 1 through 12). Items in this measure provide 12 reasons for doing a certain job (e.g., ?Because this job fulfils my career plans?) and participants indicate the extent to which each reason is accurate for them on a scale ranging from Not At All (1) to Exactly (7). The measure contains four subscales that refer to different types of behaviour regulation: intrinsic regulation, identified regulation, introjected regulation, and external regulation. The scale does not include integrated regulation because it is psychometrically difficult to differentiate from intrinsic motivation (Gagn? et al., 2010). Autonomous motivation, therefore, is represented by intrinsic and identified regulation and controlled motivation is represented by introjected and external regulation. Gagn? and colleagues (2010) found that this measure had adequate reliability (with Cronbach?s alphas between .69 and .89) and a good factor structure. Evidence of validity was also provided through comparison with hypothesised 101   antecedents and outcomes, and through a simplex pattern (Gagn? et al., 2010). In the current study, the factors had Cronbach?s alphas ranging from .73 to .91 (see Table 3.1). 3.2.2.5. Affective organisational commitment. An adapted version of the Vandenberghe and Bentein (2009) six-item scale was used to measure affective organisational commitment (see Appendix B, Section 2, Questions 15 through 20). I adapted the original scale by replacing the word organisation with school or school community. This measure consists of statements about the organisation (e.g., ?I am proud to belong to this school community?) and participants respond on a scale ranging from Strongly Disagree (1) to Strongly Agree (7). Previous studies have found adequate reliability for the scale (Cronbach?s alphas between .69 and .78; Vandenberghe et al., 2007; Vandenberghe & Bentein, 2009) and good factor structure (Vandenberghe et al., 2007). The Cronbach?s alpha for the current study with the adapted items was .86. 3.2.2.6. Demographic information. The same demographic information was collected as for Study 1 (see Appendix B, Section 1, Questions 1 through 14).   102    Table 3.1 Reliability Indexes, Means, and Standard Deviations of All Variables     Range Variable ? M SD Potential Observed Perceived autonomy support .96 5.05 1.401 1.00 ? 7.00 1.00 ? 7.00 Relatedness with colleagues .85 4.33 1.276 1.00 ? 7.00 1.00 ? 7.00 Relatedness with students .88 6.54 0.602 1.00 ? 7.00 2.25 ? 7.00 Competence .86 5.73 0.979 1.00 ? 7.00 1.17 ? 7.00 Autonomy .84 4.65 1.203 1.00 ? 7.00 1.00 ? 7.00 Flourishing .83 6.54 0.602 1.00 ? 7.00 2.25 ? 7.00 Workload well-being .85 3.43 1.055 1.00 ? 7.00 1.00 ? 7.00 Organisational well-being .82 4.38 1.065 1.00 ? 7.00 1.00 ? 7.00 Student interaction well-being .84 4.33 1.276 1.00 ? 7.00 1.00 ? 7.00 Intrinsic regulation .91 5.52 1.112 1.00 ? 7.00 1.00 ? 7.00 Identified regulation .84 5.13 1.156 1.00 ? 7.00 1.00 ? 7.00 Introjected regulation .83 2.73 1.434 1.00 ? 7.00 1.00 ? 7.00 External regulation .73 2.74 1.189 1.00 ? 7.00 1.00 ? 7.00 Job satisfaction .90 5.36 0.881 1.00 ? 7.00 1.00 ? 7.00 Organisational commitment .86 5.55 1.276 1.00 ? 7.00 1.33 ? 7.00  103   3.3. Results 3.3.1. Data Analyses The factor structure of the constructs was examined using exploratory and confirmatory factor analyses. Following this, structural equation models (SEM) were used to simultaneously examine the relationships among all factors. As described in Study 1, factor analysis involves exploring and then confirming the factor structure in the data. This provides the opportunity to test through EFA and confirm through CFA whether the items are associated with the expected constructs. Furthermore, it is an important aspect of determining internal structure and is a necessary step before conducting SEM to ensure that the structural model is based on a sound measurement model or CFA (Kline, 2011).  SEM allows researchers to examine relationships among multiple predictor and outcome variables. In that sense, it examines paths between variables like regression analysis, but instead of using observed variables, it involves the factors or latent variables determined through the CFA measurement model (Tabachnick & Fidell, 2007). SEM is an appropriate tool for research that is based on a priori specifications of relationships among variables and their directionalities (Kline, 2011). In other words, the researcher determines the structure of the model and the direction of each relationship a priori using existing research and theory (Martin, 2007). Once the model is specified, SEM is used to determine whether the solution is well-defined and fits the data, as well as whether the a priori predictions are supported (Kline, 2011; Martin, 2007). Although directional relationships are included in the integrated model, the cross-sectional design of the current study does not allow for causal claims to be made about the relationships. In other words, the findings indicate the plausibility of the directional relationships in the explanatory model (i.e., using 104   SEM we test the hypothesised causal model using sample data; Kline, 2011). Conclusive claims about directionality, however, cannot be made as they require longitudinal/experimental examination. This is particularly true in the case of the paths where previous longitudinal research does not exist.  Similar to factor analysis, it is important to consider the hierarchical nature of the data before conducting SEM. As in Study 1, the data examined in the current study were hierarchical in nature (i.e., teachers were nested in schools). Accordingly, the same procedures used in Study 1 were employed in Study 2 to address this. Namely, robust (sandwich estimator) standard errors and a robust chi-square test were used (Asparouhov & Muth?n, 2005; Muth?n & Muth?n, 1998-2012). In addition, the ?TYPE=COMPLEX? option was used with the cluster variable set as participants? school of employment. All analyses were conducted using Mplus Version 7.1 (Muth?n & Muth?n, 1998-2012). The same fit indices and guidelines stipulated in Study 1 were used in the current study and are summarised in Table 3.2. In addition, the chi-square model fit test is reported. Missing data for this study ranged from 0-6% for all the variables. Typically, cases with missing data are excluded from analyses, which can lead to reduced power and loss of important information (Tabachnick & Fidell, 2007). In order to avoid this, multiple imputation was conducted for missing values on the predictor variables using Bayesian estimation through Mplus. For the outcome variables, methods robust against missing data were utilised for data analysis.     105    Table 3.2 Guidelines for Fit Indices Fit index Evidence of adequate fit Evidence of good fit RMSEA < .10 < .06 CFI > .90 > .95 SRMR < .10 <. 08   3.3.2. Factor Analyses Data from the measures were subjected to several exploratory factor analyses (EFAs) on a randomly selected half of the dataset. The EFAs were performed with the robust weighted least squares (WLSMV) estimator and geomin oblique rotation. This method of estimation is appropriate for categorical data and is robust to missing and non-normal data. For the EFAs, the items all loaded well. Given the highly related nature of the constructs, however, cross-loading was an issue with some items. The decision was made to exclude these items. In addition, all the items associated with two constructs obtained high cross-loadings. The two constructs were intrinsic regulation and job satisfaction. This is understandable given that intrinsic regulation is driven by enjoyment or interest in the behaviour. In other words, if teachers? motivation is intrinsically regulated, they teach because it is enjoyable, and by logical extension it is highly likely they would find it satisfying. Although the relationship between the constructs is understandable, it is problematic psychometrically. It suggests that these two constructs cannot be differentiated 106   by the measures used in the current study. By including both of them in the statistical analyses, it would increase the threat of multicollinearity and reduce the robustness of results. Given this, the intrinsic regulation items were removed from analyses; however, autonomous motivation was still represented by the items relating to identified regulation that loaded well and did not cross-load.  For the final EFA, the intrinsic regulation items were excluded along with any other items that cross-loaded within .100 leaving between three and six items per construct. The model showed good fit: ?2 (1021, N = 241) = 1213.095, p < .001, RMSEA = .028, CFI = .989, and SRMR = .023. Table 3.3 shows the items that were included in the final model for each construct and their factor loadings.  Following the EFAs, a CFA was run on the other half of the dataset to confirm the factor structure. Once again, the robust WLSMV estimator was used. Given the highly related nature of the constructs involved in the model, the decision was made to combine the teacher well-being subscales into a higher-order factor to reduce the possibility of multicollinearity, which is problematic in statistical analyses as it can inflate the error terms, thus reducing the power of the analyses (Tabachnick & Fidell, 2007). Multicollinearity occurs when there is high correlation among factors. Using this second-order factor for teacher well-being, the fit indices revealed good fit of the confirmatory measurement model: ?2 (1641, N = 244) = 2377.067, p < .001, RMSEA = .043, and CFI = .959 (SRMR was not available for this estimation method). Table 3.3 shows the standardised factor loadings for the items from the CFA. The latent variable correlations among the factors using the whole dataset are shown in Table 3.4 and provide evidence that the factors are distinct constructs.   107   Table 3.3 Factor Loadings from the Factor Analyses Items Rotated factor loadings from EFA Standardised loadings from CFA Perceived autonomy support   My principal listens to how I would like to do things .897 .951 I feel that my principal provides me with choices and options .930 .944 I feel understood by my principal .960 .962 My principal conveys confidence in my ability to do well at my job .877 .864 My principal encourages me to ask questions .909 .883 My principal tries to understand how I see things before suggesting a new way of doing things .905 .903 Relatedness with colleagues   I don?t really feel connected with other people at work (R) .855 .841 I don?t really mix with colleagues at work (R) .806 .845 At work, I can talk with people about things that really matter to me .722 .678 I often feel alone when I am with my colleagues (R) .899 .793 Relatedness with students   I am very committed to my students .828 .863 Connecting with my students is an essential part of the job .934 .908 I value the relationships I build with my students .921 .960 I feel connected to my students .906 .889   108   Table 3.3 (Continued)   Items Rotated factor loadings from EFA Standardised loadings from CFA Competence   I don?t really feel competent in my job (R) .871 .844 I really master the things I have to do at work .827 .815 I feel competent at my job .924 .905 I have doubts about whether I am able to do my job properly (R) .740 .700 I am good at the things I do in my job .887 .924 I have the feeling that I can accomplish even the most difficult tasks at work .724 .819 Autonomy   At work, I often feel like I have to follow other people?s commands (R) .679 .636 If I could choose, I would do things at work differently (R) .782 .675 The things I have to do at work are in line with what I really want to do .778 .876 I feel free to do my job the way I think it should be done .859 .876 In my job, I feel forced to do things I don?t want to do (R) .819 .629 Flourishing   I lead a purposeful and meaningful life .924 .833 My social relationships are supportive and rewarding .767 .748 I am engaged and interested in my daily activities .814 .849 I actively contribute to the happiness and well-being of others .685 .796 I am a good person and live a good life .755 .717 I am optimistic about my future .571 .707 109   Table 3.3 (Continued)   Items Rotated factor loadings from EFA Standardised loadings from CFA Workload well-being   Marking work .671 .694 Fitting everything in to the allotted time .654 .638 Administrative work related to teaching .708 .692 Work I complete outside of school hours for teaching .814 .686 Working to finish my teaching tasks .703 .827 Staying late after work for meetings and activities .691 .732 Organisational well-being   Support offered by school leadership .675 .828 Recognition for my teaching .522 .680 School rules and procedures that are in place .770 .694 Communication between members of the school .734 .683 Participation in school-level decision making .730 .678 Student interaction well-being   Student behaviour .948 .921 Student motivation .739 .886 Classroom management .633 .782 Identified regulation   I chose this job because it allows me to reach my life goals .952 .885 Because this job fulfils my career plans .852 .862 Because this job fits my personal values .678 .778   110   Table 3.3 (Continued)   Items Rotated factor loadings from EFA Standardised loadings from CFA Introjected regulation   Because I have to be the best in my job, I have to be a ?winner? .784 .743 Because my work is my life and I don?t want to fail .915 .813 Because my reputation depends on it .761 .887 External regulation   Because this job affords me a certain standard of living .848 .662 Because it allows me to make a lot of money .818 .722 I do this job for the paycheque .611 .935 Job satisfaction   I am satisfied with my job .822 .916 I am satisfied with what I achieve at work .880 .887 I feel good at work .848 .909 Affective organisational commitment   This school has a great deal of personal meaning for me .865 .894 I am proud to belong to this school community .876 .922 I do not feel emotionally attached to my school (R) .829 .732 Note. (R) refers to reverse coded items.111   Table 3.4 Correlations of Latent Variables from the Confirmatory Factor Analysis  1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 1. PAS               2. Rel. with colleagues .381              3. Rel. with students .183 .283             4. Competence .146 .253 .432            5. Autonomy .546 .476 .274 .418           6. Flourishing .312 .460 .471 .644 .504          7. Workload WB .321 .250 .148 .188 .360 .275         8. Organisational WB .494 .384 .228 .289 .554 .423 .484        9. Student interaction WB .412 .321 .191 .241 .463 .353 .404 .622       10. Teacher well-being .572 .445 .264 .335 .642 .490 .561 .863 .720      11. Identified regulation .188 .318 .458 .386 .497 .521 .242 .373 .311 .432     12. Introjected regulation -.070 -.091 .051 -.059 -.095 -.085 -.045 -.069 -.057 -.079 .192    13. External regulation -.009 -.113 -.371 -.139 -.081 -.205 -.005 -.007 -.006 -.008 -.057 .372   14. Job satisfaction .350 .440 .420 .636 .637 .648 .359 .552 .461 .640 .595 -.038 -.174  15. Aff. org. commitment .426 .653 .494 .366 .457 .556 .272 .419 .349 .485 .473 .113 -.184 .539 Note. Correlations with an absolute value equal to or greater than r = .90 are significant at p < .05, those with an absolute value equal to or greater than r = .117 are significant at p < .01, and those with an absolute value equal to or greater than r = .150 are significant at p = .001. All other correlations are not significant. PAS = perceived autonomy support. Rel.  = relatedness. WB = well-being. Aff. org. commitment = affective organisational commitment. 112   3.3.3. Structural Modelling Following the factor analyses, the relationships between the variables in Figure 3.1 were analysed with SEM on the full dataset in order investigate whether they related to one another as predicted. A model was specified in Mplus 7.1 (Muth?n & Muth?n, 1998-2012) that corresponded with paths elucidated in the explanatory model. The constructs specified in this model referred to the items that were confirmed through the factor analyses (i.e., the measurement model provided by the CFA was used). By running the SEM, coefficients were estimated for the relationships specified in the explanatory model. When a parameter estimate was significant, it indicated that there was a reliable relationship between the two constructs as expected. In contrast, non-significant parameter estimates indicated that there was no significant path between those variables in the model. Non-significant paths, starting with the lowest standardised coefficients, were deleted for reasons of parsimony one at a time until only significant paths remained in the model. This provided the final model, which was then examined for goodness of fit. The fit indices suggested adequate to good fit of the final model: ?2 (1680, N = 485) = 4036.828, p < .001, RMSEA = .054, and CFI = .933 (SRMR was not available for this estimation method).  Figure 3.3 shows the final model. Given that non-significant paths were removed from the model, all of the path coefficients are statistically significant at p < .05. Standardised direct, total indirect, and total effects are shown in Table 3.5. These effects indicate the parameter estimates that occur from one variable directly to another (direct effects), the parameter estimates between two variables when they are mediated by one or more other variables (indirect effects), and the total strength of the relationship considering both direct and indirect paths (total effects). 113    Figure 3.3. Structural equation model of perceived autonomy support, need satisfaction, well-being, motivation, job satisfaction, and affective organisational commitment. Only significant paths are shown. All coefficients are significant (p < .05). Standardised coefficients are reported. Significant correlations also examined but not shown in the figure included identified regulation and flourishing (r = .48, p < .001), identified regulation and teacher well-being (r = .39, p < .001), external regulation and flourishing (r = -.196, p < .001), and job satisfaction and affective organisational commitment (r = .54, p < .001).114   Table 3.5 Standardised Effects for the Predictor Variables on Each Outcome Variable  Direct Total indirect Total Predictor: Perceived autonomy support    Relatedness with colleagues .508 ? .508 Relatedness with students .313 ? .313 Competence .284 ? .284 Autonomy .629 ? .629 Flourishing ? .473 .473 Teacher well-being ? .581 .581 Identified regulation ? .384 .384 External regulation ? -.111 -.111 Job satisfaction ? .513 .513 Organisational commitment ? .464 .464 Predictor: Relatedness with colleagues    Flourishing .224 ? .224 Teacher well-being .330 ? .330 External regulation -.204 ? -.204 Job satisfaction ? .209 .209 Organisational commitment ? .270 .270   115   Table 3.5 (Continued)     Direct Total indirect Total Predictor: Relatedness with students    Flourishing .336 ? .336 Identified regulation .322 ? .322 External regulation -.417 ? -.417 Job satisfaction ? .239 .239 Organisational commitment ? .259 .259 Predictor: Competence    Flourishing .574 ? .574 Teacher well-being .221 ? .221 Identified regulation .331 ? .331 Job satisfaction ? .419 .419 Organisational commitment ? .297 .297 Predictor: Autonomy    Flourishing .145 ? .145 Teacher well-being .557 ? .557 Identified regulation .300 ? .300 External regulation .195 ? .195 Job satisfaction ? .339 .339 Organisational commitment ? .257 .257    116   3.3.3.1. Basic psychological needs. The first findings refer to the relationships between perceived principal?s autonomy support and teachers? sense of autonomy, competence, and relatedness?their basic psychological needs. It was hypothesised that perceived autonomy support would positively predict need satisfaction (Hypothesis 1). The findings revealed that participants? perceptions of principals? autonomy support were reliably related to their sense of need satisfaction of relatedness with colleagues (? = .508, p < .001), relatedness with students (? = .313, p < .001), competence (? = .284, p < .001), and autonomy (? = .629, p < .001). All relationships were positive revealing that as perceptions of principals? autonomy support increased, so did teachers? need satisfaction. These findings provide support for Hypothesis 1. In terms of the effect sizes, the model explained different amounts of variance in the basic psychological needs. It explained the most variance in autonomy (39.6%), followed by relatedness with colleagues (25.8%), relatedness with students (9.8%), and competence (8.1%).  3.3.3.2. Well-being. The second set of findings refers to the relationship between the basic psychological needs and the two types of well-being. It was hypothesised that need satisfaction would positively predict flourishing (i.e., the global measure of well-being) and teacher well-being (i.e., work-related well-being; Hypothesis 2). The findings revealed that participants? need satisfaction of relatedness with colleagues (? = .224, p < .001), relatedness with students (? = .336, p < .001), competence (? = .574, p < .001), and autonomy (? = .145, p < .001) were reliably related to their reports of flourishing. In addition, need satisfaction of relatedness with colleagues (? = .330, p < .001), competence (? = .221, p < .001), and autonomy (? = .557, p < .001) were reliably related to reports of teacher well-being. These findings indicate that as teachers? need satisfaction increased, so did their reports of 117   flourishing and teacher well-being (providing support for Hypothesis 2). The model explained a large amount of the variance in well-being?67.8% of the variance in flourishing and 65.0% of the variance in teacher well-being?suggesting quite a large effect size for both types of well-being.  3.3.3.3. Motivation. The relationships between the basic psychological needs and the different types of motivation are reported next. It was hypothesised that need satisfaction would positively predict autonomous motivation (Hypothesis 3). As described above, autonomous motivation was measured with one type of regulation, identified regulation. Participants? need satisfaction of relatedness with students (? = .322, p < .001), competence (? = .331, p < .001), and autonomy (? = .300, p < .001) were reliably related to their reports of identified regulation. This finding indicates that as teachers? perceptions of relatedness with students, competence, and autonomy increased, so did their reports of identified regulation (providing support for Hypothesis 3).  It was also hypothesised that need satisfaction would negatively predict controlled motivation (Hypothesis 4). For controlled motivation, two types of regulation were examined: external and introjected regulation. The findings showed that relatedness with colleagues (? = -.204, p < .001), relatedness with students (? = -.417, p < .001), and autonomy (? = .195, p < .001) were reliably related to teachers? reports of external regulation, yet they had different polarities. Teachers who perceived need satisfaction with respect to relatedness with colleagues and students, reported reduced external motivation (as per Hypothesis 4). Contrary to expectations, however, teachers who perceived need satisfaction for autonomy, reported higher levels of external regulation. In addition, none of the basic psychological needs were reliably related to introjected regulation. Turning to the 118   effect sizes, the model explained 39.6% of the variance in identified regulation and 22.3% of the variance in external regulation. It did not explain any of the variance in introjected regulation because that only acted as an exogenous variable in the model (i.e., it did not act as an outcome variable). 3.3.3.4. Job satisfaction. The next findings concern the relationships that well-being and motivation had with job satisfaction. It was hypothesised that flourishing and teacher well-being would positively predict job satisfaction (Hypothesis 7), as would autonomous motivation (Hypothesis 9). The findings revealed that participants? reports of flourishing (? = .433, p < .001), teacher well-being (? = .339, p < .001), and identified regulation (i.e., the measure of autonomous motivation; ? = .291, p < .001) were reliably and positively related to their reports of job satisfaction. As reports of flourishing, teacher well-being, and identified regulation increased, so did reports of job satisfaction, providing support for Hypothesis 7 and 9.  It was also hypothesised that controlled motivation would negatively predict job satisfaction (Hypothesis 10). However, controlled motivation, measured as external and introjected regulation, did not predict lower job satisfaction. In fact, external and introjected regulation were not significantly related to job satisfaction at all. In terms of the effect size, the model explained 73.0% of the variance in job satisfaction. 3.3.3.5. Affective organisational commitment. The relationships that well-being and motivation had with affective organisational commitment are reported next. It was hypothesised that flourishing and teacher well-being (Hypothesis 8), along with autonomous motivation (i.e., identified regulation; Hypothesis 11), would positively predict affective organisational commitment. The findings revealed that participants? sense of flourishing (? = 119   .285, p < .001), teacher well-being (? = .435, p < .001), and identified regulation (? = .112, p < .001) predicted their reports of affective organisational commitment. As teachers? reports of flourishing, teacher well-being, and identified regulation increased, so did their reports of affective organisational commitment providing support for Hypotheses 8 and 11. It was also hypothesised that controlled motivation (i.e., introjected and external regulation) would negatively predict affective organisational commitment (Hypothesis 12). The findings showed that introjected regulation (? = .147, p < .001) and external regulation (? = -.306, p < .001) predicted teachers? reports of affective organisational commitment, albeit in different ways. External regulation was negatively associated with affective organisational commitment as hypothesised; however, contrary to expectations introjected regulation was positively associated. This means that participants who reported teaching with external regulation experienced lower affective organisational commitment, whereas participants who reported teaching with introjected regulation experienced greater affective organisational commitment. Together, the findings provided only partial support for Hypothesis 12. In terms of the effect size, the model explained a large amount of variance in affective organisational commitment (63.4%). 3.3.3.6. Correlational hypotheses. In addition to the directional paths shown in Figure 3.3, there were three additional hypotheses concerning correlational relationships (Hypotheses 5, 6, and 13). These were not shown in Figure 3.3 given its existing complexity with the directional paths. As described earlier, Hypothesis 5 proposed that autonomous motivation would be positively associated with flourishing and teacher well-being. Consistent with these expectations, latent variable correlations estimated from the SEM 120   revealed that autonomous motivation (measured as identified regulation) was positively associated with flourishing (r = .48, p < .001) and teacher well-being (r = .39, p < .001).  The next correlational hypothesis, however, was not supported. Hypothesis 6 indicated that controlled motivation (i.e., introjected and external regulation) would have a weak, positive relationship with well-being. External regulation, however, was negatively related with flourishing (r = -.196, p < .001) and not significantly related with teacher well-being (r = -.049, ns). Furthermore, introjected regulation, was not significantly related to flourishing (r = -.041, ns) or teacher well-being (r = -.050, ns). The final hypothesis was that job satisfaction and affective organisational commitment would be positively related (Hypothesis 13). The findings were consistent with this hypothesis: Job satisfaction was positively associated with affective organisational commitment (r = .54, p < .001).  3.4. Discussion Improving teachers? experiences at work is a crucial component of efforts to promote a whole host of positive outcomes in schools. Teachers? who are faring well, feeling motivated to teach, experiencing satisfaction with their work, and feeling committed to their school of employment tend to be more effective teachers (e.g., Duckworth et al., 2009) and better at promoting students? motivation (e.g., Pakarinen et al., 2010) and achievement (e.g., Caprara et al., 2006). However, we still have a great deal to learn about how teachers? experiences of well-being, motivation, job satisfaction, and affective organisational commitment interact with one another simultaneously. Furthermore, we do not have explanatory models to enable us to study these important variables concurrently. Study 2 aimed to address these gaps. Grounded in self-determination theory (SDT), an explanatory model of relationships between teachers? experiences of well-being, motivation, job 121   satisfaction, and affective organisational commitment was elaborated and tested. After confirming the factor pattern, SEM was conducted and provided support for the explanatory model. Not all specified paths were consistent, however, with expectations. Key findings in relation to the hypotheses are discussed below. 3.4.1. Basic Psychological Needs, Well-being, and Motivation The findings revealed that the relationships in the model that were taken from SDT were generally supported. Namely, teachers? perceptions of principal?s autonomy support predicted satisfaction of the needs for relatedness with colleagues and students, competence, and autonomy (Hypothesis 1). Furthermore, need satisfaction positively predicted well-being and identified regulation (i.e., autonomous motivation), but negatively predicted external regulation (i.e., controlled motivation; Hypotheses 2, 3, and 5). The relationship found between perceived autonomy support and need satisfaction corroborates Klassen, Perry, et al.?s (2012) research. For the relationships involving need satisfaction, well-being, and motivation, however, the findings extend the literature given that these relationships have not been examined previously among teachers.  In combination, these findings extend our understanding of teachers? experiences at work and how they interrelate. This is particularly important given that well-being and motivation have not only been linked with important teacher outcomes, such as teacher effectiveness (e.g., Duckworth et al., 2009), but also student achievement (e.g., Caprara et al., 2006). The findings also confirm the relevance of SDT for guiding research among teachers. Previous research has provided support for the use of SDT in examining well-being and motivation among students (e.g., Filak & Sheldon, 2008) and employees in other work-related settings (e.g., Deci et al., 2001). The current study corroborates and extends the 122   emerging literature on teachers by showing that SDT is relevant for examining well-being, autonomous motivation, and controlled motivation among teachers.  The implications of these findings are relevant for school administrators and policy-makers for several reasons. In particular, they highlight the importance of principals? autonomy support?or more accurately, teachers? perceptions of this. By ensuring that principals promote an autonomy-supportive working environment, teachers are more likely to perceive this to be the case. In turn, teachers will likely experience greater need satisfaction, well-being, and motivation. In order to extend our understanding, future research should examine whether and how teachers? perceptions of autonomy support compare with actual efforts by principals to be autonomy-supportive. In addition, although research has shown the types of managerial behaviours that employees in factory settings find autonomy-supportive (e.g., Deci et al., 1994), no research has examined this among teachers. This type of research would be helpful for guiding efforts to improve teachers? perceived autonomy support and the many important outcomes related with it (e.g., teacher motivation). 3.4.1.1. Autonomy and external regulation. Although support was found for the main premises of SDT, one finding that was contrary to expectations was the positive relationship between the need for autonomy and external regulation. I hypothesised that there would be a negative relationship between these two variables (Hypothesis 4). The negative latent variable correlation from the CFA supported my original hypothesis. The relationship between these two variables became positive, however, in the structural modelling, suggesting that there may be suppression effects occurring due to unobserved variables. 123   One possible reason for the changing sign is that high autonomy may have been reported by two different groups of teachers. According to Pearson and Moomaw (2005), autonomy can be viewed differently by teachers. Where one teacher may view autonomy as the freedom to adjust his or her teaching, another teacher may ?view autonomy as a means to gain substantial freedom from interference or supervision? (Pearson & Moomaw, 2005, p. 42). In this sense, perhaps one group of teachers reported high autonomy with the understanding that it allows them greater freedom to adjust their teaching methods and the content to best meet their students? needs. In contrast, perhaps a second group reported high autonomy because it means there is less monitoring of their teaching by supervisors and, as a result, less pressure to put additional effort into their teaching (e.g., in the case of teachers who are experiencing disengagement due to burnout; Demerouti, Bakker, Nachreiner, & Schaufeli, 2001). Although the data do not allow us to determine if this was the case, this interpretation is reflected in the model by the fact that autonomy was positively associated with both identified regulation and external regulation. If these two groups do exist, teachers from the first group may have reported high identified regulation, whereas teachers from the second group may have reported high external regulation. This interpretation is also supported by Gagn? et al. (2010), who found that need satisfaction of autonomy was positively associated with identified and external regulation (along with other types too). Their results suggest that while autonomy predicts autonomous motivation, it is also related?psychometrically at least?to controlled motivation. Given all this, it is conceivable that the swinging beta was caused by this double-sided view of autonomy. Unfortunately, we are not able to determine if this is the case from the current data, but future research should examine this relationship in 124   greater depth to understand why satisfaction of the need for autonomy predicts both external and identified regulation.  3.4.1.2. Introjected regulation. Another finding that was contrary to expectations was that need satisfaction did not predict introjected regulation (Hypothesis 4). Perhaps for teachers who perceive they need to avoid shame/failure or who rely on contingent self-worth?beliefs associated with introjected regulation (e.g., Deci & Ryan, 2002)?need satisfaction is not a primary concern. In other words, for these individuals, relatedness with colleagues and students, competence, and autonomy may be secondary to their need to meet the goals of introjected regulation (e.g., avoid failure or be seen positively in the eyes of others). In a related finding, reports of introjected regulation were not reliably related to flourishing or teacher well-being, which was contrary to my expectations of a weak, but positive correlation between these constructs (Hypothesis 6). One possible explanation for this is that perhaps introjected regulation for teaching has a differential relationship with well-being that depends on whether the goals of introjected regulation are met. For example, if a teacher works with the aim of avoiding failure or shame, perhaps he or she experiences a sense of well-being when this goal is met, but not otherwise.  Taken together, these findings raise concerns about how teachers come to adopt introjected regulation for teaching, and perhaps more importantly, how we can help teachers who are acting with introjected regulation for teaching move into more autonomous forms of motivation. Research that examines the characteristics of teachers who teach with introjected regulation and investigates whether there are other needs that drive their motivation (e.g., perfectionism, fear of failure) will help to advance our knowledge of this type of regulation. At the same time, introjected regulation represents the ?first step? in the process of 125   internalisation of behaviour (Gillison, Osborn, Standage, & Skevington, 2009). That is, it involves some degree of internalisation?although this relates to internal pressure and is not self-determined, but rather self-controlled (Pelletier, Fortier, Vallerand, & Bri?re, 2001). Nonetheless, it is not fully extrinsically motivated and so may act as a first step towards greater internalisation (and more self-determined motivation) suggesting the need for efforts to focus on how self-determination can be promoted among these teachers. Clearly, more research is needed to provide a greater understanding of this type of regulation.  3.4.2. Job Satisfaction Job satisfaction was the first of two variables in the model that move beyond the theoretical framework of SDT. It was hypothesised that well-being and autonomous motivation (i.e., identified regulation) would positively predict job satisfaction, whereas controlled motivation (i.e., introjected and external regulation) would negatively predict job satisfaction. The findings supported two of the hypotheses: teachers? reports of flourishing, teacher well-being, and identified regulation were positive predictors of job satisfaction (Hypotheses 7 and 9). However, the third hypothesis involving controlled motivation and job satisfaction (Hypothesis 10) was not supported. Contrary to expectations, experiences of external regulation and introjected regulation did not predict job satisfaction.  One possible reason for this non-significance is that job satisfaction may be a secondary concern for teachers who teach with introjected or external regulation. That is, the goals associated with introjected regulation (e.g., avoiding failure) and external regulation (e.g., obtaining a reward) may override a focus on job satisfaction. For example, if a teacher is motivated to teach for the paycheque (i.e., external regulation), it may not be important whether or not the job is satisfying in relation to the work tasks. This finding and the finding 126   involving identified regulation are important as they provide insight into the differential relationships that occur between the types of regulation and job satisfaction.  This result is important for practice. It suggests that school administrators, policy-makers, and even educators themselves may be wise to look at work motivation alongside job satisfaction. Efforts that focus solely on understanding and improving teacher job satisfaction may fail to address the underlying issue of work motivation. If teachers? underlying motivation is not considered simultaneously, these efforts may be ineffective among teachers with controlled work motivation. Instead, by engaging in efforts to increase teachers? autonomous motivation through steps outlined in the model (i.e., through principals? autonomy support and teachers? need satisfaction), schools will be in a better position to address teacher job satisfaction while also improving teacher well-being, motivation, and affective organisational commitment.  Building on this, the finding also has implications for research. Namely, it suggests that rather than focusing on job satisfaction by itself, researchers may gain a more complete picture by considering it alongside work motivation. Another important area of future research is the use of profile or cluster analysis techniques to examine different groups of teachers and how their experiences are explained by the explanatory model. This would advance our understanding of how teachers come to teach with the different types of regulation, how regulation influences other experiences like job satisfaction, and ultimately, how we can help them become more self-determined.  127   3.4.3. Affective Organisational Commitment Affective organisational commitment was the second variable in the model that goes beyond the premises of SDT. As hypothesised, experiences of flourishing, teacher well-being, and identified regulation (autonomous motivation) had positive relationships with affective organisational commitment (Hypotheses 8 and 11). In addition, job satisfaction and affective organisational commitment were positively correlated as expected (Hypothesis 13). The hypothesised relationship involving controlled motivation, however, was not fully supported (Hypothesis 12). Consistent with expectations, external regulation was negatively associated with affective organisational commitment. However, introjected regulation was positively associated with affective organisational commitment.  For the first part of Hypothesis 12, the negative relationship between external regulation and affective organisational commitment has important implications for school administrators, policy-makers, students, and researchers. Given the positive outcomes associated with affective organisational commitment (e.g., teachers? sense of efficacy; Bogler & Somech, 2004) and organisational commitment more broadly defined (e.g., work engagement; Hakanen et al., 2006), it is important that teachers are given the best opportunity to form affective organisational commitment. The importance of this is made even more evident when one considers the fact that organisational commitment is negatively associated with turnover and attrition among teachers (Day, Elliot, & Kington, 2005)?two issues that have severe financial and educational implications for students, schools, and societies (Collie et al., 2011; Ronfeldt, Loeb, & Wyckoff, 2013). The model described here provides school administrators and policy-makers with methods through which to support positive experiences of affective organisational commitment among teachers (e.g., by 128   supporting teachers? need satisfaction). Future research should examine the relationship between motivation and affective organisational commitment in greater detail, along with other forms of organisational commitment such as commitment driven by perceived costs of quitting a job (i.e., continuance commitment; Meyer & Herscovitch, 2001) to see how they relate to motivation. For the second part of Hypothesis 12, introjected regulation was positively associated with affective organisational commitment. Although this was unexpected, a plausible explanation for why it occurred relates to the emotional dimensions of both introjected regulation and affective organisational commitment. By looking at the items used to assess introjected regulation, we can gain a better understanding of this. Specifically, if individuals are teaching ?because [their] work is [their] life and [they don?t] want to fail? or ?because [their] reputation depends on it? (Gagn? et al., 2010), they are driven by ?internal pressure? to avoid shame/guilt or gain contingent self-worth (Deci & Ryan, 1985, 2002). It is conceivable, therefore, that they could report affective organisational commitment because it is their job that provides an opportunity to meet these two different types of internal pressure (e.g., if individuals teach because their reputation depends on it, they may feel emotionally attached to their school given that it helps to ensure their reputation is kept intact). This finding has important implications for researchers. Although it links introjected regulation to a positive outcome (i.e., higher affective organisational commitment), earlier findings showed that introjected regulation was not associated with job satisfaction or well-being, nor was it predicted by need satisfaction. Taken together, these findings underscore the need for research that examines introjected regulation among teachers more closely to understand what types of teachers work with this regulation, along with its causes and correlates. 129   3.5. Limitations and Future Directions Study 2 has several limitations that must be discussed. Given that the same methods of data collection were used as those described in Study 1, the limitations discussed in Study 1 are also relevant for Study 2. In particular, response rates were not able to be calculated during data collection raising questions about the representativeness of the data; teachers were involved in industrial action at the time of data collection, which may have affected their responses; it was assumed that participants? interpretations of questions matched the researchers? raising questions about the accuracy of responses; and the study involved a single self-report survey meaning the results are threatened by single source bias.  As mentioned in Study 1, efforts were taken to temper these limitations (e.g., comparing data to other studies as an indication of representativeness, conducting factor analysis, etc.)3. Notwithstanding this, a necessary development in future research involves finding more robust ways to address these limitations, including finding ways to collect response rates in online data collection (e.g., through personalised email invitations or face-to-face recruiting), conducting ?think-alouds? with teachers as they complete questionnaires to ascertain their thought processes and understand how they interpret the questions, and using multiple data collection methods to reduce issues of single source bias (e.g., conducting interviews or focus groups with teachers).  In addition, there were two further limitations unique to this study. First, I was not able to include intrinsic regulation in the model due to high correlation with job satisfaction. Consequently, the results here do not extend to intrinsic regulation. Combined with the 3 One exception to this is that social desirability was not controlled for in Study 2. This decision was based on the fact that social desirability had only a small influence on the results of Study 1 (explaining variance less than 2%) and because there were concerns with the reliability of the social desirability data.  130                                                    psychometric issues associated with measuring integrated regulation, this raises questions about how we can come to understand the most self-determined types of regulation among teachers (i.e., integrated and intrinsic). Moving forward, there is a need to create more sensitive measures that differentiate or consider the overlap between job satisfaction and the inherent interest of intrinsic regulation among teachers. Further attention should also be placed on integrated regulation to ascertain whether there are other methods for assessing this type of regulation (e.g., observation, interviews). This is important for further testing of the model, as well as for developing and improving our measurement of these constructs in general.  The second limitation is that I was not able to test causal relationships. SEM is used to test a hypothesised explanatory model using sample data (Kline, 2011). It cannot provide proof of causality, but rather indicates whether the model is consistent with the data. In the current study, the findings revealed that the model was consistent with the data suggesting the plausibility of the directional relationships that were decided a priori. However, there is a need to test these relationships directly using longitudinal and/or experimental designs to provide support for the causality and to accept or rule out other plausible directions. A likely first step for completing this type of examination would be to investigate smaller sections of the explanatory model over time. For example, cross-lagged analysis involving a path that does not have existing longitudinal support (e.g., the relationship between well-being and affective organisational commitment) would be a fruitful first step for examining causality in the model. 131   3.6. Conclusions In summary, Study 2 has involved elaborating and testing an explanatory model of teacher motivation, well-being, job satisfaction, and affective organisational commitment based in SDT (Deci & Ryan, 1985, 2002). Models such as this are an important step forward in the literature on teachers as they help to explain why teachers come to experience certain constructs, how efforts can be designed to promote positive work experiences among teachers, and ultimately, how we can help teachers to create the most effective learning environments and experiences for their students. Through the use of factor analysis and SEM, the findings revealed that the core relationships established through SDT were generally supported: teachers? reports of principal?s autonomy support predicted need satisfaction, which, in turn, predicted well-being and motivation. These findings corroborate the existing literature in several ways, while also extending it in relation to the examination of well-being as flourishing and the different types regulation.  Also important were the findings that showed significant relationships involving motivation, well-being, job satisfaction, and affective organisational commitment. In particular, experiences of autonomous motivation (measured as identified regulation) and well-being were positively associated with job satisfaction and affective organisational commitment. There were some unexpected relationships, however, involving autonomy and controlled motivation (measured as introjected regulation and external regulation), as well as between controlled motivation, job satisfaction, and affective organisational commitment. These findings extend our understanding of how teachers? perceptions of motivation, well-being, job satisfaction, and affective organisational commitment are interrelated and 132   highlight the need for more research in this area that examines these outcomes simultaneously. Another area of research that requires greater attention is examination of teachers? experiences of constructs like well-being and motivation over time. This is an important future step for developing our understanding of the explanatory model tested here and will provide evidence for understanding whether and how these variables have causal influence on one another (e.g., whether well-being does cause affective organisational commitment). Another advantage of research over multiple time points is that it can reveal understanding about how these variables change over time and whether there are trends in, say, teacher well-being and motivation over the course of a school term, year, or career in teaching. This type of examination can provide much-needed understanding of how these constructs develop among early career teachers and how these constructs continue to develop across the career span. Furthermore, they are a needed addition in the literature to support a comprehensive understanding of teacher well-being and motivation. The final study in this dissertation, Study 3, aims to begin this process by conducting an examination over time. That is, it involved examining teachers? experiences of well-being and motivation over multiple time points to ascertain whether and how they changed over several months and whether there were discernible trends in these constructs over time.     133   CHAPTER 4. STUDY 3: TEACHER WELL-BEING AND MOTIVATION OVER TIME 4.1. Literature Review The importance of teacher well-being and motivation for a whole host of positive outcomes relating to teachers, students, and schools has been established in previous research (e.g., Duckworth et al., 2009; Pakarinen et al., 2010; Retelsdorf et al., 2010) and further supported in Studies 1 and 2 of this dissertation. In the past, research in this area has largely involved cross-sectional examinations. Although these have advanced our understanding in many ways, there is a need for examinations that are conducted over time. After all, there is knowledge that cannot be ascertained by cross-sectional research such as understanding of the causal ordering between constructs and understanding of the development and stability of constructs over time. Given our broadening understanding of teacher well-being and motivation through cross-sectional research, a necessary next step in the development of the literature is to examine these constructs over time. Indeed, this is a call that has been made frequently in the literature on teachers (e.g., Brackett et al., 2010; Chan, 1998; Klassen et al., 2010, 2011; Klusmann, et al., 2008; Skaalvik & Skaalvik, 2009) and in the more general well-being literature (e.g., Diener, 2009). Only a handful of studies, however, have explored teacher outcomes over time (e.g., Brouwers & Tomic, 2000; Burke & Greenglass, 1995). Clearly, more research is needed to better understand how teacher outcomes change over time. This is necessary for advancing our understanding of the causes of these constructs, along with a need to understand whether and how they develop and change over time. This second point is especially important given that teachers? experiences vary regularly throughout a school year 134   including changes in relationships with students, colleagues, and parents, changes in requirements made by administrators and the government, and changes in students? social, emotional, and academic needs. The aim of Study 3, therefore, was to examine whether and how teachers? experiences of work-related well-being and motivation change over time among a new sample of teachers. As part of the investigation, I also examined whether differences in well-being and motivation among participants were predicted by their perceived satisfaction of the basic psychological needs for relatedness, competence, and autonomy. The rationale for examining this was to provide further support for the relationship between need satisfaction and well-being as examined in Study 2, and to extend our understanding of self-efficacy for teaching by analysing its relationship with need satisfaction.  4.1.1. Teachers? Experiences Over Time The examination of teacher outcomes over time has been recognised as a great need in the literature (e.g., Klassen, Tze, Betts, & Gordon, 2011). In Study 3, teachers? experiences of work-related well-being and motivation were examined over time. As highlighted in Studies 1 and 2, the definition of well-being used in this dissertation is flourishing (e.g., Ryan & Deci, 2011). For teacher well-being, this refers to flourishing at work. For the examination of motivation, the construct of self-efficacy for teaching was used in Study 3. This construct stems from social cognitive theory (Bandura, 1982, 1997) and involves a forward looking belief about one?s capabilities in specific contexts (Klassen et al., 2011). For teachers, it refers to perceptions of their ?capabilities to bring about desired outcomes of student engagement and learning, even among those students who may be difficult or unmotivated? (Tschannen-Moran & Woolfolk Hoy, 2001, p. 783). Self-efficacy 135   for teaching is considered a key motivational belief that influences teachers? actions. Like the motivational constructs examined in Study 2 (i.e., self-determination), self-efficacy for teaching has been associated with important outcomes for students, teachers, and schools (e.g., students? achievement, Caprara et al., 2006; teacher work engagement, Schaufeli & Bakker, 2004; commitment to the profession, Canrinus et al., 2012; mastery goals for teaching; Cho & Shim, 2013), thus justifying its inclusion in Study 3. In order to inform our understanding of how teacher well-being and self-efficacy function over time, I turn to the literature in these respective areas. 4.1.1.1. Changes in well-being over time. Although educational psychology research has focused on well-being related constructs over time (e.g., stress), there is a paucity of research looking at teacher well-being as its own construct over time. Accordingly, we have to turn to research on stress and burnout to help inform our understanding of how teacher well-being might function over time. Looking at changes in stress and mental health over time, Manthei, Gilmore, Tuck, and Adair (1996) collected data among New Zealand teachers at five waves over four years (at the beginning and end of the school year in two consecutive years, and again two years later). Using instruments that assessed sources of stress, overall stress, and experiences of symptoms related to mental health (e.g., crying), ANOVA revealed only two significant differences between consecutive time points. Two of the sources of stress?poor remuneration and community antagonism (e.g., public criticism of teachers)?rose significantly from Time 1 to Time 2. Although this method of comparing significant change between consecutive time points reveals whether there were significant increases or decreases in teacher stress and mental health at the group level, it does not allow us to determine if there were trends over time at the individual level. 136   Examination of trends would allow us to understand the continuing impact that the school system has on teachers? experiences.  In order to examine trends, multilevel modelling is needed. Multilevel modelling is a data analysis technique that allows researchers to estimate individual trajectories of change in a certain outcome variable along with the impact of predictor variables on changes to the starting point and growth of those trajectories (Laursen & Hoff, 2006). Multilevel modelling provides a more robust framework than repeated measures ANOVA or t-tests for examining changes over time as it takes into account the hierarchical nature of data. This is necessary to ensure interpretational and statistical errors do not occur. In addition, addressing the nested nature of data prevents the loss of power that can occur when hierarchical data are not treated as such (Tabachnick & Fidell, 2007). Another strength of multilevel modelling is that it does not require strict adherence to the assumption of independence of errors (Tabachnick & Fidell, 2007) and is more flexible in terms of missing data, non-normality, and other data complexities (Curran, Obeidat, & Losardo, 2010).  A notable example using multilevel modelling is Pas, Bradshaw, and Hershfeldt?s (2012) study that examined changes in self-reported burnout and self-efficacy for handling students with behavioural problems across two school years at three time points (six and twelve months apart; findings involving self-efficacy are discussed in the next section). The sample included practicing teachers in the US who were at various stages in their teaching careers. Using emotional exhaustion?one of Maslach and Jackson?s (1981) three components of burnout?they found a significant increase in teachers? reports of burnout over time. They also found that several other variables negatively predicted changes in both the intercept and growth of burnout (e.g., teacher preparedness for handling classroom 137   management; positive perceptions of school leadership). Taken together, the findings of this study inform our understanding of burnout by suggesting that it tends to increase across school years. A question that remains is whether it can change over shorter time frames.  In a recent study that examined this, Roeser et al. (2013) investigated the effects of mindfulness training on teachers from Canada and the US. The training involved 8-weeks of teaching activities on mindfulness and self-compassion (e.g., guided mindfulness and yoga). Measures of self-reported burnout and stress were taken at the beginning of the program (Time 1), immediately after the program finished (Time 2), and three months later (Time 3). After controlling for baseline levels, the findings revealed that teachers who completed the training reported significantly lower levels of stress and burnout than a control group at Time 2 and Time 3. The authors also collected data on physiological measures of stress (e.g., cortisol, blood pressure); however, there were no significant differences in these measures between the two groups. Combined, the findings of Roeser et al.?s study advance our understanding of the ability of stress and burnout to be impacted by interventions and to change over relatively short periods of time (eight weeks). A follow-up question that needs to be examined is whether these constructs can change over similarly short time frames without teachers? being engaged in regular exercises to promote their well-being. In other words, can we expect teacher stress and burnout to remain stable or increase/decrease in the face of normal teaching activities and no regular mindfulness exercises? This is important for understanding the stability of teacher well-being and whether we can realistically expect any immediate changes as a result of efforts by schools to promote well-being among teachers and students. 138   Broadly, Roeser et al.?s (2013) findings along with those from Manthei et al. (1996) and Pas et al. (2012) inform our understanding of how well-being related constructs function over time. More attention needs to be devoted to teacher well-being as its own construct, however, given that it is more than the absence (or opposite) of stress or burnout (Huppert & So, 2013). In order to provide some understanding of how teacher well-being may function over time, it is necessary to turn to the wider psychological literature involving subjective well-being?the relative newness of the flourishing construct means that longitudinal examinations are not yet available. This literature (e.g., Lyubomirsky, Dickerhoof, Boehm, & Sheldon, 2011; Schimmack & Lucas, 2010; Suh, Diener, & Fujita, 1996) indicates that subjective well-being remains relatively constant over time and returns to baseline levels rather quickly after changes in life circumstances?unless individuals are regularly engaged in activities to boost their well-being (e.g., mindfulness, Roeser et al., 2013). It is possible that a similar phenomenon occurs with flourishing; however, the differences between these two constructs and the differences between global measures of well-being (like subjective well-being) and teacher well-being mean that research is needed to determine if teacher well-being is generally stable over time or whether it can change, as well as whether any changes can be expected over short or longer time frames.  Taken together, previous research indicates that stress and burnout among teachers can change over time, whereas subjective well-being among individuals does not tend to vary much over time. Accordingly, these findings underscore the need for the examination of teacher well-being directly. In addition to examining that, Study 3 also examined the impact of need satisfaction on teacher well-being over time. This provides a further opportunity to examine the core relationship of self-determination theory (SDT; Deci & Ryan, 1985, 2002) 139   that was examined in Study 2. Furthermore, it will provide an understanding of the impact of contextually-related constructs on teacher well-being over time. 4.1.1.2. Changes in self-efficacy over time. The literature on self-efficacy for teaching is quite vast, and fortunately, has involved several studies examining the construct over time. Looking at changes in efficacy over approximately two years, Woolfolk Hoy and Spero?s (2005) conducted a study among pre-service teachers and continued to collect data from these teachers as they completed their first year of teaching in the profession. They examined self-efficacy for teaching using two previously published measures (Bandura, 1997; Gibson & Dembo, 1984) and one they created for their study. The participants were asked to complete these measures at the beginning (Time 1) and end (Time 2) of their teacher education program, and a year after they began teaching (Time 3). The researchers conducted repeated measures ANOVA to examine whether there were significant differences in self-efficacy between the different time points. For all measures, self-efficacy increased significantly from Time 1 to Time 2. The findings from Time 2 to Time 3 were mixed with some measures showing a significant decrease in self-efficacy for teachers and others showing no change.  The findings of Woolfolk Hoy and Spero?s (2005) research provide a revealing insight into the development of self-efficacy at the very early career stage. At the same time, concerns over the conceptual accuracy of one of the measures used?the Gibson and Dembo (1984) instrument?point to a larger issue in the literature. Woolfolk Hoy and Spero, along with Tschannen-Moran, Woolfolk Hoy, and Hoy (1998), and more recently Klassen et al. (2011) have cautioned that the Gibson and Dembo instrument focuses on locus of control rather than self-efficacy. Consequently, they question the accuracy of interpretations about 140   self-efficacy from the measure. This highlights the need for research using what Klassen et al. (2011) argue are conceptually clearer measures. Specifically, Klassen et al. (2011) have proposed that in order to be congruent with self-efficacy theory, instruments must attend to judgments related to specific outcomes (e.g., engaging students) and focus on forward-looking capabilities. Tschannen-Moran and Woolfolk Hoy?s (2001) Teacher Sense of Efficacy Scale (TSES) is one such measure that enables us to examine three factors of teacher self-efficacy?self-efficacy for student engagement, classroom management, and instructional strategies?using a conceptually and theoretically sound construction.  Notwithstanding the conceptual issues with one of the scales, the Woolfolk Hoy and Spero (2005) study provides important understanding of the malleability of self-efficacy at the pre-service level and indicates that it tends to reduce or remain constant in the first year of teaching. A question that remains is how self-efficacy functions over time among more experienced teachers. In particular, there is a need for research to determine whether it becomes relatively stable once teachers gain more experience or whether it continues to be flexible and/or malleable. One study that examined changes in self-efficacy for teaching among more experienced teachers is Carleton, Fitch, and Krockover?s (2012) research examining the impact of a two-week science professional development program on self-efficacy for science teaching. The authors examined self-efficacy at four waves: before and after the two-week summer workshop on science concepts, once in autumn when teachers returned for a one day workshop, and in spring when teachers gave a presentation at a science teachers? association meeting. Using paired t-tests, they found that participants showed a significant increase in self-efficacy for science teaching from before to after the 141   workshop. In addition, the autumn and spring waves were also significantly higher than the original measurement.  Taken together, the findings from Carleton and colleagues? (2008) study indicate that self-efficacy for teaching can change among teachers at varying career stages as a result of professional development. These changes also occurred over a short time frame (i.e., two weeks). Once again, however, concerns with the robustness of the instrument they used necessitate caution in interpreting the results. The items Carleton and colleagues (2008) used to assess self-efficacy for science teaching were based on an instrument (i.e., Friedman and Kass, 2002), which Klassen et al. (2011) argue drifts from ?theory-based understanding? of self-efficacy. For example, Carleton and colleagues used the item, ?I see myself as an interesting and motivating science and/or math teacher? (Carleton et al., 2008, p. 52), which appears to focus on something akin to self-concept as a science teacher rather than self-efficacy. Although the authors need to be commended on their efforts to inform our understanding of changes in self-efficacy as a consequence of professional development, the measurement issues highlight the need for more research in this area. Given that teachers are regularly involved in various short-term initiatives/professional development programs such as that examined by Carleton and colleagues, research that endeavours to understand whether changes in self-efficacy over short periods of time are feasible will advance our understanding of the realistic impact of these programs on self-efficacy for teaching.  Other longitudinal studies are beset by similar measurement issues (e.g., Ross, 1994). In fact, Klassen et al. (2011) claim that the research on self-efficacy for teaching has been plagued by measurement concerns. Another methodological issue that deserves attention is the data analytic techniques used to assess self-efficacy over time. Like well-being, more 142   examination of self-efficacy over time using robust multilevel modelling techniques is needed. A notable example that addresses both of these concerns is Pas and colleagues? (2012) study, which was introduced earlier. In addition to examining changes in burnout over time, they also examined changes in self-efficacy for handling students with behavioural problems. As noted, their sample included practicing teachers at various stages in their teaching careers and the authors found a significant increase in self-efficacy over time among the teachers. They also found that several other variables positively predicted changes in the intercept (e.g., positive perceptions of school leadership) and/or growth (i.e., teacher preparedness for handling classroom management) of self-efficacy. This study informs our understanding of how self-efficacy for managing disruptive behaviour can change over time and also suggests that changes do not appear to be limited to teachers who are at the beginning of their career. Research examining other types of self-efficacy (e.g., for engaging students) will further develop our understanding of this construct. Another study supporting the flexibility of self-efficacy for teaching across the career is Klassen and Chiu?s (2010) study examining the relationship between self-efficacy for teaching and years of experience. The authors used the conceptually sound TSES (Tschannen-Moran & Woolfolk Hoy, 2001) to examine three factors of self-efficacy? self-efficacy for student engagement, classroom management, and instructional strategies?among a large sample of Canadian teachers. The study was cross-sectional, but the authors examined relationships between self-efficacy and years of experience to understand how self-efficacy was different at various career stages. The findings revealed that the three factors of self-efficacy increased along with teaching experience up to mid-career, but that they decreased as teachers moved into late career. This research provides support for the idea 143   that self-efficacy for teaching can continue to change as teachers gain experience. Moving forward, research needs to examine these three types of self-efficacy using data collected at multiple time points.  Taken together, previous research highlights the need for greater examination of self-efficacy for teaching over time using robust measures and data analytic techniques. It also provides evidence that self-efficacy is flexible over both short and long time periods. Given the measurement concerns with some previous studies (e.g., Carleton et al., 2008), however, research using more conceptually sound measures is needed to confirm whether changes in self-efficacy are realistic over shorter time periods. As noted above, this will help us understand whether short-term interventions that are commonly applied in schools are likely to have an immediate effect or whether slower, more gradual changes in self-efficacy can be expected. To that end, Study 3 examined changes in self-efficacy for three factors of efficacy during a relatively short time period of two to three months. Like well-being, it also investigated the influence of need satisfaction on changes in self-efficacy to better understand how contextually-related experiences affect teachers. 4.1.2. Overview of the Current Study Cross-sectional research provides us with an understanding of constructs at one point in time, but it does not tell us how teachers? experiences change over time. For Study 3, I examined how teacher well-being and self-efficacy for teaching function over a relatively short time frame of two to three months. One particular commonality about the participants? experiences is that they all participated in a professional development program on anti-oppression. In the interest of transparency, it is noted that the program?s focus on anti-oppression may have affected teachers? perceptions of well-being and self-efficacy for 144   teaching. However, given that professional development is a typical experience for teachers in British Columbia every two to three months (Ministry of Education, 2012), any influence was treated as a feature of the environment within which the participants worked. In addition, anti-oppression is a topic with which teachers in British Columbia are familiar given the province?s performance standards on social responsibility (Ministry of Education, 2001). The performance standards, created for voluntary use in schools, refer to students? skills and awareness in contributing to the classroom and school community, solving problems in peaceful ways, valuing diversity and defending human rights, and exercising democratic rights and responsibilities (Ministry of Education, 2001). Taken together, professional development on topics like anti-oppression is a feature of the schooling system within which the participants worked and so was treated as such in Study 3.  Taking this setting into account, Study 3 involved examining changes in teacher well-being and self-efficacy for teaching over time. Moreover, the study investigated how basic psychological need satisfaction influenced these outcomes over time. Two research questions guided this study:  ? How do teacher well-being and self-efficacy for teaching function over time? (Research Question 1) ? How does need satisfaction affect teachers? initial experiences of well-being and motivation? (Research Question 2) For Research Question 1, I hypothesised that well-being would show little change over time. This is based on research showing that subjective well-being is generally stable over time and returns to baseline levels relatively quickly after changes in life circumstances (Lyubomirsky et al., 2011; Schimmack & Lucas, 2010; Suh et al., 1996; Hypothesis 1a).  145   ? Hypothesis 1a. Well-being will remain stable over the two- to three-month period of examination. For self-efficacy, however, I hypothesised that there may be some change given that previous research has provided evidence of the variability of teacher self-efficacy over relatively short amounts of time (e.g., three, six, twelve months; Carleton et al., 2009; Klassen & Chiu, 2010; Woolfolk Hoy & Spero, 2005; however, as noted earlier there are concerns about the conceptual accuracy of Carleton and colleagues? findings).  ? Hypothesis 1b. Self-efficacy will show variation over the two- to three-month period.  For Research Question 2, the hypothesised relationships between need satisfaction and well-being are based on relationships postulated in Study 1. ? Hypothesis 2a. Satisfaction of the needs for autonomy and competence will positively predict workload well-being.  ? Hypothesis 2b. Satisfaction of the needs for autonomy, competence, and relatedness with colleagues will positively predict organisational well-being. ? Hypothesis 2c. Satisfaction of the needs for autonomy, competence, and relatedness with colleagues will positively predict student interaction well-being. The relationships between need satisfaction and self-efficacy are drawn from understanding of both social cognitive theory (Bandura, 1982, 1997) and SDT (Deci & Ryan, 1985, 2002). Although motivation is defined differently in these two theories, I suggest that the relationship between need satisfaction and motivation in SDT is also relevant when defining motivation according to social cognitive theory (i.e., as self-efficacy). In essence, I posit that satisfaction of psychological needs is positively associated with confidence in future abilities (i.e., self-efficacy). Previous research supports this reasoning by highlighting 146   that self-efficacy is related to the need for competence (e.g., Lynch et al., 2005). Further, self-efficacy relates to agency (i.e., the belief that we shape our own life circumstances; Bandura, 1989), which is similar to autonomy. Together, these findings warrant further examination of the relationship between need satisfaction and self-efficacy. Rationales for the specific hypotheses between need satisfaction and self-efficacy were based on the meaning of the constructs. First, I hypothesised that competence and relatedness with students would be significant predictors of self-efficacy for student engagement because teachers are more likely to feel confident in their ability to engage their students if they have a connection with their students (i.e., because they better understand their students? learning needs) and if they feel competent in their teaching ability (i.e., confidence reflects a sense of competence).  ? Hypothesis 2d. Satisfaction of the needs for competence and relatedness with students will positively predict self-efficacy for student engagement.  Second, I hypothesised that autonomy, competence, and relatedness with students would be significant predictors of self-efficacy for classroom management given that teachers who feel connected to their students, competent, and in control are more likely to understand the best ways to successfully manage their students, feel more confident about their ability to manage the classroom, and choose classroom management strategies that work for them.  ? Hypothesis 2e. Satisfaction of the needs for autonomy, competence, and relatedness with students will positively predict self-efficacy for classroom management. For the final outcome variable, it was hypothesised that competence and autonomy would predict self-efficacy for instructional strategies because a sense of competence likely 147   relates to greater confidence in one?s abilities and because autonomy likely means that teachers are free to choose instructional strategies that they feel confident in using. ? Hypothesis 2f. Satisfaction of the needs for autonomy and competence will positively predict self-efficacy for instructional strategies. 4.2. Methods 4.2.1. Participants Participants included 71 teachers from a large, suburban school board in British Columbia, Canada. These teachers represent a different sample from the first two studies. Moreover, they worked in a different school district that was not involved in Studies 1 and 2 data collection. Approximately 69% of the sample was female and 49% taught at the elementary level (37% at the secondary level, 10% at both levels, and 1% taught at an alternative education centre). The average teaching experience was 11.6 (SD = 8.14) years and the average age of participants was 40.4 (SD = 10.28) years. Participants were asked about their ethnic origins. Among the sample, 75% reported having one, 18% reported having two, and 1% reported having three ethnic backgrounds (6% chose not to answer). The frequency of the reported ethnic backgrounds are shown in Table 4.1. Participants were asked whether they considered themselves to be a minority based on race, gender, ethnicity, class, ability and/or sexual orientation. Thirty-nine percent (39%) of participants considered themselves to be a minority, 12% were unsure, and the remainder did not consider themselves to be a minority (45%). The majority of participants were classroom teachers (58%). Other positions included working as support teachers (e.g., resource teachers, special education teachers, or counsellors, 14%), youth and family support workers (13%), multicultural liaison worker 148   (7%), administrators (e.g., principals; 3%), teacher librarians (1%), and other positions (9%; e.g., settlement workers in schools, district support staff, special education assistant). All participants worked with students and so were classified as teachers in the results.   Table 4.1 Frequency of Participants? Ethnic Origins  Northern and western European Eastern and southern European East Asian South Asian Aboriginal Other  Frequency 62% 14% 14% 11% 4% 10% Note. Northern and Western European origins refers to British, Scottish, German, Swedish, Danish, Norwegian, Dutch, etc. Eastern and Southern European origins refers to Polish, Russian, Ukrainian, Italian, Greek, Spanish, etc. Aboriginal origins refers to First Nations, Inuit, Metis, etc. East Asian origins refers to Chinese, Japanese, Korean, etc. South Asian origins refers to East Indian, Punjabi, Pakistani, etc. Other included southeast Asian, self-identified Canadian, Latin American, self-identified Jewish, and unspecified.  4.2.2. Procedures Participants for Study 3 were recruited through the booking process of a professional development program that aimed to promote anti-oppression in classrooms and schools (Appendix C shows the invitation letter provided to teachers). When teachers booked the program for their classroom or participated in it at district-level professional development days, they were invited to participate in the research. Most of the data collection occurred via online questionnaire; however, teachers who participated in the professional development days completed a paper copy of the questionnaire at Time 1 and online copies at Time 2 and 3. Data were collected over two school years. Table 4.2 shows the sample size for the three waves of data collection broken down by potential sample size (i.e., participants who consented to the research) and actual sample size (i.e., participants who actually filled out the 149   questionnaires).4 It also shows the response rates for the three time points. Details about attrition are discussed below. Teachers completed the questionnaire at three time points. Time 2 occurred around six weeks after Time 1 (M = 40.14 days, SD = 20.84 days; time varied depending on teacher availability and school vacations) and Time 3 occurred around nine weeks after Time 1 (M = 64.31 days, SD = 27.57 days; this means that Time 3 occurred about four weeks after Time 2). Data were collected over a two- to three-month period.   Table 4.2 Actual and Potential Sample Size at Each Time Point  Time 1 Time 2 Time 3 Potential 71 50a 50 Actual 65 28 22 Response rate 92% 56% 44% Note. Pro-d = professional development.  a The drop in potential sample size between Time 1 and 2 was because 21 teachers in Year 2 were not included in data collection at Time 2 and 3. See Footnote 4.   4.2.3. Measures As noted above, three self-report questionnaires were used to collect data at three times points. The questionnaires included measures of well-being, need satisfaction, self-efficacy, and questions pertaining to demographics. Table 4.3 shows the items that were used for each construct. The choice of items was based on factor analyses conducted in previous 4 The drop in potential sample size between Time 1 and Time 2 reflects a change in school district protocol which meant we were unable to collect email addresses of teachers who only participated in the program at the professional development days (i.e., they did not make a subsequent booking to have the program in their classroom).  150                                                    research among different samples from the same British Columbian population of teachers (e.g., Study 1 and 2; Collie et al., 2012). Table 4.4 shows the descriptive statistics for the different constructs and Table 4.5 shows the Cronbach?s alphas. As described in Studies 1 and 2, Table 2.5 shows the constructs that were examined in Study 3, along with those examined in the previous two studies.  4.2.3.1. Well-being. Well-being was measured with the Teacher Well-Being Scale (TWBS) developed in Study 1 (see Appendix D, Section 2). As established in Study 1 and shown in Table 4.3, six items were used to measure workload well-being (e.g., ?Staying late after work for meetings and activities?), six items were used to measure organisational well-being (e.g., ?Support offered by school leadership?), and four items were used to measure student interaction well-being (e.g., ?Relations with students in my class?). All items were scored on a 7-point scale ranging from Negatively (1) to Positively (7). The Cronbach?s alphas at all three time points provide evidence for validity based on internal structure (see Table 4.5).  4.2.3.2. Self-efficacy for teaching. The Teachers? Sense of Efficacy Scale (TSES, Tschannen-Moran & Woolfolk Hoy, 2001) was used to measure self-efficacy for teaching (see Appendix D, Section 4). Based on Collie et al.?s (2012) factor analyses, three items were used to measure self-efficacy for student engagement (e.g., ?How much can you do to get students to believe they can do well in school work??), another three items were used to measure self-efficacy for classroom management (e.g., ?How much can you do to get children to follow classroom rules??), and four items were used to measure self-efficacy for instructional strategies (e.g., ?To what extent can you use a variety of assessment strategies??). Items were scored on a scale ranging from Not At All (1) to A Great Deal (9). 151   Evidence of validity, including reliability, of the scores on this measure have been supported in a variety of samples (e.g., ? ? .64; Klassen et al., 2009; Tschannen-Moran & Woolfolk Hoy, 2001), as well as in Study 3 (see Table 4.5). 4.2.3.3. Need satisfaction. Need satisfaction was measured with the same instrument used in Study 2, the Work-related Basic Need Satisfaction scale (Van den Broeck et al., 2010; see Appendix D, Section 3). Four items were used to measure relatedness with colleagues (e.g., ?At work, I can talk with people about things that really matter to me?), three items were used to measure relatedness with students (e.g., ?I am very committed to my students?), six items were used to measure competence, (e.g., ?I feel competent at my job?), and five items were used to measure autonomy (e.g., ?I feel free to do my job the way I think it should be done?). Participants responded to these statements on a scale ranging from Strongly Disagree (1) to Strongly Agree (7). Once again, Cronbach?s alphas at all three time points were adequate. 4.2.3.4. Demographic information. Teachers were asked to supply demographic information including age, sex, ethnic/cultural heritage, whether or not they considered themselves a member of a minority group, school grades currently teaching, current teaching position, and years of teaching experience (see Appendix D, Section 1).       152    Table 4.3 Items Used to Assess the Well-Being, Self-Efficacy, and Need Satisfaction Constructs   Items Item number Workload well-being  Marking work 3 Fitting everything in to the allotted time 4 Administrative work related to teaching 7 Work I complete outside of school hours for teaching 13 Working to finish my teaching tasks 14 Staying late after work for meetings and activities 16 Organisational well-being  Relations with administrators at my school 2 Support offered by school leadership 9 Recognition for my teaching 10 School rules and procedures that are in place 11 Communication between members of the school 12 Participation in school-level decision making 15 Student interaction well-being  Relations with students in my class 1 Student behaviour 5 Student motivation 6 Classroom management 8   153   Table 4.3 (Continued)  Items Item number Self-Efficacy for Student Engagement  How much can you do to motivate students who show low interest in school work? 2 How much can you do to help your students value learning? 4 How much can you do to get students to believe they can do well in school work? 7 Self-Efficacy for Classroom Management  How much can you do to control disruptive behaviour in the classroom? 1 How much can you do to calm a student who is disruptive or noisy? 3 How much can you do to get children to follow classroom rules? 6 Self-Efficacy for Instructional Strategies  To what extent can you craft good questions for your students? 5 To what extent can you use a variety of assessment strategies? 9 To what extent can you provide an alternative explanation or example when students are confused? 10 How well can you implement alternative teaching strategies in your classroom? 12 Relatedness with colleagues  I don?t really feel connected with other people at work (R) 1 I don?t really mix with colleagues at work (R) 3 At work, I can talk with people about things that really matter to me 4 I often feel alone when I am with my colleagues (R) 5   154   Table 4.3 (Continued)  Items Item number Relatedness with students  I am very committed to my students 19 Connecting with my students is an essential part of the job 20 I feel connected to my students 22 Competence  I don?t really feel competent in my job (R) 7 I really master the things I have to do at work 8 I feel competent at my job 9 I have doubts about whether I am able to do my job properly (R) 10 I am good at the things I do in my job 11 I have the feeling that I can accomplish even the most difficult tasks at work 12 Autonomy  At work, I often feel like I have to follow other people?s commands (R) 14 If I could choose, I would do things at work differently (R) 15 The things I have to do at work are in line with what I really want to do 16 I feel free to do my job the way I think it should be done 17 In my job, I feel forced to do things I don?t want to do (R) 18   155   Table 4.4 Means, Standard Deviations, and Ranges for all Variables  Time 1  Time 2  Time 3  M SD Observed range  M SD Observed range  M SD Observed range Workload well-being 3.92 0.936 1.20 ? 6.17  4.15 1.139 1.50 ? 6.50  4.30 1.149 1.67 ? 7.00 Organisational well-being 5.05 1.070 2.67 ? 6.83  5.01 1.066 3.00 ? 7.00  4.94 1.117 3.00 ? 7.00 Student interaction well-being 5.23 1.028 3.00 ? 6.75  5.36 0.817 2.50 ? 6.75  5.48 0.862 4.00 ? 7.00 Self-efficacy for student engagement 6.77 0.984 4.33 ? 9.00  7.11 1.050 6.00 ? 9.00  6.90 1.028 5.33 ? 9.00 Self-efficacy for classroom management 7.03 1.044 4.33 ? 9.00  7.53 0.944 6.00 ? 9.00  7.46 0.963 5.67 ? 9.00 Self-efficacy for instructional strategies 7.25 1.059 5.00 ? 9.00  7.17 0.969 5.75 ? 9.00  7.13 1.130 4.75 ? 9.00 Relatedness with colleagues 3.83 0.830 1.00 ? 5.00  3.82 0.703 2.00 ? 5.00  3.80 0.892 1.00 ? 5.00 Relatedness with students 4.53 0.443 3.33 ? 5.00  4.49 0.555 3.00 ? 5.00  4.45 0.540 3.00 ? 5.00 Competence 4.10 0.571 2.33 ? 5.00  3.96 0.650 2.17 ? 5.00  3.95 0.467 3.00 ? 3.00 Autonomy  3.32 0.749 1.80 ? 5.00  3.47 0.854 1.40 ? 5.00  3.31 0.722 2.00 ? 5.00   156   Table 4.5 Cronbach?s Alphas for the Different Variables at the Three Time Points  Time 1 Time 2 Time 3 Workload well-being .84 .84 .92 Organisational well-being .85 .95 .94 Student interaction well-being .84 .87 .84 Self-efficacy for student engagement .75 .90 .87 Self-efficacy for classroom management .83 .84 .86 Self-efficacy for instructional strategies .86 .81 .90 Relatedness with colleagues .87 .71 .95 Relatedness with students .71 .90 .85 Competence .87 .92 .82 Autonomy .87 .92 .91      157   4.3. Results Before conducting the main analyses to answer the research questions, preliminary analyses were used to examine the attrition that occurred and what differences existed between participants who remained in the study and those who dropped out. Time 1 data was used to run several independent t-tests to compare these two groups. The results are shown in Table 4.6. The findings revealed no significant differences in well-being and self-efficacy at Time 1 between those who dropped out after Time 1 and those who remained in the study. Although it is possible that this non-significance reflects the small sample size, it does provide some confidence that the groups were similar in relation to their experiences with the outcome variables at Time 1. Unfortunately, this does not allow us to conclude that there were no differences in unmeasured variables and/or differences that occurred after Time 1 (e.g., if participants became very busy and could not fill out the questionnaire, which may have had implications for their motivation and well-being). Further details of this are discussed in the limitations section below. Correlations between all variables at each time point are shown in Table 4.7, Table 4.8, and Table 4.9.   158   Table 4.6 Differences in Time 1 Scores Among Participants Remaining In or Dropping Out of Study After Time 1  Teachers who only participated at Time 1 M (SD) Teachers who remained in study after Time 1  M (SD) t df p Workload well-being 3.87 (0.900) 3.97 (0.988) -0.418 57 .677 Organisational well-being 4.98 (1.076) 5.13 (1.078) -0.523 60 .603 Student interaction well-being 5.09 (1.040) 5.39 (1.007) -1.173 61 .245 Self-efficacy for student engagement 6.68 (1.096) 6.87 (0.864) -0.709 51 .481 Self-efficacy for classroom management 6.88 (1.075) 7.18 (1.008) -1.058 51 .295 Self-efficacy for instructional strategies 7.17 (1.201) 7.33 (0.897) -0.526 51 .601  159   Table 4.7 Correlations Between Variables at Time 1   1. 2. 3. 4. 5. 6. 7. 8. 9. 1. Workload WB          2. Organisational WB .47*         3. Student interaction WB .45* .50*        4. SE for student engagement .26 .38* .59*       5. SE for classroom management .25 .41* .49* .65*      6. SE for instructional strategies -.04 .31* .28* .54* .46*     7. Rel. with teachers .15 .59* .13 .06 .09 .18    8. Rel. with students .04 .20 .26 .31* .40* .42* .15   9. Competence .23 .42* .41* .36* .37* .25 .22 .40*  10. Autonomy .47* .74* .50* .35* .48* .22 .43* .32* .48* Note. WB = well-being. SE = self-efficacy. Rel. = relatedness.  * p < .05.     160   Table 4.8 Correlations Between Variables at Time 2   1. 2. 3. 4. 5. 6. 7. 8. 9. 1. Workload WB                   2. Organisational WB .60*         3. Student interaction WB .68* .44*        4. SE for student engagement .35 .26 .37       5. SE for classroom management .30 .21 .39* .52*      6. SE for instructional strategies .58* .31 .57* .54* .54*     7. Rel. with teachers .47* .66* .42* .04 .17 .02    8. Rel. with students .07 .10 .37 .35 .41* .22 .38*   9. Competence .42* .52* .79* .28 .37 .26 .61* .51*  10. Autonomy .69* .78* .68* .38* .33 .19 .69* .28 .66* Note. WB = well-being. SE = self-efficacy. Rel. = relatedness.  * p < .05.    161   Table 4.9 Correlations Between Variables at Time 3   1. 2. 3. 4. 5. 6. 7. 8. 9. 1. Workload WB                   2. Organisational WB .51*         3. Student interaction WB .56* .42        4. SE for student engagement .40 .51* .61*       5. SE for classroom management .28 .40 .60* .69*      6. SE for instructional strategies .47* .66* .62* .76* .79*     7. Rel. with teachers .26 .59* .46* .08 .43 .41    8. Rel. with students .23 .20 .57* .33 .57* .58* .47*   9. Competence .49* .76* .72* .58* .62* .74* .72* .60*  10. Autonomy .70* .73* .50* .32 .43 .60* .65* .29 .67* Note. WB = well-being. SE = self-efficacy. Rel. = relatedness.  * p < .05.   162   4.3.1. Overview of Data Analysis In order to examine how teacher well-being and motivation function over time, individual trajectories of growth were fitted using univariate multilevel modelling. This involves fitting increasingly complex models to see if the addition of variables provides a better fitting model that explains changes in the outcome variables over time.  4.3.1.1. Unconditional models. The first model that was fitted was an unconditional means model with no predictors (Model 1) to check whether there was significant variation in the outcome variables between participants. This model included a fixed component representing the average of the outcome variable across participants and the three time points (i.e., the grand mean; ?0i). In addition, there were two random components representing variation in the grand mean across individuals (between-level; u0i) and variation across the three time points within individuals (within-level; rti). The following equation represents the unconditional means model: Yti = ?0i + [u0i + rti] If the grand mean was significantly different from zero this revealed that the outcome variable, Yti, differed significantly across individuals. Furthermore, if the variance components for the two random effects were statistically significant this revealed that there was both between-person variance (variance associated with the intercept, ?00) and within-person variance (?2) to be explained. The reasons for fitting the following models was to reduce the between-person and within-person variance by adding factors into the model. By reducing these variance components, it means that the model more accurately explains individual trajectories. 163   The second set of models were unconditional growth models where time was entered as an effect. The first of these, Model 2a, involved entering time as a fixed effect to ascertain whether there was significant change over time in participants? scores on the outcome variable. The following equation represents the unconditional growth model with time as a fixed effect: Yti = [? 0i + ? 1iTimeti] + [u0i + rti] where [? 0i + ? 1iTimeti] represented the fixed effects of intercept and slope, respectively, and [u0i + rti] represents the random effects of intercept and the growth rate (rti).  For the second unconditional growth model, Model 2b, time was allowed to be a random effect (Model 2b). This captured any differences in the slope of the trajectories across participants. The following equation represents the unconditional growth model with time as a random effect: Yti = [? 0i + ? 1iTimeti] + [u0i + u1i Timeti + rti] where the addition of u1i Timeti represents the random effect of slope. There were two additional between-person variance components in this model: the variance associated with the slope (?11) and the covariance between the slope and intercept (?01).  The third unconditional model, Model 2c, involved testing the curvature of the trajectory by including a fixed factor of Time2. The following equation represents this model: Yti = [? 0i + ? 1iTimeti + ? 2iTimeti2] + [u0i + u1i Timeti + rti] where the addition of ? 2iTimeti2 represents the fixed effect of curvature. Following this, the final unconditional model, Model 2d, involved allowing Time2 to vary randomly. This captured any differences in the curvature of the trajectories across participants. The equation for this model is as follows: 164   Yti = [? 0i + ? 1iTimeti + ? 2iTimeti2] + [u0i + u1i Timeti + u1i Timeti2 + rti] where the addition of u1i Timeti2 represents the random effect of curvature.  4.3.1.2. Conditional model. Once the intercept, slope, and curvature were identified through the unconditional models, conditional models were run where variables of interest were included to see if they explained the between-person and within-person variance. In Study 3, the first conditional model, Model 3, involved the addition of the need satisfaction variables: relatedness with colleagues, relatedness with students, competence, and autonomy.5 Namely, participants? scores on the need satisfaction variables at Time 1 were entered as a time-invariant predictor. The aim of this model was to assess the effect of need satisfaction on participants? well-being and self-efficacy. The final model, Model 4, only included the need satisfaction variables that were significant. The rationale for excluding the non-significant need satisfaction variables was for parsimony given that they did not significantly add to the explanation of the model. 4.3.2. Results for Multilevel Findings In order to conduct the multilevel analyses, I used Mplus Version 7.11 (Muth?n & Muth?n, 1998-2012). The method of estimation used was MLR, which involves maximum likelihood estimation with robust standard errors and a robust chi-square test. This method of estimation is robust in the face of non-normality and missing data. In order to assess model fit, a chi-square difference test based on the loglikelihood values was used. This method is appropriate for MLR estimation and incorporates the scaling correction factor of each model 5 Covariates of age, teaching experience, teaching load, perceptions of workshop (i.e., ?perceived usefulness?), number of workshops, sex, and identified minority were included in preliminary conditional models; however, none were significant. Thus, these models were not reported in the results and the covariates were excluded so that they did not compromise power given the small sample size.  165                                                    (Muth?n & Muth?n, 2005). The df value used in this method is the difference in the number of parameters between the two models. In order to aid interpretability, time was centred so that Time 1 = 0, and the need satisfaction variables were centred around the mean at Time 1.  4.3.2.1. Workload well-being. The first outcome examined was workload well-being. Table 4.10 shows the results. As detailed above, the first model tested was the unconditional means model (Model 1, Table 4.10). The results revealed that workload well-being varied significantly among participants (as seen in the significant intercept in Table 4.10, p < .001). The variance estimated for the mean intercept (?00) was 0.803 (p < .001) and that for the within-person random effect (?2) was 0.208 (p < .001). This highlights that there was more variance between participants than within participants. Model 2a, an unconditional growth model, involved the addition of time as a fixed effect to see if it explained the two variance components. It was not significant (p = .088) revealing that there was no significant change in workload well-being over time. Given this, there was no need to run an unconditional model with time as a random effect (Model 2b described above). The curvature of the trajectories was also tested (Model 2c described above) to assess whether there was a significant curvilinear trend that was not captured by the time variable. To that end, time2 was entered in the model, but it was not significant confirming that there was no change over time (results not presented). Despite there being no significant changes in workload well-being over time, conditional models were run to help explain the between-person variance. In order to provide contextual timing of the data, time was left in the model despite being non-significant. The first conditional model involved the addition of the need satisfaction variables (see Model 3, Table 4.10). The results showed that autonomy was a significant predictor of workload well-166   being (p = .002). The final conditional model (Model 4) excluded the non-significant need satisfaction variables and confirmed that autonomy was still a significant predictor (p < .001). As teachers? reports of autonomy increased, so did their workload well-being. This finding highlights that individual differences in workload well-being were explained in part by autonomy. This was mirrored in the between-person variance (?00), which was reduced from 0.789 (p < .001) in Model 2a to 0.568 (p < .001) in Model 3. Put another way, the addition of autonomy into the model explained 28% of the between-person variance. The fact that the between-person variance remained significant suggests that there are other variables that explain differences in workload well-being between participants (e.g., perhaps school climate or autonomy support). The within-person variance (?2) was not greatly reduced by the addition of factors into the model (0.208 [p < .001] in Model 1 to 0.193 in Model 4 [p < .001]), which is not surprising given that time (the within-person variable) was not significant. A chi-square difference test showed that the addition of autonomy significantly improved the fit of model: ?2 (1) = 71.86, p < .001. In order to graph the predicted values of workload well-being, the following equation was used: Yti = ?00 + ?01Timeti + ?05Autonomyi  This equation was solved for three different levels of autonomy: teachers reporting low autonomy (one standard deviation below the mean), average autonomy, and high autonomy (one standard deviation above the mean). Figure 4.1 shows the relationship between workload well-being and autonomy based upon this equation. Time is not considered in the graph given that it was not a significant factor in the model. As the graph shows, teachers 167   who reported high autonomy experienced significantly greater workload well-being than teachers reporting average or low autonomy.  Table 4.10 Multilevel Models for Examining Change in Workload Well-Being Over Time  Model 1 Model 2a Model 3 Model 4 Fixed effects     Intercept (?00) 4.039*** 3.975*** 3.878*** 3.880*** Timeti (?01)  0.127 0.100 0.101 Rel. with colleagues (?02)   -0.103?  Rel. with students (?03)   -0.254  Competence (?04)   0.097  Autonomy (?05)   0.702** 0.637*** Random effects     ?00 0.803*** 0.789*** 0.555*** 0.568*** ?2 0.208*** 0.198*** 0.192*** 0.193*** Fit statistics     -2LL -131.951 -130.198 -97.538 -98.186 AIC 269.902 268.396 211.076 206.373 ?p = .055, *p < .050, ** p < .010, *** p < .001.     168       Figure 4.1. Predicted values of workload well-being for different levels of autonomy.   33.544.555.56Low Average HighWorkload Well-BeingAutonomy169   4.3.2.2. Organisational well-being. The second variable to be examined was organisational well-being. The results are shown in Table 4.11. The unconditional means model (Model 1) revealed that organisational well-being varied statistically significantly among participants (p < .001). In addition, the between-person variance (?00 = 0.870, p < .001) and the within-person variance (?2 = 0.250, p < .001) were statistically significant, with more variance occurring at the between-level than the within-level. The unconditional growth model (Model 2a) revealed that time as a fixed effect did not add significantly to the explanation of the outcome variable (p = .302). The addition of time2 to test the curvature of the model was also non-significant, confirming that there was no change over time (results not presented). As with workload well-being, the time variable was left in subsequent conditional models to contextualise the data. The first conditional model involved the addition of the four need satisfaction variables (Model 3). Two need satisfaction variables were significant and so the final model (Model 4) was run with only the two significant need satisfaction variables in the model. Relatedness with colleagues (p = .001) and autonomy (p < .001) were the significant predictors and they both had positive relationships with the outcome variable. As teachers? reports of relatedness with colleagues and autonomy increased, so did their organisational well-being. A chi-square test revealed that the final model provided significantly greater fit than Model 2a: ?2 (2) = 117.82, p < .001. The two need satisfaction variables explained 72% of the between-person variance (?00 was reduced from 0.887 in Model 2a to 0.247 in Model 4).   170   The following equation was used to graph the predicted values of organisational well-being: Yti = ?00 + ?01Timeti + ?02Relatedness with Colleaguesi + ?05Autonomyi This equation was solved for three different levels of need satisfaction: low, average, and high levels of relatedness with colleagues and autonomy. Figure 4.2 shows the relationship between organisational well-being and need satisfaction. Teachers who reported high relatedness with colleagues and autonomy experienced markedly greater organisational well-being than teachers reporting average or low relatedness with colleagues or autonomy.   171    Table 4.11 Multilevel Models for Examining Change in Organisational Well-Being Over Time  Model 1 Model 2a Model 3 Model 4 Fixed effects     Intercept (?00) 5.025*** 5.062*** 4.958*** 4.959*** Timeti (?01)  -0.075 -0.048 -0.051 Rel. with teachers (?02)   0.348*** 0.347** Rel. with students (?03)   -0.173  Competence (?04)   0.256  Autonomy (?05)   0.845*** 0.907*** Random effects     ?00 0.870*** 0.877*** 0.227** 0.247** ?2 0.250*** 0.243*** 0.232*** 0.229*** Fit statistics     -2LL -144.415 -143.900 -86.393 -87.466 AIC 294.830 295.800 188.787 186.931 *p < .05, ** p < .01, *** p < .001.     172       Figure 4.2. Predicted values of organisational well-being for different levels of relatedness with colleagues and autonomy.   33.544.555.56Low Average HighOrganisational Well-BeingRelatedness with Colleagues and Autonomy173   4.3.2.3. Student interaction well-being. The results of the multilevel models predicting change in student interaction well-being are shown in Table 4.12. The unconditional means model (Model 1) revealed that student interaction well-being varied significantly among participants (p < .001). The between-person (?00 = 0.661, p < .001) and within-person (?2 = 0.268, p < .001) variance components were both significant, with more variance occurring between participants. Model 2a revealed that the addition of time as a fixed effect was not significant (p = .867), nor was the inclusion of time2 (results not presented). We can conclude that student interaction well-being did not change significantly over time. Model 3 revealed that autonomy was a significant factor in the model. Model 4 confirmed autonomy as a significant and positive predictor (p < .001) while excluding the non-significant need satisfaction variables. As teachers? reports of autonomy increased, so did their student interaction well-being. Furthermore, autonomy explained 27% of the between-person variance (?00 was reduced from 0.658 in Model 2a to 0.479 in Model 4). The chi-square difference test revealed that Model 4 significantly improved model fit: ?2 (1) = 44.61, p < .001. The following equation was used to graph the predicted values of student interaction well-being:  Yti = ?00 + ?01Timeti + ?05Autonomyi Figure 4.3 shows the relationship between student interaction well-being and autonomy. Teachers who reported high levels of autonomy experienced greater student interaction well-being than teachers reporting low or average autonomy.  174    Table 4.12 Multilevel Models for Examining Change in Student Interaction Well-Being Over Time  Model 1 Model 2a Model 3 Model 4 Fixed effects     Intercept (?00) 5.272*** 5.266*** 5.180*** 5.185*** Timeti (?01)  0.012 0.004 -0.001 Rel. with colleagues (?02)   -0.083  Rel. with students (?03)   0.093  Competence (?04)   0.397  Autonomy (?05)   0.523* 0.646*** Random effects     ?00 0.661*** 0.658*** 0.426** 0.479** ?2 0.268*** 0.269*** 0.239*** 0.238*** Fit statistics     -2LL -140.780 -140.767 -98.529 -100.690 AIC 287.560 289.534 213.058 211.380 *p < .05, ** p < .01, *** p < .001.       175       Figure 4.3. Predicted values of student interaction well-being for different levels of autonomy.   33.544.555.56Low Average HighStudent Interaction Well-BeingAutonomy176   4.3.2.4. Self-efficacy for student engagement. The first self-efficacy variable tested was self-efficacy for student engagement. The results of the multilevel models are shown in Table 4.13. Self-efficacy for student engagement varied significantly (p < .001) among participants as shown in the unconditional means model (Model 1). The two variance components were significant (?00 = 0.616, p = .001; ?2 = 0.427, p < .001). Once again, there was more between-person variance than within-person variance. The addition of time as a fixed effect in Model 2a was not significant (p = .238), nor was the addition of time2 (results not presented). The conditional models (Models 3 and 4) revealed that relatedness with students was a significant and positive predictor (p = .003) of the outcome variable. As teachers? reports of relatedness with students increased, so did their self-efficacy for student engagement. Relatedness with students explained 31% of the between-person variance (?00 was reduced from 0.616 in Model 2a to 0.424 in Model 4). The chi-square difference test revealed that Model 4 significantly improved model fit: ?2 (1) = 57.79, p < .001. The equation used to graph the predicted values of self-efficacy for student engagement was as follows: Yti = ?00 + ?01Timeti + ?03Relatedness with Studentsi  Figure 4.4 shows the relationship between self-efficacy for student engagement and relatedness with students. Teachers who reported high levels of relatedness with students experienced the greatest self-efficacy for student engagement.     177    Table 4.13 Multilevel Models for Examining Change in Self-Efficacy for Student Engagement Over Time  Model 1 Model 2a Model 3 Model 4 Fixed effects     Intercept (?00) 6.895*** 6.830*** 6.800*** 6.793*** Timeti (?01)  0.115 0.104 0.107 Rel. with colleagues (?02)   -0.096  Rel. with students (?03)   0.555? 0.787** Competence (?04)   0.208  Autonomy (?05)   0.313  Random effects     ?00 0.616** 0.616*** 0.370** 0.424** ?2 0.427*** 0.417*** 0.473*** 0.480*** Fit statistics     -2LL 136.487 -135.766 -110.723 -113.168 AIC 278.975 279.532 237.446 236.336 ? p = .05, *p < .05, ** p < .01, *** p < .001.       178       Figure 4.4. Predicted values of self-efficacy for student engagement for different levels of relatedness with students.   6.46.66.877.27.47.67.8Low Average HighSelf-Efficacy for Student EngagementRelatedness with Students179   4.3.2.5. Self-efficacy for classroom management. Table 4.14 shows the results for self-efficacy for classroom management. Model 1 revealed that participants? scores on the outcome variable varied significantly (p < .001) and that there was significant between-person (?00 = 0.575, p < .001) and within-person variance (?2 = 0.478, p = .001). Unlike previous models, the two variance components were close in size suggesting that there were almost equal amounts of variance to explain at the two levels.  In Model 2a, time was entered as a fixed effect. It was significant (p = .035), suggesting that there was variation in participants? scores on the outcome variable over time. The fixed time effect explained 6% of the within-person variance (?2 was reduced from 0.478 in Model 2a to 0.451 in Model 2a). Furthermore, the chi-square difference test revealed that the inclusion of time as a fixed variable significantly improved the fit of the model: ?2 (1) = 5.48, p = .019. In two subsequent models, time was allowed to vary randomly and time2 was entered as a fixed predictor, but neither effects were significant (results not presented). It was concluded that self-efficacy for classroom management changed linearly over time and that there was no variation in the trend across participants.  The conditional models (Models 3 and 4) revealed two significant need satisfaction variables. Relatedness with students (p = .001) and autonomy (p = .002) were both significant and positive predictors. The addition of the two predictors in Model 4 significantly improved model fit (?2 [2] = 51.28, p < .001) and explained 56% of the between-person variance (?00 was reduced from 0.554 in Model 2a to 0.244 in Model 4). The equation used to graph the predicted values of self-efficacy for classroom management over time was as follows: Yti = ?00 + ?01Timeti + ?03Relatedness with Studentsi + ?05Autonomyi 180   where the time coefficient (? 01) was 0.233 as specified in Model 4. Figure 4.5 shows the trajectories for teachers with low, average, and high levels of relatedness with students and autonomy. This is similar to the graphs shown for the other outcome variables in that need satisfaction was positively correlated with the outcome; however, where the other graphs did not include time, this one did. It shows how self-efficacy for classroom management changed over time for teachers reporting low, average, and high levels of relatedness with students and autonomy. The trajectories show that teachers reported greater self-efficacy over time. Moreover, the results show that the starting point for teachers reporting high relatedness with students and autonomy was greater than the final point for teachers reporting low and average levels of these two needs, suggesting that need satisfaction plays a major role in self-efficacy for classroom management.   181    Table 4.14 Multilevel Models for Examining Change in Self-Efficacy for Classroom Management Over Time  Model 1 Model 2a Model 3 Model 4 Fixed effects     Intercept (?00) 7.223*** 7.094*** 7.062*** 7.063*** Timeti (?01)  0.223* 0.230* 0.233* Rel. with colleagues (?02)   -0.089  Rel. with students (?03)   0.737** 0.751** Competence (?04)   0.048  Autonomy (?05)   0.473** 0.440** Random effects     ?00 0.575*** 0.554*** 0.243 0.244 ?2 0.478** 0.451*** 0.495** 0.499*** Fit Statistics     -2LL -138.498 -135.970 -106.857 -107.094 AIC 282.996 279.939 229.714 226.188 *p < .05, ** p < .01, *** p < .001.      182       Figure 4.5. Predicted values of self-efficacy for classroom management for different levels of relatedness with students and autonomy.   66.577.588.5Time 1 Time 2 Time 3Self-Efficacy for Classroom ManagementLow Relatedness with Students and AutonomyAverage Relatedness with Students and AutonomyHigh Relatedness with Students and Autonomy183   4.3.2.6. Self-efficacy for instructional strategies. Table 4.15 shows the results of the multilevel models for self-efficacy for instructional strategies. Model 1 showed that scores on the outcome varied significantly between participants (p < .001), and that the between-person variance was significant (?00 = 0.854, p < .001), as was the within-person variance (?2 = 0.291, p = .006). The addition of time as a fixed effect in Model 2a was not significant (p = .830), nor was the addition of time2 (results not presented). The conditional models (Models 3 and 4) revealed that relatedness with students was a significant and positive predictor (p = .001) of the outcome variable. The addition of relatedness with students in the final model explained 27% of the between-person variance (?00 was reduced from 0.854 in Model 2a to 0.620 in Model 4). The chi-square difference test confirmed that the addition of relatedness with students in the model significantly improved model fit: ?2 (1) = 52.65, p < .001. The equation used to graph the predicted values of self-efficacy for instructional strategies was as follows: Yti = ?00 + ?01Timeti + ?03Relatedness with Studentsi Figure 4.6 shows the relationship between self-efficacy for instructional strategies and relatedness with students. Teachers reporting greater relatedness with students also reported greater self-efficacy for instructional strategies.   184    Table 4.15 Multilevel Models for Examining Change in Self-Efficacy for Instructional Strategies Over Time  Model 1 Model 2a Model 3 Model 4 Fixed effects     Intercept (?00) 7.200*** 7.184*** 7.221*** 7.213*** Timeti (?01)  -0.014 -0.008 -0.009 Rel. with colleagues (?02)   0.172  Rel. with students (?03)   0.853** 0.982** Competence (?04)   0.048  Autonomy (?05)   0.102  Random effects     ?00 0.854*** 0.854*** 0.567** 0.620** ?2 0.291** 0.291** 0.325* 0.319* Fit statistics     -2LL -133.511 -133.497 -107.958 -109.178 AIC 273.022 274.994 231.917 228.356 *p < .05, ** p < .01, *** p < .001.      185     Figure 4.6. Predicted values of self-efficacy for instructional strategies for different levels of relatedness with students.  4.3.2.7. Summary of need satisfaction effects. Table 4.16 summarises the results of the multilevel models showing the observed predictors, along with the original hypotheses. As shown, Hypotheses 2a to 2e were partially supported. In contrast, no support was found for Hypothesis 2f. Autonomy and relatedness with students had the most consistent effect, having a significant relationship with the outcome variables in four and three models, respectively. Relatedness with colleagues had a significant effect in one model (i.e., organisational well-being). The results for competence were unexpected. Although it was hypothesised as having a significant relationship with all of the outcome variables, it was not a significant effect in any of the models. 6.46.66.877.27.47.67.8Low Average HighSelf-Efficacy for Instructional StrategiesRelatedness with Students186   Table 4.16 Hypothesised and Observed Significant Need Satisfaction Effects on the Outcome Variables Hypothesis number Outcome Hypothesised predictors Observed predictors 2a Workload well-being Competence Autonomy  Autonomy 2b Organisational well-being Relatedness with colleagues Competence Autonomy  Relatedness with colleagues Autonomy 2c Student interaction well-being Relatedness with students Competence Autonomy  Autonomy 2d Self-efficacy for student engagement Relatedness with students Competence  Relatedness with students 2e Self-efficacy for classroom management Relatedness with students Competence Autonomy  Relatedness with students Autonomy 2f Self-efficacy for instructional strategies Competence Autonomy Relatedness with students  187   4.4. Discussion The need to improve our understanding of how teacher well-being and motivation function over time is a key priority regularly highlighted in the literature (e.g., Klassen et al., 2011). Study 3 aimed to address this issue by conducting growth curve modelling. Results showed that self-reported well-being was stable (Hypothesis 1a), whereas self-efficacy for classroom management showed a positive linear trend over time (Hypothesis 1b; self-efficacy for student engagement and instructional strategies were both stable over time). In addition, results showed that both well-being and self-efficacy were predicted by need satisfaction (Hypotheses 2a through 2f). Key findings from the study are discussed below. 4.4.1. Self-efficacy for Classroom Management The findings of the study showed that self-efficacy for classroom management changed over time among the teachers in the sample and that it showed a positive linear trend increasing over the three time points. There are several reasons why this may have occurred, the first of which is a natural growth in teachers? confidence as the school year progresses and as they get to know their students better (e.g., they have more understanding of how to best interact with individual students). Hence, the passing of two to three months between the first and third time point of data collection may have been enough to precipitate this change in self-efficacy for classroom management. Collie and colleagues? (2012) research showing that positive perceptions of student behaviour and motivation among teachers predicted greater self-efficacy for teaching supports this interpretation, as does longitudinal research showing positive changes in self-efficacy for teaching over the school year (e.g., Pas et al., 2012). 188   Previous research examining the stability of self-efficacy over time has also suggested that professional development can have a positive effect on self-efficacy (e.g., Carleton et al., 2008). Based on the reasons discussed earlier, however, the interpretations of Carleton and colleagues? (2008) research is hampered by the conceptually troubled instrument that was used. Nonetheless, another possible explanation for the change in self-efficacy found in Study 3 is teachers? experiences of professional development during the course of data collection. This may include the anti-oppression workshop in which they participated as well as other professional development sessions they likely attended. In order to determine whether professional development influences self-efficacy for teaching, there is a need for intervention research involving multiple forms of data collection (e.g., questionnaires, observations, interviews). This will help us to identify whether and how professional development influences self-efficacy for teaching.  Notwithstanding the fact that we can only speculate why the change occurred, the finding broadens understanding of self-efficacy for classroom management in a couple of key ways. First, it extends previous research that has looked at changes in self-efficacy for teaching among early career teachers (e.g., Woolfolk Hoy & Spero, 2005) by revealing that changes among more experienced teachers are possible. In that sense, it supports Klassen and Chiu?s (2010) cross-sectional research showing a positive relationship between self-efficacy for teaching and years of experiences up to the mid-point in the career, while extending this to consider multiple time points. Second, this finding extends previous research that looked at changes across longer time frames (e.g., Pas et al., 2012) by suggesting that self-efficacy for classroom management can change over short periods of time. Furthermore, by using a 189   conceptually sound instrument it provides much needed support for similar findings in Carleton and colleagues? (2008) study. The implications of this finding extend to students, educators, and administrators. Self-efficacy for teaching has been associated with several important outcomes for teachers (e.g., work engagement; Schaufeli & Bakker, 2004) and students (e.g., student achievement; Caprara et al., 2006). More specifically, self-efficacy for classroom management has been negatively associated with teachers? reports of stress, burnout (Brouwers & Tomic, 1999, 2000), and the use of punitive strategies for classroom management (Morin & Battalio, 2004). Consequently, efforts that can help to improve teacher self-efficacy affect not only the teachers themselves, but their students? learning and achievement. In addition, Collie and colleagues (2012) showed that self-efficacy played a mediating role in the relationship between teacher stress from student behaviour and job satisfaction. They concluded that stress appears to have a detrimental effect on teachers when it is accompanied by low self-efficacy. Efforts that can help to bolster teacher confidence should be considered by administrators to help offset damage caused by teacher stress levels. This is particularly important given that teaching is recognised as a highly stressful career (e.g., Kyriacou, 2000, 2001) and that this stress results in lower work engagement among teachers (Hakanen et al., 2006) and lower learning motivation among students (Pakarinen et al., 2010).  4.4.2. Teacher Well-Being The three types of teacher well-being showed no changes over time. For this finding, it is possible that no changes occurred because teacher well-being is relatively stable over short time periods. This is supported by the literature on subjective well-being (e.g., Schimmack & Lucas, 2010; Suh et al., 1996), which has shown that unless individuals are 190   regularly engaged in positive activities designed to improve well-being it remains relatively stable and returns to baseline levels within three months of any major life changes (Lyubomirsky et al., 2011). An alternative explanation for why no changes in well-being were observed over time relates to the level of statistical power in Study 3. Given the small sample size, it is possible that non-significant findings were due to inadequate power. Future research with larger sample sizes is needed to examine teacher well-being over time to see whether similar results are found or whether time becomes significant with an increase in power. The inclusion of questions asking teachers about any recent changes in work or life circumstances is also a fruitful avenue for future research as it would help to improve our understanding of whether and how life events affect teacher well-being in ways that are similar to subjective well-being (i.e., minimally).  In spite of the concerns about statistical power, this finding is important for research and practice. It extends our understanding of teacher well-being in two key ways. First, it suggests that it acts differently from stress and burnout, at least in the short-term. Research has revealed that stress and burnout change over the course of a school year (e.g., Pas et al., 2012) and across even shorter time frames in the case of an intervention (e.g., Roeser et al., 2013). This differential functioning between stress/burnout and well-being over time provides support for the examination of teacher well-being as a distinct construct and not just as the absence of stress or burnout. Second, this finding suggests that teacher well-being may function in similar ways to individuals? global assessments of their subjective well-being. In other words, it may be a stable construct that is largely unaffected by everyday occurrences. For practice, this finding suggests that in the short-term teacher well-being may be somewhat robust in the face of negative events that occur in teaching (e.g., sudden increases in 191   disruptive behaviour among students). The reverse of this, however, is that teachers reporting low well-being may not experience a long-term or significant increase in well-being when positive events occur (e.g., an increase in salary). This raises concerns about how to improve teacher well-being over the long term and, thus, highlights the importance of considering more stable, contextual factors like need satisfaction so as to promote lasting improvements in teacher well-being. Another implication of these findings is that they provide further support for the three factors of teacher well-being developed in Study 1 with a different sample of teachers. Briefly, scores on workload well-being were much lower than scores on organisational and student interaction well-being. This corroborates similar findings from Study 1 and provides further evidence for the conceptual meaning of the factors. In addition, it reiterates the fact the workload generally affects teacher well-being less positively than organisational issues and student interaction.  4.4.3. Need Satisfaction The inclusion of the need satisfaction variables as time-invariant predictors in the multilevel models explained significant amounts of between-level variance for all models. In all cases, need satisfaction predicted higher levels of well-being and self-efficacy. In particular, satisfaction of the needs for autonomy and relatedness with students had the most consistent effect on the outcome variables. Autonomy had a significant relationship with workload well-being, organisational well-being, student interaction well-being, and self-efficacy for classroom management. These relationships were expected and likely occurred because autonomy allows teachers to feel a sense of choice and control over their workload, their role in the school, how they interact with students, and how they manage their 192   classroom. These findings are supported by previous research showing links between autonomy and other measures of well-being (i.e., a negative association with stress; Collie et al., 2012) and motivation (i.e., a positive association with work engagement; Klassen, Perry, et al., 2012).  Relatedness with students had, as expected, a significant relationship with self-efficacy for student engagement and classroom management. Contrary to expectations, it also had a significant relationship with self-efficacy for instructional strategies. For these findings, it is possible that a positive connection with students allows teachers to feel more confident in their ability to engage, manage, and effectively instruct those students. Indeed, several scholars (Hamre & Pianta, 2006; Jennings & Greenberg, 2009; Martin & Dowson, 2009) have highlighted the importance of positive teacher-student relationships for student engagement, effective classroom management, and effective teaching and learning, which likely explains the unexpected relationship between relatedness with students and self-efficacy for instructional strategies.  These findings extend our understanding of how autonomy and relatedness with students are associated with teacher well-being and self-efficacy. They support Klassen, Perry, et al.?s (2012) assertion of the relative importance of relatedness with students over relatedness with colleagues for explaining teacher outcomes (i.e., relatedness with colleagues only predicted one outcome?organisational well-being). For well-being, the findings corroborate results from Study 2 that linked autonomy with a composite teacher well-being variable, while also extending our understanding by revealing relationships from autonomy to each of the three separate factors of teacher well-being. For self-efficacy, the findings add 193   to our understanding of the needs for autonomy and relatedness with students by showing that they are relevant for self-efficacy.  These findings are also important for administrators and policy-makers who should be aware of the significant role that autonomy and relatedness with students play in teacher well-being and self-efficacy. In particular, the findings support the need for efforts at the school-level to create autonomy-supportive school climates. Furthermore, they suggest that administrators and policy-makers should pay greater attention to the importance of positive teacher-student relationships. The final need satisfaction variable to be discussed briefly is competence. Contrary to expectations it was not significantly related to any of the outcomes. This is intriguing given the significant relationship found in Study 2 for well-being. Perhaps it speaks to the small sample size or some uniqueness about the sample of teachers in Study 3.  For self-efficacy, perhaps competence was not related because it is more aligned with self-concept and ability based on past/current experiences (e.g., I am competent at my job), rather than judgements of future ability that are context-specific (e.g., I believe can do a lot to control disruptive behaviour in my classroom). For example, if a teacher has worked for many years, he or she may feel competent in the job. However, if the teacher happens to have a particularly difficult class in one year, his or her confidence about how much can be done to engage that group of students or manage the classroom may be quite low. Future research is needed that examines this relationship among larger samples and in more detail, including investigation of whether a reversed relationship between these variables is more relevant.  194   4.5. Limitations and Future Directions Study 3 has several limitations that must be discussed. The first concerns the sample size. Although significant effects were found, a larger sample size would have provided more power and given stronger assurances that non-significant effects were not due to low power. In order to provide further support for the findings, samples involving a larger number of teachers are a necessary step in future research. This will provide more conclusive evidence of whether teacher well-being is a stable construct (like subjective well-being) or whether changes can be expected over time.  The second limitation concerns the time frame. Although the short time frame did not affect the findings per se, it did not allow me to examine the impact of the different stages of a school year (beginning, end, report writing, etc.). Given that different points in the school year are associated with unique and changing demands on teachers, there is a need for research that examines these variables over the course of a full school year (and beyond) with more waves of data collection. This would provide a greater understanding of whether and how teacher well-being and motivation fluctuate across (and beyond) a school year, along with the influence of the changing contextual and personal experiences of teachers. In spite of this limitation, the current findings do provide important understanding about the flexibility of these constructs in the short term.  Another limitation is that I was not able to ascertain why teachers dropped out of Study 3 and whether they were identifiable by a confounding variable (e.g., whether they were too busy to fill out the questionnaire which could affect workload well-being scores). The findings may be conservative, therefore, in that only teachers functioning reasonably well were able to continue participating in the study. The statistical comparisons of teachers 195   who remained in the study with those who dropped out provide evidence that there were no differences between the groups at Time 1; however, we do not know if any relevant changes occurred after this point (e.g., because of a negative experience with the workshop or major changes in teachers well-being or motivation). In the future, efforts that attempt to track why participants drop out will be helpful to rule out and/or help understand reasons for attrition. This will also provide greater understanding of the representativeness of the data provided by participants involved in all waves of data collection. The final limitation to be discussed is that the participants were recruited through the booking process of a professional development program. As such, the sample may have been unique in some way compared with the broader population. The fact that reports of well-being and self-efficacy were comparable with previous research (e.g., Study 1 and 2 in this dissertation; Collie et al., 2012) provides support for the generalisability of the findings. Nonetheless, there is a need for research that uses different samples of teachers in order to corroborate the findings of Study 3. In addition, it would be interesting to involve groups of teachers at different careers stages so as to understand how years of teaching experience influence changes in well-being and self-efficacy for teaching (e.g., whether changes in these constructs occur more readily among early career teachers than more experienced teachers).  4.6. Conclusions In summary, Study 3 has involved examining how reports of teacher well-being and self-efficacy for teaching functioned over two to three months. This type of research provides an important addition to the understanding we have gained from studies conducted at one point in time. By using multilevel modelling, results revealed a positive change in self-efficacy for classroom management over time; however, there were no changes over time for 196   the remaining variables. Also investigated was the influence of basic psychological need satisfaction on teachers? initial reports of well-being and self-efficacy. The results showed that teachers? sense of autonomy, relatedness with students, and relatedness with colleagues predicted greater well-being and self-efficacy among the participants. Combined, these findings extend our understanding of the flexibility of self-efficacy for classroom management over short time periods. In addition, they indicate that well-being is a stable construct as measured by the TWBS. Study 3, along with the other two studies, provides important implications for research and practice. These are discussed in the General Discussion below.       197   CHAPTER 5. GENERAL DISCUSSION The aim of the dissertation research was to provide a comprehensive examination of teacher well-being and motivation. Three related studies were conducted that developed a measure of well-being, tested an explanatory model of teachers? experiences of well-being, motivation, job satisfaction, and organisational commitment, and examined teachers? reports of well-being and motivation over time. More specifically, the first study involved developing and testing a measure of teacher well-being based on teachers? experiences at work. Results revealed three factors of teacher well-being. The first was workload well-being, which refers to teachers? perceptions of the amount of marking, teaching, and administrative work they have, the work/meetings they have to do after school hours, and their ability to fit everything into the allotted time. The second was organisational well-being, which refers to teachers? perceptions of support offered by and relations with school leadership, the culture towards teachers and teaching, and the level of participation they have in decision making. The third was student interaction well-being, which refers to teachers? perceptions of student motivation, student behaviour, classroom management, and their relations with students. Comparisons showed that demographic characteristics did not play a substantial role in determining scores on the three factors of well-being, indicating that the measure functions similarly across the different demographic groups that were present in the sample of teachers. Furthermore, correlational analyses provided support for the three factors of well-being showing that they were related as expected with three other measures of teachers? experiences: stress, job satisfaction, and flourishing (i.e., general life well-being). The second study involved describing and testing an explanatory model of teacher well-being, motivation, job satisfaction, and affective organisational commitment. The 198   results provided support for the model. They showed that the relationships established by SDT were supported: perceptions of autonomy support predicted need satisfaction that, in turn, predicted well-being and motivation. Furthermore, the relationships involving well-being, motivation, job satisfaction, and affective organisational commitment were generally supported. Namely, teachers? reports of autonomous motivation (measured as identified regulation) and well-being were positively associated with job satisfaction and affective organisational commitment. There were some unexpected relationships, however, involving autonomy and controlled motivation (measured as introjected regulation and external regulation), as well as between controlled motivation, job satisfaction, and affective organisational commitment. The third study involved an examination of teachers? reports of well-being and self-efficacy for teaching over time. Results showed that self-efficacy for classroom management changed over time, whereas the three types of teacher well-being (i.e., workload, organisational, and student interaction well-being) and the other two types of self-efficacy (i.e., self-efficacy for student engagement and instructional strategies) did not. Findings also revealed the significance of basic psychological need satisfaction in predicting teacher well-being and self-efficacy. In particular, need satisfaction of autonomy, relatedness with students, and relatedness with colleagues positively predicted greater well-being and self-efficacy for teaching.  Taken together the findings from these three studies advance knowledge in the field of teacher well-being and motivation. Before discussing the major contributions of the dissertation research, however, there are several findings that refer to the dissertation as a whole and warrant discussion first. The first was support found for the Teacher Well-Being 199   Scale (TWBS) developed in Study 1. The factor analyses and correlations in Study 1, along with the SEM in Study 2, provided validity evidence based on internal structure and relations to other variables. Furthermore, the multilevel analyses in Study 3 provided further support for the three factors by revealing their stability over time and how they relate to need satisfaction. Taken together, these findings provide important initial evidence of validity, including reliability, to support the interpretations of scores on the TWBS as a measure of teacher well-being. Future research is needed to examine the measure among different samples and in relation to other constructs to provide further evidence of validity. The second finding relevant across the research as a whole was the relationship that sense of need satisfaction for autonomy had with reported motivation. The results of Study 2 revealed a double-sided view of autonomy. It was positively associated with autonomous motivation and controlled motivation suggesting that high autonomy can be reported by teachers acting with very different motivations (e.g., ?I do this job for the paycheque?, which is associated with external regulation, or ?because this job fits my personal values?, which is associated with identified regulation; Gagn? et al., 2010). The findings of Study 3 revealed a positive relationship between experiences of autonomy and self-efficacy: autonomy predicted greater self-efficacy for student engagement, classroom management, and instructional strategies. Taken together, these findings highlight the importance of examining all three needs simultaneously in both research and practice. That is, if we focus only on promoting autonomy we may be supporting not only identified regulation and self-efficacy for teaching, but also external regulation (i.e., because autonomy positively predicted both positive and negative forms of motivation). Given that external regulation was negatively associated with the needs for relatedness and competence, a focus on promoting all three 200   needs simultaneously will help to ensure that we are encouraging teachers to move towards positive forms of motivation and away from external regulation. This finding also highlights the need for future research using person-centred approaches to examine teachers? experiences. Person-centred approaches examine differences among individuals in how variables relate to one another (Laursen & Hoff, 2006). Latent profile analysis and cluster analysis are two means for providing understanding of the different groups of teachers in the population and how they experience the different constructs. For example, these techniques would allow for the identification of groups of teachers, such as those reporting high autonomy but also high external regulation, in order to understand predictors and outcomes that we can expect among this group. The third common finding reflected across the three studies was that contextual factors need to be considered when attempting to understand teachers? experiences at work. For the development of the TWBS in Study 1, contextual factors made up the bulk of all items in the final subscales (e.g., ?Support offered by school leadership?). In Study 2, contextual factors of perceived autonomy support and need satisfaction were central in teachers? reports of their well-being and motivation. Finally, in Study 3 the contextual factors of need satisfaction provided important depth to our understanding of teacher well-being and self-efficacy. This extends previous research that has highlighted the importance of school climate for teacher outcomes (e.g., Collie et al., 2012; Grayson & Alvarez, 2008) and indicates the importance of considering a wide range of contextual experiences?at both the classroom and school level?when attempting to understand and promote teacher well-being and motivation.  201   5.1. Conclusions and Implications The overall purpose of the dissertation research was to address several significant gaps in the literature through a three-part examination of teacher well-being and motivation. Various researchers have indicated the need for educational psychology research to focus on advancing our scientific understanding, while also providing useful and practical applications that are accessible to schools and relevant stakeholders (e.g., Martin, 2007; Pintrich, 2000). This dissertation has aimed to address both of these goals and they are discussed below in turn, along with implications for research and practice. 5.1.1. Contributions for Scientific Understanding The main contribution of this dissertation research is that it has extended our understanding of the constructs of teacher well-being and motivation in several key ways. In particular, the findings have provided support for a multidimensional construct of teacher well-being that is distinct from related variables such as stress, job satisfaction, and flourishing (i.e., general life well-being). This is important because it highlights how different aspects of teachers? work have varying polarities of influence on their well-being. In particular, the findings have highlighted the negative influence of workload, and the positive influence of organisational-level and student-related aspects of teachers? work on their well-being. An implication of this contribution is that it has highlighted the value of allowing participants to respond in ways that consider the full range of influences (both the positive and negative). By allowing this, the TWBS provided a more complete picture of participants? experiences than would have been possible if considering only the negative or positive influence of the items. 202   In relation to motivation, the findings have provided corroborating evidence to support the relevance of the different types of regulation from SDT (Deci & Ryan, 1985, 2002) for understanding teacher motivation, while at the same time raising questions about the psychometric differences between intrinsic regulation and job satisfaction. Indeed, the findings indicate the need for more nuanced methods for assessing the differences between these constructs (e.g., interviews, more sensitive measures).  The dissertation has also provided evidence of how teacher well-being and self-efficacy function over time. Given that teacher well-being appeared to be stable over time, it may not only be robust to negative events, but also to positive events or efforts to improve it (practical implications of this are discussed below). This indicates the need to examine teacher well-being in greater detail to understand whether other factors can help to improve it over the long term. Future research examining need satisfaction as a time-varying covariate would help to extend our understanding in this area as it would provide an indication of whether and how changes in need satisfaction influence changes in well-being. Additional avenues of research that would be revealing include attempts to control for individual base-line levels of well-being and examinations of whether regular well-being exercises can help to improve teacher well-being (as per Roeser et al., 2013).  For self-efficacy, the findings have extended our understanding about the flexibility of self-efficacy for classroom management over a relatively short time frame. This finding is important given that self-efficacy appears to buffer the negative effects of stress (e.g., Collie et al., 2012) and given that it is positively associated with student achievement (e.g., Caprara et al., 2006). Moving forward, research that examines why this occurred (e.g., by conducting 203   interviews with participants) and whether this increase holds over a longer time frame would further advance our understanding. Another contribution of the dissertation research is that it has extended our understanding of how teachers? experiences of well-being and motivation relate to other important constructs (e.g., job satisfaction). Where previous research has revealed that these constructs are related (e.g., Hakanen et al., 2006; Klassen & Chui, 2011), this dissertation research has shown how these constructs are related via the explanatory model. In addition, the dissertation has extended previous research on job satisfaction by highlighting that motivation may be an important determinant in whether or not job satisfaction is relevant for certain teachers. Furthermore, the findings provide a greater understanding of autonomy, the relationship this can have with different types of motivation, and the need for more sensitive measures that differentiate between the positive and negative reasons underlying autonomy. Taken together, these findings have underscored the complexity of teachers? work and the value of considering multiple constructs simultaneously in order to gain a more complete picture of teachers? experiences (Collie et al., 2012).   This dissertation research has also provided evidence of the relevance of existing theory for understanding teachers? experiences at work. In particular, the findings have provided corroborating evidence for previous research using SDT (Deci & Ryan, 1985, 2002) to examine teachers? experiences at work (e.g., Klassen, Perry, et al., 2012; Roth et al., 2007; Taylor et al., 2008). Moreover, it has extended the literature by suggesting that SDT is an appropriate framework for examining how well-being and motivation relate to job satisfaction and affective organisational commitment. In addition, the study raises several 204   questions for future research about the causes and correlates of introjected regulation among teachers.  Another area in which this dissertation provides research contributions is in relation to contextual factors. In particular, the findings have deepened our understanding of the contexts within which teachers work and how these contexts can influence teacher well-being and motivation. The dissertation research has revealed that different contextual factors have varying degrees of importance for teachers? reports of well-being and motivation. It has revealed that contextual factors such as perceptions of principals? autonomy support and interactions with students played substantial roles in teachers? experiences of well-being and motivation, as well as other outcomes (e.g., need satisfaction, job satisfaction).  The findings have also revealed that the examined demographic characteristics did not play a major role in determining teachers? reports of well-being. In fact, the findings indicate that researchers may want to consider other demographic characteristics (e.g., teaching positions) and other factors (e.g., beliefs) when attempting to understand what drives differences in teacher well-being. There is a need to conduct research among different samples of teachers to understand whether demographic variables play a more substantial role among more diverse populations. Notwithstanding this, an important contribution of the dissertation research is that it suggests that efforts by administrators and policy makers to improve teacher well-being should have a broad influence among teachers who are demographically similar to the participants who were involved in this research. Additional practical contributions of the dissertation are discussed below. 205   5.1.2. Contributions for Practice The findings of this dissertation research provide several important contributions for practice. Overall, it is my hope that the findings will help to cultivate awareness of the importance of teacher well-being and motivation and, in turn, promote supportive and positive work experiences for teachers. The first contribution relevant to practice concerns the development of the TWBS. The practical nature of the instrument?involving the examination of core aspects of teachers? work?means that it can be used to inform practice. Potentially, it could be used by school administrators, policy-makers, and even educators themselves in order to understand what aspects of teaching work positively or negatively influence teacher well-being. Thus, while extending our understanding of teacher well-being, possible uses of the TWBS also extend to diagnosis and intervention.  The second contribution relevant to practice concerns the explanatory model of teacher well-being, motivation, job satisfaction, and affective organisational commitment. The model provides a framework through which school administrators and policy-makers can better understand the experiences of well-being and motivation among teachers. It also describes how these experiences are affected by and, in turn, affect other experiences such as need satisfaction and job satisfaction. Consequently, the model has the potential to help guide efforts to promote positive work experiences among teachers. Given the links between teacher well-being/motivation and student outcomes (e.g., achievement; Caprara et al., 2006), such efforts also have the potential to positively impact students. The findings concerning changes in teachers? reports of well-being and self-efficacy over time, although more removed from direct practical application, do provide relevant contributions. In particular, they provide an understanding of what types of changes can be 206   expected in teacher well-being and self-efficacy across relatively short time frames. This is important for school administrators and policy-makers who want to ensure that professional development efforts have the potential to affect change among teachers. More specifically, the positive change over time observed in self-efficacy for classroom management suggests that interventions designed to improve teachers? skills in this area can realistically expect to result in changes in teachers? confidence over short periods of time. Considering that self-efficacy for teaching is associated with teaching effectiveness (e.g., Cho & Shim, 2013), this is a worthwhile construct to target in the attempt to improve teaching quality.  In contrast, the findings for well-being suggest that perhaps changes over a short period of time cannot be expected. This has positive and negative ramifications for practice given that teacher well-being may not drop significantly when troubling events occur, but that it may not rise when efforts are made to improve work circumstances for teachers. The small sample size may have been partially responsible for no significant changes in well-being. Additional research with larger sample sizes is needed, therefore, along with research that investigates whether increases in teacher well-being can be fostered over the short- and long-term through other methods (e.g., basic psychological need support; regular well-being exercises as per Roeser et al., 2013). 5.1.3. Final Conclusions Taken together, this dissertation research has important implications for research and practice regarding teacher well-being and motivation. However, it is also important to mention that many, if not all, of these implications ultimately relate to students as well. The contributions of the dissertation to our understanding of teacher well-being and motivation extend to students given that these variables are related to teaching effectiveness (Duckworth 207   et al., 2009), resiliency in teaching (Klusmann et al., 2008), students? motivation (Pakarinen et al., 2010; Taylor & Ntoumanis, 2007) and students? achievement (Caprara et al., 2006). Furthermore, the practical implications and applications that this dissertation research can help promote and guide also have the potential to affect students in positive ways. Future research that examines links between the constructs examined in this dissertation (i.e., teacher well-being, autonomous motivation, self-efficacy for teaching, job satisfaction, and organisational commitment) and student outcomes more directly will help to extend our knowledge of how teachers? experiences relate to students? outcomes and, ultimately, how we can best support students? academic, social, and emotional development.    208   REFERENCES Albuquerque, I., Pedroso de Lima, M., Figueiredo, C., & Matos, M. (2011). Subjective well-being structure: Confirmatory factor analysis in a teachers? Portuguese sample. Social Indicators Research, Advance online publication. doi: 10.1007/s11205-011-9789-6  Allen, N. J., & Meyer, J. P. (1996). Affective, continuance, and normative commitment to the organization: An examination of construct validity. Journal of Vocational Behavior, 49, 252-276. doi: 10.1006/jvbe.1996.0043  American Psychological Association, American Educational Research Association, and National Council on Measurement in Education. (1999). Standards for educational and psychological testing. Washington, DC: American Psychological Association.  Asparouhov, T., & Muth?n, B. (2005) Multivariate statistical modeling with survey data. Proceedings of the Federal Committee on Statistical Methodology Research Conference. Retrieved from http://www.fcsm.gov/05papers/Asparouhov_Muthen_IIA.pdf Baard, P. P., Deci, E. L., & Ryan, R. M. (2004). Intrinsic need satisfaction: A motivational basis of performance and well-being in two work settings. Journal of Applied Social Psychology, 34, 2045-2068. Bandura, A. (1982). Self-efficacy mechanism in human agency. American Psychologist, 37, 122-147. doi: 10.1037/0003-066X.37.2.122 Bandura, A. (1989). Human agency in social cognitive theory. American Psychologist, 44, 1175-1184. doi: 10.1037/0003-066X.44.9.1175 Bandura, A. (1997). Self-efficacy: The exercise of control. New York: W. H. Freeman and Company. Barnab?, C., & Burns, M. (1994). Teachers' job characteristics and motivation. Educational Research, 36, 171.  Beretvas, S. N., Meyers, J. L., & Leite, W. L. (2002). A reliability generalization study of the Marlowe-Crowne Social Desirability Scale. Educational and Psychological Measurement, 62, 570-589. Billingsley, B. S., & Cross, L. H. (1992). Predictors of commitment, job satisfaction, and intent to stay in teaching: A comparison of general and special educators. Journal of Special Education, 25, 453-471.  Black, A. E., & Deci, E. L. (2000). The effects of instructors' autonomy support and students' autonomous motivation on learning organic chemistry: A self-determination theory perspective. Science Education, 84, 740-756.  Bogler, R., & Somech, A. (2004). Influence of teacher empowerment on teachers? organisational commitment, professional commitment and organisational citizenship 209   behaviour in schools. Teaching and Teacher Education, 20, 277-289. doi: 10.1016/j.tate.2004.02.003 Borg, M. G., & Riding, R. J. (1991). Occupational stress and satisfaction in teaching. British Educational Research Journal, 17, 263.  Bowling, N., Eschleman, K. J., & Wang, Q. (2010). A meta-analytic examination of the relationship between job satisfaction and subjective well-being. Journal of Occupational and Organisational Psychology, 83, 915-934. doi: 10.1348/096317909X478557 Boyle, G. J., Borg, M. G., Falzon, J. M., & Baglioni, A. J. (1995). A structural model of the dimensions of teacher stress. British Journal of Educational Psychology, 65, 49-67. Brackett, M., Palomera, R., Mojsa-Kaja, J., Reyes, M. R., & Salovey, P. (2010). Emotion-regulation ability, burnout, and job satisfaction among British secondary-school teachers. Psychology in the Schools, 47, 406-417. doi: 10.1002/pits.20478 Brenner, C. A., Perry, N. E., & Collie, R. J. (2012, September). Student teachers' developing practices that promote self-regulated learning: Linking efficacy and utility beliefs to effectiveness. In N. Perry & B. Kramarski (Co-chairs), Metacognition and self-regulation in developing professionals. Symposium presented at the biennial meeting of the European Association for Research on Learning and Instruction, Milan, Italy. Bronfenbrenner, U. (1978). The ecology of human development: Experiments by nature and design. Cambridge, MA: Harvard University Press.  Bronfenbrenner, U. (1994). Ecological models of human development. International Encyclopedia of Education (2nd ed., Vol. 3, pp. 1643-1647). Oxford, England: Elsevier Sciences. Brouwers, A., & Tomic, W. (1999). Teacher burnout, perceived self-efficacy in classroom management, and student disruptive behaviour in secondary education. Curriculum and Teaching, 14, 7-26. Brouwers, A., & Tomic, W. (2000). A longitudinal study of teacher burnout and perceived self-efficacy in classroom management. Teaching and Teacher Education, 16, 239-253. doi: 10.1016/S0742-051X(99)00057-8 Brown, T. A. (2006). Confirmatory factor analysis for applied research. New York, NY: The Guildford Press.  Bryk, A. S., & Schneider, B. (2003). Trust in schools: A core resource for school reform. Educational Leadership, 60, 40-44. Burke, R. J., & Greenglass, E. (1995). A longitudinal study of psychological burnout in teachers. Human Relations, 48, 187-202. doi: 10.1177/001872679504800205 210   Butler, R. (2007). Teachers' achievement goal orientations and associations with teachers' help seeking: Examination of a novel approach to teacher motivation. Journal of Educational Psychology, 99, 241-252. doi: 10.1037/0022-0663.99.2.241  Butler, R., & Shibaz, L. (2008). Achievement goals for teaching as predictors of students' perceptions of instructional practices and students' help seeking and cheating. Learning and Instruction, 18, 453-467. doi: 10.1016/j.learninstruc.2008.06.004  Butt, G., Lance, A., Fielding, A., Gunter, H., Rayner, S., & Thomas, H. (2005). Teacher job satisfaction: Lessons from the TSW pathfinder project. School Leadership and Management, 25, 455-471. doi: 10.1080/13634230500340807  Canrinus, E. T., Helms-Lorenz, M., Beijaard, D., Buitink, J., & Hofman, A. (2012). Self-efficacy, job satisfaction, motivation and commitment: Exploring the relationships between indicators of teachers? professional identity. European Journal of Psychology of Education, 27, 115-132. doi: 10.1007/s10212-011-0069-2 Caprara, G. V., Barbaranelli, C., Borgogni, L., & Steca, P. (2003). Efficacy beliefs as determinants of teachers' job satisfaction. Journal of Educational Psychology, 95, 821-832. doi: 10.1037/0022-0663.95.4.821 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, 473-490. doi: 10.1016/j.jsp.2006.09.001 Carleton, L. E., Fitch, J. C., & Krockover, G. H. (2008). An in-service teacher education program's effect on teacher efficacy and attitudes. Educational Forum, 72, 46-62. doi: 10.1080/00131720701603628 Chan, D. W. (1998). Stress, coping strategies, and psychological distress among secondary school teachers in Hong Kong. American Educational Research Journal, 35, 145-163. Chaplain, R. P. (2008). Stress and psychological distress among trainee secondary teachers in England. Educational Psychology, 28, 195-209. Cho, Y., & Shim, S. S. (2013). Predicting teachers? achievement goals for teaching: The role of perceived school goal structure and teachers? sense of efficacy. Teaching and Teacher Education, 32, 12-21. Clunies-Ross, P., Little, E., & Kienhuis, M. (2008). Self-reported and actual use of proactive and reactive classroom management strategies and their relationship with teacher stress and student behaviour. Educational Psychology, 28, 693-710. doi: 10.1080/01443410802206700  Collie, R. J. (2010). Social and emotional learning and school climate: Predictors of teacher stress, job satisfaction, and sense of efficacy (Unpublished master?s thesis). University of British Columbia, Vancouver, Canada. 211   Collie, R. J., Perry, N. E., & Brenner, C. A. (2013, August). Does stress have to be bad? Examining stress and coping among early career teachers. Paper presented at the annual meeting of the European Association for Research on Learning and Instruction, Munich, Germany. Collie, R. J., Shapka, J. D., & Perry, N. E. (2011). Predicting teacher commitment: The impact of school climate and social-emotional learning. Psychology in the Schools, 48, 1034-1048.  Collie, R. J., Shapka, J. D., & Perry, N. E. (2012). School climate and social?emotional learning: Predicting teacher stress, job satisfaction, and teaching efficacy. Journal of Educational Psychology, 104, 1189-1204. doi: 10.1037/a0029356 Cortina, J. M. (1993). What is coefficient alpha? An examination of theory and applications. Journal of Applied Psychology, 78, 98-104. Curran, P. J., Obeidat, K., & Losardo, D. (2010). Twelve frequently asked questions about growth curve modeling. Journal of Cognition and Development, 11, 121-136. doi:10.1080/15248371003699969  Darling-Hammond, L. (2000). Solving the demands of teacher supply, demand and standards: How we can ensure a competent, caring, and qualified teacher for every child. New York: National Commission on Teaching and America's Future.  Day, C., Elliot, B., & Kington, A. (2005). Reform, standards and teacher identity: Challenges of sustaining commitment. Teaching and Teacher Education, 21, 563-577. doi:10.1016/j.tate.2005.03.001  Deci, E. L., Eghrari, H., Patrick, B. C., & Leone, D. R. (1994). Facilitating internalisation: The self-determination theory perspective. Journal of Personality, 62, 119-142. Deci, E. L., & Ryan, R. M. (1985). Intrinsic motivation and self-determination in human behaviour. New York, NY: Plenum Press.  Deci, E. L., & Ryan, R. M. (Eds.). (2002). Handbook of self-determination research. Rochester, NY: University of Rochester Press.  Deci, E. L., & Ryan, R. M. (2008a). Facilitating optimal motivation and psychological well-being across life's domains. Canadian Psychology/Psychologie Canadienne, 49, 14-23. doi: 10.1037/0708-5591.49.1.14  Deci, E. L., & Ryan, R. M. (2008b). Hedonia, eudaimonia, and well-being: An introduction. Journal of Happiness Studies, 9, 1-11. doi: 10.1007/s10902-006-9018-1  Deci, E. L., & Ryan, R. M. (2008c). Self-determination theory: A macrotheory of human motivation, development, and health. Canadian Psychology/Psychologie Canadienne, 49, 182-185. doi: 10.1037/a0012801  212   Deci, E. L., & Ryan, R. M. (2012). Motivation, personality, and development within embedded social contexts: An overview of self-determination theory. In R. M. Ryan (Ed.), The oxford handbook of human motivation (pp. 85-110). New York, NY: Oxford University Press.  Deci, E. L., Ryan, R. M., Gagn?, M., Leone, D. R., Usunov, J., & Kornazheva, B. P. (2001). Need satisfaction, motivation, and well-being in the work organisations of a former eastern bloc country: A cross-cultural study of self-determination. Personality and Social Psychology Bulletin, 27, 930-942. doi: 10.1177/0146167201278002 Deci, E. L., Schwartz, A. J., Sheinman, L., & Ryan, R. M. (1981). An instrument to assess adults' orientations toward control versus autonomy with children: Reflections on intrinsic motivation and perceived competence. Journal of Educational Psychology, 73, 642-650. doi: 10.1037/0022-0663.73.5.642  Deci, E. L., Vallerand, R. J., Pelletier, L. G., & Ryan, R. M. (1991). Motivation and education: The self-determination perspective. Educational Psychologist, 26, 325.  Demerouti, E., Bakker, A. B., Nachreiner, F., & Schaufeli, W. B. (2001). The job demands-resources model of burnout. Journal of Applied Psychology, 86, 499-512. doi: 10.1037/0021-9010.86.3.499 Diener, E. (1984). Subjective well-being. Psychological Bulletin, 95, 542-575. doi: 10.1037/0033-2909.95.3.542  Diener, E. (2009). Conclusion?The well-being science needed now. Social Indicators Research Series, 37, 267-271. doi: 10.1007/978-90-481-2350-6 1 Diener, E., Emmons, R. A., Larsen, R. J., & Griffin, S. (1985). The satisfaction with life scale. Journal of Personality Assessment, 49, 71-75.  Diener, E., Suh, E. M., Lucas, R. E., & Smith, H. L. (1999). Subjective well-being: Three decades of progress. Psychological Bulletin, 125, 276-302. doi: 10.1037/0033-2909.125.2.276  Diener, E., Wirtz, D., Tov, W., Kim-Prieto, C., Choi, D., Oishi, S., & Biswas-Diener, R. (2010). New well-being measures: Short scales to assess flourishing and positive and negative feelings. Social Indicators Research, 97, 143-156. doi: 10.1007/s11205-009-9493-y Dinham, S., & Scott, C. (1998). A three domain model of teacher and school executive career satisfaction. Journal of Educational Administration, 36, 319-331.  Duckworth, A. L., Quinn, P. D., & Seligman, M. E. P. (2009). Positive predictors of teacher effectiveness. Journal of Positive Psychology, 4, 540-547. 213   Feather, N. T., & Rauter, K. A. (2004). Organizational citizenship behaviours in relation to job status, job insecurity, organizational commitment and identification, job satisfaction and work values. Journal of Occupational & Organizational Psychology, 77, 81-94.  Fernet, C., Guay, F., & Sen?cal, C. (2004). Adjusting to job demands: The role of work self-determination and job control in predicting burnout. Journal of Vocational Behavior, 65, 39-56. doi: 10.1016/S0001-8791(03)00098-8  Filak, V. F., & Sheldon, K. M. (2008). Teacher support, student motivation, student need satisfaction, and college teacher course evaluations: Testing a sequential path model. Educational Psychology, 28, 711-724. doi: 10.1080/01443410802337794 Fischer, D. G., & Fick, C. (1993). Measuring social desirability: Short forms of the Marlowe-Crowne Social Desirability Scale. Educational and Psychological Measurement, 53, 417-424. doi: 10.1177/0013164493053002011  Fortier, M. S., Vallerand, R. J., & Guay, F. (1995). Academic motivation and school performance: Toward a structural model. Contemporary Educational Psychology, 20, 257-274. Friedman, I. A., & Kass, E. (2002). Teacher self-efficacy: A classroom-organization conceptualization. Teaching and Teacher Education, 18, 675-686.  Gagn?, M., Chemolli, E., Forest, J., & Koestner, R. (2008). A temporal analysis of the relation between organisational commitment and work motivation. Psychologica Belgica, 48, 219-241. Gagn?, M., & Deci, E. L. (2005). Self-determination theory and work motivation. Journal of Organizational Behavior, 26, 331-362. doi: 10.1002/job.322  Gagn?, M., Forest, J., Gilbert, M., Aub?, C., Morin, E., & Malorni, A. (2010). The motivation at work scale: Validation evidence in two languages. Educational and Psychological Measurement, 70, 628-646. doi: 10.1177/0013164409355698 Gagn?, M., Ryan, R. M., & Bargmann, K. (2003). Autonomy support and need satisfaction in the motivation and well-being of gymnasts. Journal of Applied Sport Psychology, 15, 372-390. doi: 10.1080/714044203  Geving, A. M. (2007). Identifying the types of student and teacher behaviours associated with teacher stress. Teaching and Teacher Education, 23, 624-640. doi: 10.1016/j.tate.2007.02.006  Gibson, S., & Dembo, M. H. (1984). Teacher efficacy: A construct validation. Journal of Educational Psychology, 76, 569-582. doi:10.1037/0022-0663.76.4.569  214   Gillison, F., Osborn, M., Standage, M., & Skevington, S. (2009). Exploring the experience of introjected regulation for exercise across gender in adolescence. Psychology of Sport and Exercise, 10, 309-319. doi: 10.1016/j.psychsport.2008.10.004 Grayson, J. L., & Alvarez, H. K. (2008). School climate factors relating to teacher burnout: A mediator model. Teaching and Teacher Education, 24, 1349-1363. doi: 10.1016/j.tate.2007.06.005 Hakanen, J. J., Bakker, A. B., & Schaufeli, W. B. (2006). Burnout and work engagement among teachers. Journal of School Psychology, 43, 495-513. doi: 10.1016/j.jsp.2005.11.001 Hakanen, J. J., Schaufeli, W. B., & Ahola, K. (2008). The job demands-resources model: A three-year cross-lagged study of burnout, depression, commitment, and work engagement. Work & Stress, 22, 224-241. doi:10.1080/02678370802379432  Hamre, B. K., & Pianta, R. C. (2006). Student-teacher relationships. In G. G. Bear, & K. Minke (Eds.), Children's needs III: Development, prevention, and intervention (pp. 59-71). Bethesda, MD: NASP. Hayton, J. C., Allen, D. G., & Scarpello, V. (2004). Factor retention decisions in exploratory factor analysis: A tutorial on parallel analysis. Organizational Research Methods, 7, 191-205. doi: 10.1177/1094428104263675  Herzberg, F., Mausner, B., & Snyderman, B. (1959). The motivation to work. New York: John Wiley.  Hobfall, S. E. (1989). Conservation of resources: A new attempt at conceptualizing stress. American Psychologist, 44, 513-524.  Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling: A Multidisciplinary Journal, 6, 55. Huppert, F. A., & So, T. T. (2013). Flourishing across Europe: Application of a new conceptual framework for defining well-being. Social Indices Research, 110, 837-861. Ilardi, B. C., Leone, D., Kasser, T., & Ryan, R. M. (1993). Employee and supervisor ratings of motivation: Main effects and discrepancies associated with job satisfaction and adjustment in a factory setting. Journal of Applied Social Psychology, 23, 1789-1805. doi: 10.1111/j.1559-1816.1993.tb01066.x  Ingersoll, R. M. (2002). Holes in the teacher supply bucket. School Administrator, 59, 42-43.  Ingersoll, R. M., & May, H. (2012). The magnitude, destinations, and determinants of mathematics and science teacher turnover. Educational Evaluation and Policy Analysis, 34, 435-464. doi:10.3102/0162373712454326  215   Jackson, L. T. B., Rothmann, S., & van de Vijver (2006). A model of work-related well-being for educators in South Africa. Stress & Health: Journal of the International Society for the Investigation of Stress, 22, 263-274. doi: 10.1002/smi.1098  Jamal, M. (1990). Relationship of job stress and type-A behavior to employees' job satisfaction, organizational commitment, psychosomatic health problems, and turnover motivation. Human Relations, 43 727-738. doi: 10.1177/001872679004300802  Jang, H., Reeve, J., & Deci, E. L. (2010). Engaging students in learning activities: It is not autonomy support or structure but autonomy support and structure. Journal of Educational Psychology, 102, 588-600. doi: 10.1037/a0019682 Jennings, P. A., & Greenberg, M. T. (2009). The prosocial classroom: Teacher social and emotional competence in relation to student and classroom outcomes. Review of Educational Research, 79, 491-525. doi: 10.3102/0034654308325693 Jepson, E., & Forrest, S. (2006). Individual contributory factors in teacher stress: The role of achievement striving and occupational commitment. British Journal of Educational Psychology, 76, 183-197. doi: 10.1348/000709905X37299  Joinson, A. (1999). Social desirability, anonymity, and internet-based questionnaires. Behavior Research Methods, Instruments, and Computers, 31, 433-438.  Karabenick, S. A., Woolley, M. E., Friedel, J. M., Ammon, B. V., Blazevski, J., Rhee Bonney, C., ? & Kelly, K. L. (2007). Cognitive processing of self-report items in educational research: Do they think what we mean? Educational Psychologist, 42, 139-151. Klassen, R. M., Aldhafri, S., Mansfield, C., Purwanto, E., Siu, a. F. Y., Wong, M. W., Woods-McConnery, A. (2012). Teachers? engagement at work: An internationals validation study. The Journal of Experimental Education, 80, 317-337. doi: 10.1080/00220973.2012.678409 Klassen, R. M., Bong, M., Usher, E. L., Chong, W. H., Huan, V. S., Wong, I. Y. F., & Georgiou, T. (2009). Exploring the validity of a teachers? self-efficacy scale in five countries. Contemporary Educational Psychology, 34, 67-76. doi: 10.1016/j.cedpsych.2008.08.001 Klassen, R. M., & Chiu, M. M. (2010). Effects on teachers? self-efficacy and job satisfaction: Teacher gender, years of experience, and job stress. Journal of Educational Psychology, 102, 741-756. doi: 10.1037/a0019237 Klassen, R. M., & Chiu, M. M. (2011). The occupational commitment and intention to quit of practicing and pre-service teachers: Influence of self-efficacy, job stress, and teaching context. Contemporary Educational Psychology, 36, 114-129. doi: 10.1016/j.cedpsych.2011.01.002 216   Klassen, R. M., Perry, N. E., & Frenzel, A. C. (2012). Teachers' relatedness with students: An underemphasized component of teachers' basic psychological needs. Journal of Educational Psychology, 104, 150-165. doi: 10.1037/a0026253 Klassen, R. M., Tze, V. M. C., Betts, S. M., & Gordon, K. A. (2011). Teacher efficacy research 1998?2009: Signs of progress or unfulfilled promise? Educational Psychology Review, 23, 21-43. Klassen, R. M., Usher, E. L., & Bong, M. (2010). Teachers? collective efficacy, job satisfaction, and job stress in cross-cultural context. Journal of Experimental Education, 78, 464-486. doi: 10.1080/00220970903292975 Kline, R. B. (2011). Principles and practice of structural equation modeling (3rd ed.). New York, NY: The Guildford Press.  Klusmann, U., Kunter, M., Trautwein, U., L?dtke, O., & Baumert, J. (2008). Engagement and emotional exhaustion in teachers: Does the school context make a difference? Applied Psychology: An International Review, 57, 127-151. doi: 10.1111/j.1464-0597.2008.00358.x Konu, A., Viitanen, E., & Lintonen, T. (2010). Teachers' wellbeing and perceptions of leadership practices. International Journal of Workplace Health Management, 7, 44-57. doi: 10.1108/17538351011031939  Kyriacou, C. (2000). Stressbusting for teachers. London, England: Nelson Thornes. Kyriacou, C. (2001). Teacher stress: Directions for future research. Educational Review, 53, 27-35. doi: 10.1080/00131910120033628 Kyriacou, C., & Sutcliffe, J. (1977). The prevalence of stress among teachers in medium-sized mixed comprehensive schools. Research in Education, 18, 75-79.  Laursen, B., & Hoff, E. (2006). Person-centered and variable-centered approaches to longitudinal data. Merrill-Palmer Quarterly, 52, 377-389.  Lavian, R. H. (2012). The impact of organizational climate on burnout among homeroom teachers and special education teachers (full classes/individual pupils) in mainstream schools. Teachers and Teaching: Theory and Practice, 18, 233-247. Leiter, M. P., & Bakker, A. B. (2010). Work engagement: Introduction. In A. B. Bakker & M. P. Leiter (Eds.), Work engagement: A handbook of essential theory and research (pp. 1-9). New York, NY: Psychology Press. Lent, R. W., Nota, L., Soresi, S., Ginevra, M. C., Duffy, R. D., & Brown, S. D. (2011). Predicting the job and life satisfaction of Italian teachers: Test of a social cognitive model. Journal of Vocational Behavior, 79, 91-97. doi: 10.1016/j.jvb.2010.12.006  217   Linley, P. A., Maltby, J., Wood, A. M., Osborne, G., & Hurling, R. (2009). Measuring happiness: The higher order factor structure of subjective and psychological well-being measures. Personality and Individual Differences, 47, 878-884. doi: 10.1016/j.paid.2009.07.010  Locke, E. A. (1969). What is job satisfaction? Organizational Behavior and Human Performance, 4, 309-336.  Loo, R., & Thorpe, K. (2000). Confirmatory factor analyses of the full and short versions of the Marlowe-Crowne social desirability scale. Journal of Social Psychology, 140, 628-635.  Lynch, M. F. J., Plant, R. W., & Ryan, R. M. (2005). Psychological needs and threat to safety: Implications for staff and patients in a psychiatric hospital for youth. Professional Psychology: Research and Practice, 36, 415-425. doi: 10.1037/0735-7028.36.4.415 Lyubomirsky, S., Dickerhoof, R., Boehm, J. K., & Sheldon, K. M. (2011). Becoming happier takes both a will and a proper way: An experimental longitudinal intervention to boost well-being. Emotion, 11, 391-402. doi: 10.1037/a0022575 MacNeil, A. J., Prater, D. L., & Busch, S. (2009). The effects of school culture and climate on student achievement. International Journal of Leadership in Education, 12, 73-84. doi: 10.1080/13603120701576241  Manthei, R., & Gilmore, A., Tuck, B., & Adair, V. (1996). Teacher stress in intermediate schools. Educational Research, 38, 3-19. Martin, A. J. (2007). Examining a multidimensional model of student motivation and engagement using a construct validation approach. British Journal of Educational Psychology, 77, 413-440. doi:10.1348/000709906X118036  Martin, A. J., & Dowson, M. (2009). Interpersonal relationships, motivation, engagement, and achievement: Yields for theory, current issues, and educational practice. Review of Educational Research, 79, 327-365. doi: 10.3102/0034654308325583 Maslach, C., & Jackson, S. E. (1981). The measurement of experienced burnout. Journal of Occupational Behavior, 2, 99-113.  McCarthy, C. J., Lambert, R. C., O'Donnell, M., & Melendres, L. T. (2009). The relation of elementary teachers' experience, stress, and coping resources to burnout symptoms. Elementary School Journal, 109, 282-300.  Mertler, C. A. (2002). Job satisfaction and perception of motivation among middle and high school teachers. American Secondary Education, 31, 43-53.  218   Mertler, C. A. (2003). Patterns of response and nonresponse from teachers to traditional and web surveys. Practical Assessment, Research & Evaluation, 8. Retrieved from http://PAREonline.net/getvn.asp?v=8&n=22  Meyer, J. P., Becker, T. E., & Vandenberghe, C. (2004). Employee commitment and motivation: A conceptual analysis and integrative model. Journal of Applied Psychology, 89, 991-1007.  Meyer, J. P., & Herscovitch, L. (2001). Commitment in the workplace: Toward a general model. Human Resource Management Review, 11, 299-326. Meyer, J. P., & Maltin, E. R. (2010). Employee commitment and well-being: A critical review, theoretical framework and research agenda. Journal of Vocational Behavior, 77, 323-337. doi: 10.1016/j.jvb.2010.04.007  Milkie, M. A., & Warner, C. H. (2011). Classroom learning environments and the mental health of first grade children. Journal of Health and Social Behavior, 52, 4-22. Miller, L. A., McIntire, S. A., & Lovler, R. L. (2011). Foundations of psychological testing: A practical approach. Thousand Oaks, CA: SAGE Publications.  Ministry of Education (2001). BC Performance Standards: Social Responsibility ? A Framework. Retrieved from http://www.bced.gov.bc.ca/perf_stands/social_resp.htm Ministry of Education (2012). School calendar regulation. Retrieved from http://www.bced.gov.bc.ca/legislation/schoollaw/d/bcreg_114-02.pdf  Mo?, A., Pazzaglia, F., & Ronconi, L. (2010). When being able is not enough. the combined value of positive affect and self-efficacy for job satisfaction in teaching. Teaching and Teacher Education, 26, 1145-1153. doi: 10.1016/j.tate.2010.02.010  Morin, J., & Battalio, R. (2004). Construing misbehaviour: The efficacy connection in responding to misbehaviour. Journal of Positive Behaviour Interventions, 6, 251-254.kay Muth?n, L. K. & Muth?n, B. O. (1998-2012). Mplus User?s Guide. Seventh Edition. Los Angeles, CA: Muth?n & Muth?n. Muth?n, L. K. & Muth?n, B. O. (2005) Chi-square difference testing using the Satorra-Bentler scaled chi-square. Retrieved from http://www.statmodel.com/chidiff.shtml Naylor, C. (2001). Teacher workload and stress: An international perspective on human costs and systemic failure. Vancouver, BC: British Columbia Teachers' Federation.  Pakarinen, E., Kiuru, N., Lerkkanen, M., Poikkeus, A., Siekkien, M., & Nurmi, J. (2010). Classroom organisation and teacher stress predict learning motivation in kindergarten children. European Journal of Psychology of Education, 25, 281-300. 219   Pas, E. T., Bradshaw, C. P., & Hershfeldt, P. A. (2012). Teacher- and school-level predictors of teacher efficacy and burnout: Identifying potential areas for support. Journal of School Psychology, 50, 129-145. doi: 10.1016/j.jsp.2011.07.003  Patrick, B. C., Hisley, J., Kempler, T., & College, G. (2000). `What's everybody so excited about?': The effects of teacher enthusiasm on student intrinsic motivation and vitality. Journal of Experimental Education, 68, 217.  Pearson, L. C., & Moomaw, W. (2005). The relationship between teacher autonomy and stress, work satisfaction, empowerment, and professionalism. Educational Research Quarterly, 29, 38-54. Pelletier, L. G., Fortier, M. S., Vallerand, R. J., & Bri?re, N. M. (2001). Associations among perceived autonomy support, forms of self-regulation, and persistence: A prospective study. Motivation & Emotion, 25, 279-306.  Perry, N. E., & Winne, P. H. (2006). Learning from learning kits: gStudy traces of students? self-regulated engagements with computerized content. Educational Psychology Review, 18, 211-228. Pintrich, P. R. (2000). Educational psychological at the millennium: A look back and a look forward. Educational Psychologist, 35, 221-226. Podsakoff, P. M., MacKenzie, S. B., Lee, J., & Podsakoff, N. P. (2003). Common method biases in behavioral research: A critical review of the literature and recommended remedies. Journal of Applied Psychology, 88, 879-903. doi:10.1037/0021-9010.88.5.879  Rath, T., & Harter, J. (2010). Well being: The five essential elements. New York, NY: Gallup Press. Reeve, J., Jang, H., Carrell, D., Jeon, S., & Barch, J. (2004). Enhancing students? engagement by increasing teachers? autonomy support. Motivation and Emotion, 28, 147-169. Reis, H. T., Sheldon, K. M., Gable, S. L., Roscoe, J., & Ryan, R. M. (2000). Daily well-being: The role of autonomy, competence, and relatedness. Personality and Social Psychology Bulletin, 26, 419-435. doi: 10.1177/0146167200266002  Retelsdorf, J., Butler, R., Streblow, L., & Schiefele, U. (2010). Teachers' goal orientations for teaching: Associations with instructional practices, interest in teaching, and burnout. Learning and Instruction, 20, 30-46. doi: 10.1016/j.learninstruc.2009.01.001 Richardson, P. W., & Watt, H. M. G. (2010). Current and future directions in teacher motivation research. In T. C. Urdan, & S. A. Karabenick (Eds.), The decade ahead: Applications and contexts of motivation and achievement (pp. 139-173). Emerald Group Publishing Limited.  220   Roeser, R. W., Schonert-Reichl, K., Jha, A., Cullen, M., Wallace, L., Wilensky, R., ... & Harrison, J. (2013). Mindfulness training and reductions in teacher stress and burnout: Results from two randomized, waitlist-control field trials. Journal of Educational Psychology, 105, 787-804. doi: 10.1037/a0032093 Ronfeldt, M., Loeb, S., & Wyckoff, J. (2013). How teacher turnover harms student achievement. American Educational Research Journal, 50, 4-36. doi:10.3102/0002831212463813  Ross, J. A. (1994). The impact of an inservice to promote cooperative learning on the stability of teacher efficacy. Teaching and Teacher Education, 10, 381-394. doi:10.1016/0742-051X(94)90020-5  Roth, G., Assor, A., Kanat-Maymon, Y., & Kaplan, H. (2007). Autonomous motivation for teaching: How self-determined teaching may lead to self-determined learning. Journal of Educational Psychology, 99, 762-774. doi: 10.1037/0022-0663.99.4.761 Ryan, R. M., & Connell, J. P. (1989). Perceived locus of causality and internalization: Examining reasons for acting in two domains. Journal of Personality and Social Psychology, 57, 749-761. doi: 10.1037/0022-3514.57.5.749  Ryan, R. M., & Deci, E. L. (2000). Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being. American Psychologist, 55, 68-78. doi: 10.1037/0003-066X.55.1.68  Ryan, R. M., & Deci, E. L. (2001). On happiness and human potentials: A review of research on hedonic and eudaimonic well-being. Annual Review of Psychology, 52, 141-166.  Ryan, R. M., & Deci, E. L. (2011). A self-determination theory perspective on social, institutional, cultural, and economic supports for autonomy and their importance for well-being. In V. I. Chirkov, R. M. Ryan & K. M. Sheldon (Eds.), Human autonomy in cross-cultural context (pp. 45-64). Netherlands: Springer. doi: 10.1007/978-90-481-9667-8_3 Ryff, C. D. (1989). Happiness is everything, or is it? explorations on the meaning of psychological well-being. Journal of Personality and Social Psychology, 57, 1069-1081. doi: 10.1037/0022-3514.57.6.1069  Schaufeli, W. B., & Bakker, A. B. (2004). Job demands, job resources, and their relationship with burnout and engagement: A multi-sample study. Journal of Organizational Behavior, 25, 293-315. doi: 10.1002/job.248  Schimmack, U., & Lucas, R. E. (2010). Environmental influences on well-being: A dyadic latent panel analysis of spousal similarity. Social Indices Research, 98, 1-21. doi: 10.1007/s11205-009-9516-8 221   Schleicher, D. J., Hansen, S. D., & Fox, K. E. (2011). Job attitudes and work values. In S. Zedeck, & et al. (Eds.), APA handbook of industrial and organizational psychology: Vol. 3. maintaining, expanding, and contracting the organization (pp. 137-189). Washington, DC: American Psychological Association.  Seligman, M. E. P. (2012). Flourish: A visionary new understanding of happiness and well-being. New York, NY: Free Press.  Selye, H. (1974). Stress without distress. Philadelphia: Lippincott.  Shann, M. H. (1998). Professional commitment and satisfaction among teachers in urban middle schools. Journal of Educational Research, 92, 67-73.  Simms, L. J. (2008). Classical and modern methods of psychological scale construction. Social and Personality Psychology Compass, 2, 414-433.  Sinclair, C. (2008). Initial and changing student teacher motivation and commitment to teaching. Asia-Pacific Journal of Teacher Education, 36, 79-104. Skaalvik, E. M., & Skaalvik, S. (2009). Does school context matter? Relations with teacher burnout and job satisfaction. Teaching and Teacher Education, 25, 518-524. doi: 10.1016/j.tate.2008.12.006 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, 1038. doi: 10.1016/j.tate.2011.04.001  Skinner, E., & Edge, K. (2002). Self-determination, coping, and development. In E. L. Deci, & R. M. Ryan (Eds.), Handbook of self-determination research (pp. 297-337). Rochester, NY: University of Rochester Press.  Solinger, O. N., van Olffen, W., & Roe, R. A. (2008). Beyond the three-component model of organizational commitment. Journal of Applied Psychology, 93, 70-83. doi: 10.1037/0021-9010.93.1.70  Spector, P. E. (1997). Job satisfaction: Application, assessment, causes and consequences. London: Sage Publications.  Spector, P. E., & Brannick, M. T. (2009) Common method variance or measurement bias? The problem and possible solutions. In D. A. Buchanan & A. Bryman (Eds.), The Sage Handbook of Organization Research Methods (346-362). London, UK: Sage Publications. Standage, M., Duda, J. L., & Ntoumanis, N. (2005). A test of self-determination theory in school physical education. The British Journal of Educational Psychology, 75, 411-433.  222   Steinhardt, M. A., Smith Jaggers, S. E., Faulk, K. E., & Gloria, C. T. (2011). Chronic work stress and depressive symptoms: Assessing the mediating role of teacher burnout. Stress and Health, 27, 420-429. Suh, E., Diener, E., & Fujita, F. (1996). Events and subjective well-being: Only recent events matter. Journal of Personality and Social Psychology,70, 1091-1102. doi: 10.1037/0022-3514.70.5.1091 Tabachnick, B., & Fidell, L. (2007). Using multivariate statistics (5th ed.). Boston, MA: Pearson Education.  Taylor, I. M., & Ntoumanis, N. (2007). Teacher motivational strategies and student self-determination in physical education. Journal of Educational Psychology, 99, 747-760. doi: 10.1037/0022-0663.99.4.747  Taylor, I. M., Ntoumanis, N., & Standage, M. (2008). A self-determination theory approach to understanding the antecedents of teachers? motivational strategies in physical education. Journal of Sport and Exercise Psychology, 30, 75-94.  Thomas, N., Clark, V., & Lavery, J. (2003). Self-reported work and family stress of female primary teachers. Australian Journal of Education, 47, 73-87.  Thoresen, C. J., Kaplan, S. A., Barsky, A. P., Warren, C. R., & de Chermont, K. (2003). The affective underpinnings of job perceptions and attitudes: A meta-analytic review and integration. Psychological Bulletin, 129, 914-945. doi: 10.1037/0033-2909.129.6.914  Thornton, B., Peltier, G., Medina, R. (2007). Reducing the special education teacher shortage. The Clearing House, 80, 233-238. Tov, W., & Diener, E. (2009). The well-being of nations: Linking together trust, cooperation, and democracy. In E. Diener (Ed.), The science of well-being: The collected works of Ed Diener, Social Indicators Research Series, 37, 155-173. doi: 10.1007/978-90-481-2350-6 7 Tschannen-Moran, M., & Woolfolk Hoy, A. E. (2001). Teacher efficacy: Capturing an elusive construct. Teaching and Teacher Education, 17, 783-805. doi: 10.1016/S0742-051X(01)00036-1 Tschannen-Moran, M., Woolfolk Hoy, A. E., & Hoy, W. K. (1998). Teacher efficacy: Its meaning and measure. Review of Educational Research, 68, 202-248.  Uziel, L. (2010). Rethinking social desirability scales: From impression management to interpersonally oriented self-control. Perspectives on Psychological Science, 5, 243-262. doi:10.1177/1745691610369465  Vallerand, R. J., Fortier, M. S., & Guay, F. (1997). Self-determination and persistence in a real-life setting: Toward a motivational model of high school dropout. Journal of Personality and Social Psychology, 72, 1161-1176. 223   Van den Broeck, A., Vansteenkiste, M., De Witte, H., Soenens, B., & Lens, W. (2010). Capturing autonomy, competence, and relatedness at work: Construction and initial validation of the work-related basic need satisfaction scale. Journal of Occupational and Organisational Psychology, 83, 981-1002. doi: 10.1348/096317909X481382 Van Horn, J. E., Schaufeli, W. B., & Taris, T. W. (2001). Lack of reciprocity among Dutch teachers: Validation of reciprocity indices and their relation to stress and well-being. Work & Stress, 15, 191-213. doi: 10.1080/02678370110066571  Van Horn, J. E., Taris, T. W., Schaufeli, W. B., & Schreurs, P. J. G. (2004). The structure of occupational well-being: A study among Dutch teachers. Journal of Occupational & Organizational Psychology, 77, 365-375.  Vandenberghe, C., & Bentein, K. (2009). A closer look at the relationship between affective commitment to supervisors and organisations and turnover. Journal of Occupational & Organisational Psychology, 82, 331-348. Vandenberghe, C., Bentein, K., Michon, R., Chebat, J., Tremblay, M., & Fils, J. (2007). An examination of the role of perceived support and employee commitment in employee-customer encounters. Journal of Applied Psychology, 92, 1177-1187. doi: 10.1037/0021-9010.92.4.1177  Velicer, W. F., Eaton, C. A., & Fava, J. L. (2000). Construct explication through factor of component analysis: A review and evaluation of alternative procedures for determining the number of factors or components. In Goffin, R. D., & Helmes, E. (Eds.), Problems and solutions in human assessment: Honoring Douglas N. Jackson at seventy (pp. 41-71). Boston, MA: Kluwer. Vivoll Straume, L. V., & Vitters?, J. (2012). Happiness, inspiration and the fully functioning person: Separating hedonic and eudaimonic well-being in the workplace. The Journal of Positive Psychology, 7, 387-398. doi: 10.1080/17439760.2012.711348  Watt, H. M. G., & Richardson, P. W. (2007). Motivational factors influencing teaching as a career choice: Development and validation of the FIT-choice scale. Journal of Experimental Education, 75, 167-202. doi: 10.3200/JEXE.75.3.167-202  Wei, H., & Chen, J. (2010). School attachment among Taiwanese adolescents: the roles of individual characteristics, peer relationships, and teacher well-being. Social Indicators Research, 95, 241-436. doi: 10.1007/s11205-009-9529-3  Weiqi, C. (2007). The structure of secondary school teacher job satisfaction and its relationship with attrition and work enthusiasm. Chinese Education and Society, 40, 17-31. doi: 10.2753/CED 1061-1932400503  Winne, P. H., Jamieson-Noel, D. L., & Muis, K. (2002). Methodological issues and advances in researching tactics, strategies, and self-regulated learning. In P. R. Pintrich and M. L. 224   Maehr (Eds.), Advances in motivation and achievement: New directions in measures and methods (Vol. 12, pp. 121-155). Greenwich, CT: JAI Press. Wolpin, J., Burke, R. J., & Greenglass, E. R. (1991). Is job satisfaction an antecedent or a consequence of psychological burnout? Human Relations, 44, 193-209. doi: 10.1177/001872679104400205 Woolfolk Hoy, A. E., & Spero, R. B. (2005). Changes in teacher efficacy during the early years of teaching: A comparison of four measures. Teaching and Teacher Education, 21, 343-356. doi: 10.1016/j.tate.2005.01.007      225   APPENDICES  Appendix A: Invitation Letter for Studies 1 and 2          Dear Teachers,  Did you know your well-being as a teacher is important for you and your students? Would you like to help us learn more about teacher well-being with the aim of improving working conditions for teachers? Do you want to win an iPad?  My name is Rebecca Collie and I am a doctoral student at the University of British Columbia. Under the supervision of Dr. Jennifer Shapka, I am currently conducting an online research study entitled, ?Well-Being and its Influence on Teachers,? and I am writing to request your professional assistance in this research. Teacher well-being is not only critically important for teachers' health and teaching effectiveness, but also the motivation and achievement of students. Despite the clear importance of this issue, relatively little is known about what aspects of teaching work impact teachers? well-being. My research study aims to uncover more about this with the long-term aim of helping to improve the profession for teachers and students alike.   With this email, I am inviting you to participate in my research. Your participation will involve only 15-20 minutes of your time by completing an online questionnaire. If you agree to participate, you can access the questionnaire at this link: [URL]  How does this research benefit teachers? By participating, you will be helping to provide important information for how best to support your well-being. It is my hope that this research will help to provide tangible strategies for schools, districts, and policy-makers to help improve teachers? experiences of well-being. I hope that these, in turn, will also help to improve the experiences of students, which are clearly linked with teacher well-being. The results of the study will be made available to your school district (without identifying any schools or individuals) with the aim of promoting working conditions that support teacher well-being.  Department of Educational and Counselling Psychology, and Special Education The University of British Columbia Faculty of Education  2125 Main Mall Vancouver BC Canada V6T 1Z4 Tel  604-822-0242  Fax  604-822-3302 www.ecps.educ.ubc.ca 226   Please be assured that your responses will be anonymous, and that your participation in this study is voluntary. To ensure confidentiality, your completed survey will be stored on a secure and password protected server in the Lower Mainland of British Columbia. Please be assured that your decision to participate or not participate in this study will have no impact on your relationship with the [NAME OF TEACHERS? ASSOCIATION] or your job standing. In addition, your teachers? association will not know who did/did not participate in the research. By completing and submitting the survey, you are giving your consent to participate. It is recommended that you keep a copy of this letter for your own records. If you do not wish to participate, simply disregard this message. If you have any questions regarding the study, please contact me at [EMAIL].   Let me thank you in advance for your valuable time and assistance in this study, which aims to advance our understanding of teacher well-being. As a thank you for your interest in our study, click below and enter your name in a draw to win an iPad. Your eligibility to enter the draw is not linked to your participation. [URL]  Please note: If you have any questions or concerns about this questionnaire, please contact Dr. Jennifer Shapka, the principal investigator and the supervisor of this study (CONTACT DETAILS). You may also contact the Research Subject Information Line in the UBC Office of Research Services at [PHONE NUMBER].  Best of luck in the remainder of the school year.  Sincerely,  Rebecca Collie      227   Appendix B: Questionnaire for Studies 1 and 2  Section 1: Demographics 1. What school level are you teaching at?   Elementary   Middle   Secondary   Both Elementary and Middle   Both Elementary and Secondary   Other:    2. What is your current position?   Pre-service teacher   Regular teacher   Resource teacher   Special education teacher    Administrator (e.g., principal, vice- principal, director, school head)   Teacher librarian   Teacher counsellor    Other:     3. How much of your time at work is spent teaching students or planning for teaching in a week?   0%    1-25%    26-50%    51-75%    76-100%   4. What are the major subjects that you teach?   5. What is your highest degree?   6. What is your age?   7. What is your gender?   F   M  8. What is the zipcode(s) of the school(s) you teach at?   9. What was your country of birth?    10. What is your ethnicity?   Northern and Western European origins (e.g., British, Scottish, German, Swedish, Danish, Norwegian, Dutch)   Eastern and Southern European origins (e.g., Polish, Russian, Ukranian, Italian, Greek, Spanish)   Aboriginal origins (e.g., First Nations, Inuit, Metis)   South Asian origins (e.g., East Indian, Punjabi, Pakistani)   East Asian origins (e.g., Chinese, Japanese, Korean)   Southeast Asian origins (e.g., Filipino, Thai, Vietnamese)   Other (please list):    11. How many years have you been teaching (including as a TOC)?     228   12. What is the socio-economic status of students in your school  (e.g., average family income level compared to most people in your province)? 1 Low 2 3 Average 4 5 High       13. What is the academic ability of students in your school  (e.g., average ability compared to other schools in your district/province)? 1 Low 2 3 Average 4 5 High       14. What is your school location? Urban Suburban Rural Small Town       Section 2: Flourishing, Autonomy Support, and Commitment Please indicate the degree to which you agree with the following statements currently in your life.   1 Strongly Disagree 2  3  4 Mixed or Neither Agree nor Disagree 5  6  7 Strongly Agree 1. My principal listens to how I would like to do things        2. I feel that my principal provides me with choices and options        3. I feel understood by my principal        4. My principal conveys confidence in my ability to do well at my job        5. My principal encourages me to ask questions        6. My principal tries to understand how I see things before suggesting a new way of doing things        7. I lead a purposeful and meaningful life        8. My social relationships are supportive and rewarding        9. I am engaged and interested in my daily activities        10. I actively contribute to the happiness and well-being of others        11. I am competent and capable in the activities that are important to me        12. I am a good person and live a good life        13. I am optimistic about my future        14. People respect me        229   15. This school has a great deal of personal meaning for me        16. I really feel a sense of 'belonging' to the school(s) at which I work        17. I am proud to belong to this school community        18. I do not feel emotionally attached to my school        19. I really feel as if my school?s problems are my own        20. I do not feel like 'part of the family at my school           Section 3: Need Satisfaction The following statements aim to tap your personal experiences at work. Would you please indicate the degree to which you currently agree with these statements?    1 Totally disagree 2 Disagree 3 Somewhat disagree / somewhat agree 4 Agree 5 Totally agree 1. I don?t really feel connected with other people at my job      2. At work, I feel part of a group      3. I don?t really mix with other people at my job      4. At work, I can talk with people about things that really matter to me      5. I often feel alone when I am with my colleagues      6. Some people I work with are close friends of mine      7. I don?t really feel competent in my job.      8. I really master my tasks at my job      9. I feel competent at my job      10. I doubt whether I am able to execute my job properly      11. I am good at the things I do in my job      12. I have the feeling that I can even accomplish the most difficult tasks at work      13. I feel like I can be myself at my job      14. At work, I often feel like I have to follow other people?s commands      15. If I could choose, I would do things at work differently      16. The tasks I have to do at work are in line with what I really want to do      230   17. I feel free to do my job the way I think it could best be done      18. In my job, I feel forced to do things I do not want to do      19. I am very committed to my students      20. Connecting with my students is an essential part of the job      21. I value the relationships I build with my students      22. I feel connected to my students        Section 4: Motivation at Work Using the scales below, please indicate for each of the following statements to what degree they presently correspond to one of the reasons for which you are doing this specific job.  Currently, why are you doing your current job?   1 Not at all 2 Very little 3 A little 4 Moderately 5 Strongly 6 Very strongly 7 Exactly 1. Because I enjoy this work very much        2. Because I have fun doing my job        3. For the moments of pleasure that this job brings me        4. I chose this job because it allows me to reach my life goals        5. Because this job fulfils my career plans        6. Because this job fits my personal values        7. Because I have to be the best in my job, I have to be a ?winner?        8. Because my work is my life and I don?t want to fail        9. Because my reputation depends on it        10. Because this job affords me a certain standard of living        11. Because it allows me to make a lot of money        12. I do this job for the paycheque         231    Section 5: Job Satisfaction, Stress, and Social Desirability Please indicate the degree to which the following statements are part of your experiences with your job.   1 Never 3 Almost never 3 Rarely 4 Sometimes 5 Often 6 Very often 7 Always 1. I am satisfied with my job        2. I am happy with the way my colleagues and superiors treat me        3. I am satisfied with what I achieve at work        4. I feel good at work         5. In general, how stressful do you find being a teacher? 0 Not at all stressful 1 Mildly stressful 2 Moderately stressful 3 Very stressful 4 Extremely stressful       Listed below are a number of statements concerning personal attitudes and traits. Read each item and decide whether the statement is true or false as it pertains to you personally.   True False 6. I like to gossip at times   7. There have been occasions when I took advantage of someone   8. I?m always willing to admit it when I make a mistake   9. I sometimes try to get even rather than forgive and forget   10. At time I have really insisted on having things my own way   11. I have never been irked when people expressed ideas very different from my own   12. I have never deliberately said something that hurt someone?s feelings      232    Section 6: Teacher Well-Being Currently, how do the following aspects of being a teacher affect your well-being as a teacher? Well-being refers to open, engaged, and healthy functioning as a teacher. For example, in your current working environment how do your relations with fellow teachers influence your well-being? Do your relations have a positive or negative influence on your well-being as a teacher?  If the current job action means that you are not currently engaging in some of the different items below, please answer based on your experiences prior to job action, or if you are a beginning teacher, please leave these items blank.    1 Negatively 2 Mostly Negatively 3 More Negatively than Positively 4 Neither Positively  nor Negatively 5 More Positively than Negatively 6 Mostly Positively 7 Positively 1. Relations with fellow teachers        2. Relations with students in my class        3. Relations with students not in my class (e.g., in hallways, etc.)        4. Relations with leadership at my school        5. Relations with administrators at my school        6. Relations with my students? parents        7. Relations with my own family at home        8. Marking work        9. Lesson planning        10. Knowledge of subject-matter        11. Interruptions to my teaching in my classroom (e.g., from people from outside of my class)        12. Fitting everything in to the allotted time (e.g., required curriculum, extra-curricular activities)        13. Implementing educational        233   innovations or new programs 14. Using technology for administrative work (e.g., report cards, class lists)        15. Using technology for teaching (e.g., the Internet, electronic whiteboards, etc.)        16. The challenge of teaching work (e.g., being intellectually challenged)        17. The nature of teaching work (e.g., the meaningfulness of the job)        18. Satisfaction with my current job        19. Meeting learning goals         20. Student behaviour        21. Student motivation        22. My salary and benefits        23. Responsibility for student success        24. Rest periods/breaks during the day        25. Noise(s) around the classroom/school        26. Administrative work related to teaching        27. Classroom management        28. Expectations of parents        29. Expectations of leadership or district        30. Support offered by parents        31. Support offered by leadership        32. The amount of time to spend with individual students        234   33. The availability of equipment, facilities, or materials for teaching        34. Number of students in my class(es)        35. Recognition for my teaching        36. Opportunities for promotion        37. School rules and procedures that are in place        38. Follow-through by staff and leadership of school rules        39. Communication between members of the school        40. Work I complete outside of school hours for teaching        41. The community or neighbourhood in which I live        42. My commute to school        43. My physical health        44. The state of affairs in the world        45. Working to finish my teaching tasks        46. The ability to control what/how I teach         47. Instances of bullying between students and/or staff        48. Student/staff use of social media/Internet        49. Using the proscribed curriculum and syllabi        50. Participation in school-level decision making        235   51. The health of my family and/or friends        52. Uncontrolled events outside of school (e.g., car breaking down)        53. My or my families extracurricular activities outside of school (e.g., going to the gym, taking children to activities)        54. Looking after family members        55. Inclusion of students with special needs        56. Availability of support for students with special needs        57. Staying late after work for meetings and activities        58. Standardised testing (e.g., FSA)        59. Expectations of the provincial government         60. Communication between all members of my school          236   Appendix C: Invitation Letter for Study 3     Dear Teachers,  In the next few weeks, you will be participating in the YouthMADE curriculum?a program that uses short films made by youth to explore issues of racism and discrimination. At the same time, you are being invited to participate in a UBC-based evaluation research project about YouthMADE, which aims to examine the impact of the YouthMADE curriculum on your well-being and development. This research is being conducted by Dr. Jennifer Shapka, who is an Associate Professor in the Department of Educational and Counselling Psychology, and Special Education at the University of British Columbia (her contact information is below).   To participate, all you will need to do is fill out an online questionnaire before you take part in the YouthMADE workshops. You can fill it out right now?details for accessing it are in the accompanying email. This questionnaire will ask questions about your well-being, as well as how you feel about your school and your experiences as a teacher. Demographic questions such as age, gender and ethnicity will also be asked. It will take about 15 minutes to complete. If you consent, please complete the first online questionnaire as honestly and openly as you can.   After you have the YouthMADE curriculum in your classroom, we will contact you with details about two follow-up online questionnaires. These questionnaires will be spread over two months and will include similar questions to the first questionnaire.   There are no known risks for your participation in this study, and you have the right to withdraw from the study at any time. Also, the study is voluntary and it will not impact your job security. Please be assured that your responses will be kept confidential.   Contact for information about the study: If you have any questions about this study, please contact Dr. Jennifer Shapka at PHONE NUMBER, or her research associate, Rebecca Collie at PHONE NUMBER.  Contact for concerns about the rights of research subjects: If you have any concerns about your child?s treatment or rights as a research subject, you may contact the Research Subject Information Line in the UBC Office of Research Services at PHONE NUMBER.  Thank you for your time and support!  Sincerely, Jennifer Shapka (on behalf of the YouthMADE research team)    Department of Educational and Counselling Psychology, and Special Education The University of British Columbia Faculty of Education  2125 Main Mall Vancouver BC Canada V6T 1Z4 Tel  604-822-0242  Fax  604-822-3302 www.ecps.educ.ubc.ca 237   Appendix D: Questionnaire for Study 3  Section 1: Demographics  1. What is your age?     2. What is your gender?   F   M   3. What is your ethnicity?   Northern and Western European origins (e.g., British, Scottish, German, Swedish, Danish, Norwegian, Dutch)   Eastern and Southern European origins (e.g., Polish, Russian, Ukranian, Italian, Greek, Spanish)   Aboriginal origins (e.g., First Nations, Inuit, Metis)   South Asian origins (e.g., East Indian, Punjabi, Pakistani)   East Asian origins (e.g., Chinese, Japanese, Korean)   Southeast Asian origins (e.g., Filipino, Thai, Vietnamese)   Other (please list):    4. Do you consider yourself to be a member of a minority group (e.g., based on race, gender, ethnicity, class, ability, and/or sexual orientation)?   Yes   No   I don?t know / No answer  5. What grades do you teach? Check all that apply.   Kinder   Grade 1   Grade 2   Grade 3   Grade 4    Grade 5   Grade 6   Grade 7   Grade 8    Grade 9   Grade 10   Grade 11   Grade 12   6. What is your current position?   Pre-service teacher   Regular teacher   Resource teacher   Special education teacher    Administrator (e.g., principal, vice- principal, director, school head)   Teacher librarian   Teacher counsellor    Other:     7. How many years have you been teaching (including as a TOC)?      238   Section 2: Well-being Currently, how do the following aspects of being a teacher affect your well-being as a teacher? Well-being refers to open, engaged, and healthy functioning as a teacher. For example, in your current working environment how do your relations with fellow teachers influence your well-being? Do your relations have a positive or negative influence on your well-being as a teacher?   If the current job action means that you are not currently engaging in some of the different items below, please answer based on your experiences prior to job action, or if you are a beginning teacher, please leave these items blank.   1 Negatively 2 Mostly Negatively 3 More Negatively than Positively 4 Neither Positively  nor Negatively 5 More Positively than Negatively 6 Mostly Positively 7 Positively 1. Relations with students in my class        2. Relations with administrators at my school        3. Marking work        4. Fitting everything in to the allotted time (e.g., required curriculum, extra-curricular activities)        5. Student behaviour        6. Student motivation        7. Administrative work related to teaching        8. Classroom management        9. Support offered by leadership        10. Recognition for my teaching        11. School rules and procedures that are in place        12. Communication between members of the school        13. Work I complete outside of school hours for teaching        239   14. Working to finish my teaching tasks        15. Participation in school-level decision making        16. Staying late after work for meetings and activities         Section 3: Need Satisfaction The following statements aim to tap your personal experiences at work. Would you please indicate the degree to which you currently agree with these statements?    1 Totally disagree 2 Disagree 3 Somewhat disagree / somewhat agree 4 Agree 5 Totally agree 1. I don?t really feel connected with other people at my job      2. At work, I feel part of a group      3. I don?t really mix with other people at my job      4. At work, I can talk with people about things that really matter to me      5. I often feel alone when I am with my colleagues      6. Some people I work with are close friends of mine      7. I don?t really feel competent in my job.      8. I really master my tasks at my job      9. I feel competent at my job      10. I doubt whether I am able to execute my job properly      11. I am good at the things I do in my job      12. I have the feeling that I can even accomplish the most difficult tasks at work      13. I feel like I can be myself at my job      14. At work, I often feel like I have to follow other people?s commands      15. If I could choose, I would do things at work differently      16. The tasks I have to do at work are in line with what I really want to do      17. I feel free to do my job the way I think it could best be done      18. In my job, I feel forced to do things I do not want to do      19. I am very committed to my students      240   20. Connecting with my students is an essential part of the job      21. At work, I often feel like I have to follow other peoples? commands      22. I feel connected to my students        Section 4: Self-efficacy for Teaching  Please indicate your opinion about each of the questions below by marking any one of the nine responses in the columns on the right side, ranging from (1) ?Nothing? to (9) ?A Great Deal?. Please respond to each of the questions by considering the combination of your current ability, resources, and opportunity to do each of the following in your teaching position.     1 Nothing 2  3 Very Little 4 5 Some Degree 6 7 Quite a Bit 8 9 A Great Deal 1. How much can you do to control disruptive behaviour in the classroom?          2. How much can you do to motivate students who show low interest in school work?          3. How much can you do to calm a student who is disruptive or noisy?          4. How much can you do to help your students value learning?          5. To what extent can you craft good questions for your students?          6. How much can you do to get children to follow classroom rules?          7. How much can you do to get students to believe they can do well in school work?          8. How well can you establish a classroom management system with each group of students?          9. To what extent can you use a variety of assessment strategies?          10. To what extent can you provide an alternative explanation or example when students are confused?          11. How much can you assist families in helping their children do well in school?          12. How well can you implement alternative teaching strategies in your classroom?           241  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0165878/manifest

Comment

Related Items