Self-efficacy as a Predictor of Commitment to the Teaching Profession_a Meta-Analysis

Review

Self-efficacy as a predictor of commitment to the teachingprofession: A meta-analysisSteven Randall Chesnut *, Hansel BurleyTexas Tech University, Lubbock, Texas 79409-1071, USA

A R T I C L E I N F O

Article history:Received 8 May 2014Received in revised form 14 January 2015Accepted 14 February 2015Available online 28 February 2015

Keywords:Self-efficacyCommitment to teachingMeta-analysis

A B S T R A C T

This meta-analysis examined research on the effects of preservice and inservice teachers’self-efficacy beliefs on commitment to the teaching profession. Unlike previous studies onself-efficacy and commitment, this review systematically examines the effects found withinthe literature and highlights important theoretical and methodological issues. A total of33 qualified studies were included in the final analysis, including 16,122 preservice andinservice teachers. Findings suggest that preservice and inservice teachers’ self-efficacy beliefsinfluence their commitment to the teaching profession (ES = +0.32). However, these effectsvary based upon the conceptual accuracy of the self-efficacy measure and the origin of data.Conceptually accurate self-efficacy measures resulted in significantly higher effect sizes.Additionally, the specificity of questionnaire items and conceptual accuracy of the self-efficacy measure positively predicted the relationships between self-efficacy beliefs andcommitment to teaching. Implications for the measurement of self-efficacy and interpre-tation of preservice and inservice teacher self-efficacy beliefs are presented.

© 2015 Elsevier Ltd. All rights reserved.

Contents

1. Introduction .............................................................................................................................................................................................................................. 21.1. Teacher self-efficacy .................................................................................................................................................................................................. 21.2. Role of self-efficacy in career decisions ............................................................................................................................................................. 31.3. Commitment to the teaching profession .......................................................................................................................................................... 41.4. Measuring self-efficacy beliefs ............................................................................................................................................................................. 51.5. Rationale for current study .................................................................................................................................................................................... 5

2. Methods ..................................................................................................................................................................................................................................... 62.1. The search .................................................................................................................................................................................................................... 62.2. Inclusion ....................................................................................................................................................................................................................... 62.3. Coding ........................................................................................................................................................................................................................... 72.4. Effect size calculations and statistical analyses .............................................................................................................................................. 82.5. Limitations of current meta-analysis ................................................................................................................................................................. 8

3. Findings ...................................................................................................................................................................................................................................... 84. Discussion ............................................................................................................................................................................................................................... 12

References ............................................................................................................................................................................................................................... 14

Authors’ note: This research, conducted by Steven Randall Chesnut and Hansel Burley, was completed in the Department of Educational Psychology atTexas Tech University: 3008 18th Street, Lubbock, Texas, 79409, United States of America.

* Corresponding author. Tel.: +1 405 928 8310.E-mail address: [email protected] (S.R. Chesnut).

http://dx.doi.org/10.1016/j.edurev.2015.02.0011747-938X/© 2015 Elsevier Ltd. All rights reserved.

Educational Research Review 15 (2015) 1–16

Contents lists available at ScienceDirect

Educational Research Review

journal homepage: www.elsevier.com/ locate /edurev

mailto:[email protected]

http://www.elsevier.com/locate/EDUREV

http://crossmark.crossref.org/dialog/?doi=10.1016/j.edurev.2015.02.001&domain=pdf

1. Introduction

The commitment that preservice and inservice teachers have to enter and remain in the teaching profession has been aconstruct of great interest among teacher educators (Chesnut & Cullen, 2014; Klassen & Chiu, 2011; Schlechty & Vance, 1981).Teacher education programs and professional development training prepare future and current educators to meet the needsof their students through a variety of mastery and vicarious experiences (e.g., Tschannen-Moran & McMaster, 2009). Whilemuch of this effort on the part of teacher educators and professional development specialists is rewarded with improvedteacher performance, over half of those who become teachers end up leaving the profession (Ingersoll, 2003; Tait, 2008;Tynjälä & Heikkinen, 2011). This attrition is especially prevalent in urban schools, characterized by a high poverty, high mi-nority student population (DeAngelis & Presley, 2011).

Throughout the years, researchers have attempted to quantify (e.g., track, measure) teacher attrition and explore the un-derlying reasons for teachers’ decisions to remain in or leave the profession. The decisions teachers make regarding entranceinto the profession, remaining in the profession, and leaving the profession have been suggested to originate from the directand indirect influence of occupation-related beliefs (e.g., self-efficacy, outcome expectations), interests, and distal and prox-imal choice goals (Brown & Lent, 2006; Lent, Brown, & Hackett, 1994; Siwatu & Chesnut, 2014). One of the most controversialfactors mentioned has been self-efficacy. While most teacher self-efficacy researchers believe that self-efficacy beliefs canpredict, to a large extent, an individual’s behaviors and performances (Bandura, 1977, 1986, 1997, 2006; Klassen, 2010; Klassen& Chiu, 2010, 2011; Schunk & Pajares, 2009; Schunk & Usher, 2011; Usher, 2009; Usher & Pajares, 2006a, 2006b, 2008), somehave failed to observe that connection (Chapman, 1984; Friedman, 2003). Bandura (1997, 2006) and Bong (2006) have sug-gested that this lack of connection is due to weaknesses in the accuracy of the self-efficacy measures, the specificity of theitems, and the alignment of the prompted behaviors with the actual outcome measures.

Prior studies that have investigated the psychometric properties of the relationships between self-efficacy beliefs andcommitment have suggested that increased variability in the self-efficacy scale provides greater explanatory potential forthe variation in commitment responses (e.g, Pajares, Hartley, & Valiante, 2001). Additionally, the level of specificity of theself-efficacy items has also been found to increase the explained variability in certain outcomes. While the field has grownsubstantially and we as researchers have been able to learn quite a bit about the measurement of self-efficacy beliefs, someof the arguments posed about the characteristics of self-efficacy measures and their problems remain theoretical. Thesetheoretical arguments provide false comfort when attempting to interpret self-efficacy measures aimed at improving teachereducation and development.

1.1. Teacher self-efficacy

Teacher self-efficacy has come to be one of the most commonly examined factors believed to influence preservice andinservice teacher commitment, burnout, student achievement, and willingness to adopt and implement reform efforts (Chesnut& Cullen, 2014; Tschannen-Moran & Woolfolk Hoy, 2001, 2007; Wheatley, 2000, 2002; Woolfolk Hoy & Davis, 2006). In hisseminal work, Bandura defined self-efficacy as “beliefs in one’s capabilities to organize and execute the courses of actionrequired to produce given attainments” (Bandura, 1997, p. 3). In essence, self-efficacy is an individual’s belief about whathe or she can do successfully (e.g., Bong, 2006). In regard to teaching, Dellinger, Bobbett, Olivier, and Ellett defined teacherself-efficacy as “individual’s beliefs in their capabilities to perform specific teaching tasks at a specified level of quality in aspecified situation” (Dellinger, Bobbett, Olivier, & Ellett, 2008, p. 752). Capturing these beliefs can prove beneficial to re-searchers and teacher educators because they represent the underlying self-beliefs of teachers regarding what can be successfullydone in the classroom.

The perspectives on efficacy and how it has been conceptualized and operationalized has evolved since the earliest re-corded studies in the 1960s. While Tschannen-Moran and Woolfolk Hoy (2001) provide a comprehensive history of teacherself-efficacy, our purpose is to highlight the milestones and major events that lead up to our contemporary conundrum.Based upon Rotter’s (1966) work in teacher beliefs, the Rand Corporation developed two items that sought to measure teach-ers’ beliefs as they concern their ability to influence student achievement. One of these items focused on the teacher as theinfluential factor in student achievement, and the second focused on environmental factors as influential in student achieve-ment. Responses to these two items proved to be very powerful predictors of student achievement. More specifically, teachersthat believed they could influence student achievement were more likely to have students with higher achievement scores.With this view of teacher efficacy, researchers pressed forward examining the influence of teachers’ perceptions of theirinfluence in student learning and student outcomes. Subsequent studies continued to show strong predictive relationshipsbetween teachers’ efficacy beliefs and student outcomes (Armor et al., 1976; Guskey, 1981, 1982, 1988). The problem withmeasures that looked teacher efficacy from this perspective is that they focused on the teachers’ beliefs that students’ changeswere based upon things that they can do, in its most general sense, juxtaposed to things that were outside of their control.While subsequent research went on to examine teacher efficacy from the perspective of locus of control, and even receiv-ing distinguishing titles such as personal and general teaching efficacy (Gibson & Dembo, 1984; Hoy & Woolfolk, 1990), theyignored a crucial aspect of Bandura’s (1977) interpretation: the actual teaching behaviors that lead to changes in studentoutcomes.

Before continuing, we want to clarify that the line of research focusing on teaching efficacy as locus of control has beenbeneficial to research on teacher education. It is beneficial to know that teachers who believe that they are responsible for

2 S.R. Chesnut, H. Burley / Educational Research Review 15 (2015) 1–16

student gains and that, on a global level, they can influence student development make for more effective instructors. Thistype of perspective can enhance orientations toward teacher education. In fact, Jerald (2007) reviewed the research andreported that teachers with a stronger sense of self-efficacy had better levels of planning and organization, were more re-silient when classroom strategies did not go well, were more open to experimentation, were less critical of students, andreferred students to special education less often. However, the field needs more research in this area, including the role ofmastery experiences, modeling by principals, and social persuasion, for example, that all students can succeed. When Bandura(1977) began discussing the role of self-efficacy as perceptions of ability and its predictive role in human behavior, he didnot do so with intention of measuring an individual’s beliefs that they can enact change. Instead, his arguments centeredon the existence of individuals’ beliefs in their ability to successfully engage and maintain a behavior that, without neces-sarily stating, would lead to them being successful agents in their environment.

Highlighted in his 1986 publication, Bandura discussed the role of human agency as an individual’s beliefs in his or herabilities to enact change in an environment. Broadly speaking, Bandura utilized agency as a term for locus of control. Whenan individual believes he or she is able to enact change in an environment, the subsequently observed change is likely tobe attributed to something that the individual did successfully. However, before someone believes he/she can influence theenvironment, he/she must be confident in his/her ability to execute and sustain the independent and combined behaviorsnecessary for that change (Bandura, 1997). This focus on the behaviors has not been forgotten since its conceptual and op-erational work in the late 1970s, but it has been frequently ignored or inappropriately cited to give credence to the otherline of research.

Since the late 1990s, following Bandura’s (1997) publication, teaching self-efficacy research has began to focus more ofits attention on the behaviors that teachers need to be able to successfully engage in order to be effective instructors. Thatis, instruments developed since 1997 have focused substantially more attention on effective teaching behaviors, such as “planactivities that accommodate the range of individual differences among my students” (Dellinger et al., 2008), “use my stu-dents’ cultural background to help make learning meaningful” (Siwatu, 2007), and “assist families in helping their childrendo well in school” (Tschannen-Moran & Woolfolk Hoy, 2001). Scales that utilize conceptually accurate items, as envisionedand defined by Bandura (1977, 1997, 2006), have frequently shown stronger relationships between teacher self-efficacy andstudent performance and learning (Tschannen-Moran & Woolfolk Hoy, 2001), and occupational related outcomes such ascommitment (Chesnut & Cullen, 2014; Klassen et al., 2012). Additionally, scales that utilize the context-specific aspect ofself-efficacy by increasing the specificity of the situation in which teachers must be able to enact a behavior (e.g., such asin a second language classroom), is an aspect of Bandura’s operationalization of self-efficacy that should enhance a scale’sability to predict an outcome, but is rarely given consideration. The problem with using the alternative lines of research,under the guise of self-efficacy, is that it does not provide the appropriate orientation to learn about teachers’ beliefs to engagein the behaviors needed to be successful, and the inferences drawn from these incorrect operational definitions are of limiteduse in further research and teacher education (Klassen, Tze, Betts, & Gordon, 2011; Skaalvik & Skaalvik, 2008; Wyatt, 2012).

1.2. Role of self-efficacy in career decisions

Self-efficacy beliefs have a foundational role in social cognitive theory (Bandura, 1986, 1997). Self-efficacy also plays afundamental role in Lent et al.’s (1994) social cognitive career theory (SCCT). In SCCT models, self-efficacy beliefs contrib-ute to the explanation of the development of an individual’s career interests, occupational and educational choices, and successin academic and career engagements (viz., Brown & Lent, 2006; Lent & Brown, 1996). As previously discussed, self-efficacybeliefs are context-specific appraisals that can fluctuate based upon an individual’s interpretations of everyday engage-ments, observations, and interactions (Bandura, 1977, 1997; Usher & Pajares, 2008). As SCCT models suggest, these beliefsinfluence preservice and inservice teachers’ decisions to pursue a career in teaching and remain in the profession (Brown& Lent, 2006; Lent et al., 1994).

SCCT’s choice model (Lent et al., 1994) depicts a process by which an individual’s occupational-related goals influencehis or her decision to pursue a career path. In particular, influenced by occupation-related beliefs (i.e., self-efficacy and outcomeexpectations), occupational interests develop, leading to occupational choice goals (Brown & Lent, 2006; Lent et al., 1994).These choice goals, in turn, motivate behaviors that will help individuals achieve their career-related goals. For example, aprospective teacher has high teaching self-efficacy and believes in the outcomes associated with being a teacher. Conse-quently, these beliefs influence his or her interest in the teaching profession. This interest influences career-related goalsof becoming a teacher. With the goal and the intentions of becoming a teacher, the prospective teacher will likely enroll ina traditional or alternative teacher preparation program and engage in the appropriate actions that will help him or herbecome a teacher.

Lent and colleagues (1994) used the choice model to explain how career-related goals can change as a result of positiveor negative experiences related to pursuing a particular career. For example, after declaring an elementary education major,the prospective teacher may engage in a wide variety of experiences in the classroom and in the field. When reflecting onthese experiences, the prospective teacher may realize the complexity of teaching math in an urban school. Should theseexperiences decrease self-efficacy beliefs, preservice teachers may modify their career-related goals (e.g., preference for teach-ing in a suburban school rather than an urban school, goal of becoming a math teacher) or alter their choice of occupation.

Ultimately, the choice model examines an individual’s commitment to an occupation. Commitment is a complex, mul-tifaceted concept that has come to be identified as the psychological bond and identification of an individual to an organization

3S.R. Chesnut, H. Burley / Educational Research Review 15 (2015) 1–16

or occupation that has special meaning (Chan, Lau, Nie, Lim, & Hogan, 2008; Coladarci, 1992; Firestone & Pennell, 1993;Fresko, Kfir, & Nasser, 1997). Given this definition, someone who is committed might show a strong affiliation or connec-tion to a career or organization. For teachers, this commitment might be evidenced by extended hours at school, a genuinelevel of care toward students and colleagues, and searching for new ways to reach students through professional develop-ment experiences. A lack of commitment, on the other hand, might be evidenced by a weakened bond to an organization,career, or colleagues that could give rise to a myriad of subsequent personal decisions and behaviors, such as leaving theprofession.

1.3. Commitment to the teaching profession

From an organizational psychology perspective, organizational commitment is an employee’s affective bond to an orga-nization (Mowday, Porter, & Steers, 1982). A leading framework proposed by Mowday et al. (1982) asserts that organizationalcommitment can be characterized by strong confidence in the organization’s goals, willingness to exert considerable effort,and a desire to stay with the organization. Commitment to the teaching profession, much like any other occupational com-mitment, is the psychological bond that an individual has with teaching, as a role, as an occupation, and as an institution(Chesnut & Cullen, 2014). Ware and Kitsantas (2011) described teacher commitment as the intention to stay in teaching.They also reported that teachers were found to have a higher level of commitment when they have efficacy to garner thesupport of their principals, influence policies at the workplace, and control instruction in their classes (Ware & Kitsantas,2007), with teachers who associate instructional outcomes to factors they control confronting new challenges with opti-mism (Rosenholtz, 1989). When teachers feel that they cannot control the terms of their work, the sense of powerlessnessthis causes results in disassociation from the products of work, including educational outcomes for students and avoid-ance of new work challenges (Rosenholtz, 1989).

Commitment has been, and can continue to be, considered a form of motivation (Rosenholtz, 1989). Motivation is broadlydefined as the process by which goal-directed activities are undertaken and sustained (Schunk, 2012). While the factorsthat influence this goal-directed behavior can be further subtyped into intrinsic and extrinsic motivators (Ryan & Deci, 2000),commitment should be interpreted as the goal-directed behavior that is influenced by both intrinsic (e.g., self-efficacy, value)and extrinsic motivators (e.g., salary, workload). Teacher commitment is related specifically to internal motivation, with highmotivation related to performance (Rosenholtz, 1989). Positive performance becomes self-rewarding while poor perfor-mance may result in alienation and disengagement, leading to poor performance (Ware & Kitsantas, 2011).

In the literature surrounding preservice teacher commitment, researchers have focused primarily on the positive natureof the term. More specifically, when viewing commitment with a positive frame of reference, researchers focus on aspectsof increasing psychological attachment, longevity of teaching career, and promoting the entrance of preservice teachers intothe teaching profession. In measurement, some researchers utilize questions that prompt teachers to indicate whether theyplan to return to or enter teaching in the next year (e.g., Evans & Tribble, 1986). Others seek answers to how many yearsteachers plan to stay in the profession (e.g., Bruinsma & Jansen, 2010). Most researchers, however, tend to focus on psy-chological self-report measures that seek to clarify how individuals perceive the value of the profession, attachment to thejob, and perceptions of fit into the expectations that are held for those in the profession (e.g., Chesnut & Cullen, 2014; Klassenet al., 2012).

In the literature surrounding inservice teacher commitment, researchers tend to focus primarily on the negative natureof the term. More specifically, when viewing commitment with a negative frame of reference, researchers focus on aspectsthat influence detachment, depersonalization, and ultimately withdrawal from the profession. While some researchers havebeen known to look at intentions to leave the profession (e.g., Coladarci, 1992), many utilize psychological self-report mea-sures that examine the amount of emotional exhaustion, depersonalization, accomplishment, and stress (e.g., Betoret, 2009;Chan, 2008; Chwalisz, Altmaier, & Russell, 1992). Ultimately, the combination of these elements conceptually defines burnout.

Teacher burnout has been examined nearly as much as teacher commitment. While there are different subcategories ofburnout (e.g., depersonalization, reduced personal accomplishment; Huberman & Vandenberghe, 1999), the most com-monly utilized is emotional burnout. Emotional burnout has been defined as “the state of physical and emotional depletionresulting from the conditions of work” (Freudenberger, 1974, p. 160). This depletion is the process of demotivation in whichteachers reevaluate the intrinsic and extrinsic motivators, ultimately resulting in lower identification and psychological bondwith the teaching profession, institution, or role as an educator. Teachers experiencing emotional burnout are likely to exhibitsigns opposite to teachers who are highly committed to teaching. For example, teachers experiencing high levels of emo-tional burnout are more likely to depersonalize their students and colleagues, reduce their desire for personal accomplishment,and exhibit fewer emotional reactions to events in the learning environment (Fimian & Blanton, 1987; Fives, Hamman, &Olivarez, 2007; Maslach, Jackson, & Schwab, 1986; Maslach & Pines, 1977). Research examining emotional burnout tendsto focus on inservice teachers. This bias makes conceptual sense, as preservice teachers are less likely to have experiencedenough negative events in the profession that would demotivate them from pursuing their career.

Conceptually similar to, but unique from, commitment, burnout describes the attachment and identification of an in-dividual to the profession. For example, items from the commonly used Maslach Burnout Inventory (Maslach et al., 1986)ask teachers if they feel that they perceive their daily tasks negatively, if they talk to others negatively about the job, whetherthe job is draining, and whether they need increasing amounts of time to relax after work. On the other hand, items fromcommon commitment scales ask teachers whether they identify and take problems with the teaching profession as theirs


to solve, if teaching has personal meaning, and if teaching would be a good career even if money were no issue. Under theumbrella of motivation to enter and remain in the teaching profession can be viewed on a loose continuum. At the nega-tive end of the continuum, studies that investigate burnout are able to fill in information about lower and decreasing levelsof motivation to remain in the profession. At the positive end of the continuum, studies that investigate commitment areable to fill in information about higher and increasing levels of motivation to enter and continue in the teaching profes-sion. While the visualization and conceptualization of this continuum are easy, it remains unclear why researchers chooseto examine burnout over commitment and how this selection influences the results that are reported.

With much of the focus on trying to increase teacher commitment and reduce teacher attrition (turnover), some re-searchers have investigated the fluctuations of commitment during teaching. Skinner (1996) suggested that in many instances,researchers could conclude the strength of commitment and the onset of emotional burnout by examining a teacher’s per-ception of the impact that they can make on their own environment. That is, teachers who believe that they can make adifference in their environment through personal engagements tend to exhibit strong and adaptive self-efficacy beliefs. Thisagency can greatly influence a teacher’s commitment to the profession (Bandura, 1997). However, when weak and mal-adaptive, teacher self-efficacy beliefs can also be the source of decreased confidence in perceived abilities and agency (Bandura,2001). While weak and maladaptive self-efficacy beliefs can be detrimental to a teacher’s longevity (Klassen & Chiu, 2011),these beliefs can be used to structure interventions to help prepare preservice and train inservice teachers (Siwatu & Chesnut,2014). Being able to use self-efficacy beliefs to structure instruction and interventions is dependent upon knowledge of teach-ers’ self-efficacy beliefs, which could be biased if they are improperly measured.

1.4. Measuring self-efficacy beliefs

In the appraisal of self-efficacy beliefs, individuals tend to draw upon their self-knowledge, the prompted task, and thestrategies that might make engagement in the task a successful endeavor (Bandura, 1997; Knoblauch & Woolfolk Hoy, 2008).In order to make an accurate appraisal, individuals must be able to compare what they can do to what is being asked ofthem. While it is likely that self-efficacy appraisals will be slightly inaccurate (e.g., over- or underestimate abilities), inter-preting and drawing implications from these appraisals is a greater concern (Bandura, 1997; Schunk & Pajares, 2009). It isimportant to understand what exactly is being measured in a self-efficacy scale.

Bandura (1997) and Wheatley (2005) have suggested that self-efficacy instruments not only ignore the underlying factorsthat influence self-efficacy appraisals (e.g., sources of information, perspectives of task), but that the construction of manyfamous instruments has been wrought with inaccurate conceptualizations and operational definitions (Klassen et al., 2011;Wyatt, 2012). Basing decisions and interpretations of self-efficacy research upon studies that utilized inaccurate concep-tualizations and operational definitions of self-efficacy beliefs might not only bias the outcomes of such processes, but mayfurther promulgate the use of inaccurate measurement methods. Instruments that claim to measure self-efficacy beliefs,yet measure only locus of control, self-esteem, or some other self construct perpetuate these misconceptions (Bong, 2006).

Interpreting self-efficacy scores and ratings is no easy task, especially for conceptually inaccurate measures. When tryingto interpret high versus low ratings on individual items, it is important to determine what perspective the respondent tookand whether or not the items truly captured the nature of self-efficacy as Bandura (1977, 1997) envisioned. Additionally,the spread of possible responses may bias outcomes. For example, a scale that measures self-efficacy on a 5-point Likert-type scale does not offer as much variability as a scale based upon a 0–100 scale because respondents tend to avoid extremepositions, so scales with just a few steps shrink quickly (Bandura, 2006). This lack of variability may lead to inaccurate de-terminations of who has higher self-efficacy (Pajares et al., 2001). Similarly, interpreting instrument composite scores, especiallyof an instrument that poorly measures self-efficacy, may give way to inaccurate conclusions about the nature of self-efficacy and its role in achievement, performance, or commitment. Without following up or interviewing participants withina temporally acceptable time frame (e.g., Bong, 2006), sole interpretation of self-efficacy ratings may fail to portray accu-rately the self-efficacy beliefs and perceptions of the participants (Wheatley, 2005; Wyatt, 2012). In terms of research andresponsive interventions for preservice and inservice teachers, it is important to understand what is being measured in self-efficacy scales and to be confident in their accuracy.

1.5. Rationale for current study

Our purpose in the present meta-analysis is to validate the above arguments made by self-efficacy researchers, as partof an effort to encourage accurate and appropriate self-efficacy measures for research and intervention. To date, stringentreviews of the relationship between teacher self-efficacy and commitment are absent in the literature. As self-efficacy growsinto an increasingly important factor of study in teacher commitment and burnout, it is important that we gain an over-view of the field and establish directions for future research. The present study will answer the following questions:

1. Are self-efficacy beliefs positively related to teacher commitment to the profession?2. Does the relationship between self-efficacy and commitment differ for preservice and inservice teachers?3. Does the relationship between self-efficacy and commitment differ when measuring commitment with a positive ori-

entation (e.g., commitment, intention) compared to a negative orientation (e.g., burnout, attrition)?4. Does the relationship between self-efficacy and commitment differ for teachers from diverse regions in the world?


5. Does the relationship between self-efficacy and commitment differ when measuring commitment with a positive orientation(e.g., commitment, intention) compared to a negative orientation (e.g., burnout, attrition)?

6. Does the relationship between self-efficacy and commitment differ for studies that accurately measure self-efficacy (ac-cording to Bandura, 1997) compared to those that inaccurately measure self-efficacy?

7. Does the specificity of self-efficacy measures and accuracy of conceptualizations influence the effects of the relation-ship between self-efficacy and teacher commitment?

2. Methods

The current meta-analysis employed techniques proposed by Lipsey and Wilson (2001). Multiple software suites wereutilized to carry out this meta-analysis. SPSS (v. 22) was utilized to facilitate data entry into databases and running the mul-tiple regression analysis. Microsoft Excel and Comprehensive Meta Analysis (Borenstein, Rothstein, & Cohen, 2012) wereutilized to calculate effect sizes (e.g., correlation coefficients to Fisher’s Z) and test Q and association parameters (pre-sented as χ2 estimates).

The procedures for this study followed the steps outlined by Lipsey and Wilson, (2001): (1) establishing a reason for thestudy, (2) locating all studies that might fit, (3) filtering the studies to keep only those that fit within the pre-establishedcriteria, (4) coding these studies based upon a code sheet into a database, (5) calculating values that can be used in metaanalyses (e.g., Fisher Z, inverse variance weight), and (6) conducting the statistical analyses testing for homogeneity, sig-nificance, and subgroup effects.

2.1. The search

All of the studies in this meta-analysis came from keyword-guided searches from online journal and dissertation data-bases and from searches guided by citations in relevant studies. Broad searches of published and unpublished literature onteacher commitment and attrition were conducted using several strategies and databases. The following databases weresearched, focusing on manuscripts, theses, and dissertations from 1980 to 2013: ERIC, PsycINFO, ProQuest Dissertations andTheses, and Google Scholar to ensure coverage of studies left out in previous searches. The following search terms wereutilized in the search for studies:

1. A mixture of teacher self-efficacy and terminology related to commitment (e.g., commitment, retention, intention, enterteaching).

2. A mixture of teacher self-efficacy and terminology related to burnout (e.g., withdrawal, dropout, attrition, turnover).3. A mixture of teacher self-efficacy and terminology related to commitment and burnout in studies that utilized popular

instruments to measure self-efficacy (e.g., Teacher Sense of Efficacy Scale, Teacher Efficacy Scale, Teachers’ Efficacy BeliefsSystem; Dellinger et al., 2008; Gibson & Dembo, 1984; Tschannen-Moran & Woolfolk Hoy, 2001)).

Searches returned thousands of published articles, and nearly 300 dissertations and conference proceedings on the topicsof self-efficacy, teachers, and commitment. Of the studies that were returned in our searches, a majority of them were removedon the premise of a qualitative framework (e.g., Hong, 2010). Those that utilized a quantitative framework were retained,but many of them were a function of noise within the search algorithms. More specifically, studies containing one of thecomponents (e.g., self-efficacy), but not the other (e.g., burnout), were returned. Additionally, studies were returned thatmentioned the relationship between self-efficacy and commitment in the manuscript, but focused the analysis on self-efficacy and some correlate of commitment (e.g., job satisfaction). In the end, 33 independent studies from 27 manuscriptswere retained, and appropriately weighted, for analysis. Of these 33 studies, 31 were published manuscripts and two weredissertation projects. Table 1 summarizes the studies retained for inclusion in the study.

2.2. Inclusion

To be included in this meta-analysis, the following inclusion criteria were established:

1. Studies involved preservice and inservice teachers’ perceptions and confidence in ability (self-efficacy) as it relates totheir subsequent commitment to the teaching profession.a. Commitment could either be interpreted positively as intention, retention, and commitment, or negatively as burnout,

attrition, and turnover.2. Studies were written in English. While international studies were specifically coded for nationality of sample, only those

written in English were retained.3. Studies were completed between 1980 and the present.4. Studies reported relationships between teacher self-efficacy and commitment that could be ultimately converted to effect

sizes meaningful to a multiple regression meta-analysis.5. Both published and unpublished studies, including peer-reviewed journal articles, dissertations, and conference presentations.


2.3. Coding

To examine the relationship between teacher self-efficacy beliefs and their commitment to the profession, studies thatremained after the final selection screenings were coded. Descriptive and substantive features, such as type of population,nationality of teacher sample, accuracy of self-efficacy measures, breadth of self-efficacy measures, and the correlation (e.g.,relationship) between self-efficacy and commitment, were placed in the database. The authors of this study trained withthe coding scheme before manuscripts were coded for this study. While disagreements initially arose, the iterative processof meeting and realignment facilitated an interrater agreement of 93% on the codes for 10 random studies. Minute dis-agreements on the specificity of the self-efficacy were the most common. Given that specificity was coded on a 5-point scale,the discrepancies tended to arise when one author coded the study a 4 and the other a 5. In situations where there was adiscrepancy in coding, the authors met to discuss and mutually decide upon a rating. The study features utilized in the anal-yses were coded in the following way:

1. Type of population: Preservice or inservice teachers.2. Type of study: Dissertation, thesis, conference, or publication.3. Origin of data: Countries were coded and combined into continents.4. Accuracy of self-efficacy items: Conceptually accurate or inaccurate (Bandura, 1997, 2006; Bong, 2006). Accurate ques-

tionnaire items will reflect what a respondent can do. These items will not reflect intentionality, self-esteem, locus ofcontrol, or outcome expectancies (Bandura, 2006).

5. Type of motivation to enter or remain in the profession: Commitment or burnout.6. Specificity of items: Scale of 1–5 indicating global (1), behaviorally specific (3), behaviorally and contextually specific

(5) self-efficacy items. Items at level 1 will have no clear domain or clear behavioral task, and may look like questions inmeasures of self-concept or locus of control (e.g., I can affect students’ learning). Items at level 3 will lack a clear domain,

Table 1Studies included in the meta-analysis.

Author (year) Type N Sample Origin Outcome ES (r) SE measure Acc. Spec.

Bruinsma and Jansen (2010) Journal 198 Preservice Europe Intent 0.210 TSES – Modified Yes 3Chesnut and Cullen (2014) Journal 209 Preservice N. America Commit 0.350 TSES – Original Yes 3Rots, Aelterman, Vlerick, and Vermeulen

(2007)Journal 209 Preservice Europe Commit 0.350 TSES – Modified Yes 3

Evans and Tribble (1986) Journal 179 Preservice N. America Commit 0.230 TES No 1Klassen and Chiu (2011) Journal 379 Preservice N. America Commit 0.310 TSES – Original Yes 3Rots and Aelterman (2009) Journal 301 Preservice Europe Commit 0.310 TSES – Translated Yes 3Rocca (2005) Dissertation 215 Preservice N. America Intent 0.421 TSES – Original Yes 3Chan et al. (2008) Journal 3715 Inservice Asia Commit 0.330 OSTES – Modified Yes 3Coladarci (1992) Journal 364 Inservice N. America Commit 0.280 TES No 1Fives et al. 2007 Journal 49 Preservice N. America Burnout −0.570 TSES Yes 3Betoret (2006) Journal 247 Inservice Europe Commit 0.403 Author Created No 3Betoret (2009) Journal 724 Inservice Europe Burnout −0.322 No 4Brouwers and Tomic (2000) Journal 558 Inservice Europe Burnout −0.450 Yes 5Klassen et al. (2012) – 1 Journal 379 Preservice N. America Commit 0.350 TSES – Original Yes 3Klassen et al. (2012) – 2 Journal 203 Preservice Europe Commit 0.300 TSES – Original Yes 3Klassen et al. (2012) – 3 Journal 211 Preservice Asia Commit 0.160 TSES – Original Yes 3Klassen et al. (2012) – 4 Journal 394 Preservice Asia Commit 0.310 TSES – Original Yes 3Schwarzer and Hallum (2008) – 1 Journal 608 Inservice Europe Burnout −0.170 TSE Yes 1Schwarzer and Hallum (2008) – 2 Journal 595 Inservice Europe Burnout −0.480 TSE Yes 1Watt and Richardson (2007) Journal 294 Preservice Australia Commit 0.160 FIT – Choice No 1Chapman (1984) Journal 1282 Inservice N. America Commit 0.070 Not Described No 1Shen (2009) Journal 530 Inservice Asia Burnout −0.168 GSES – Shortened No 1Brouwers, Evers, and Tomic (2001) Journal 277 Inservice Europe Burnout −0.340 Author Created Yes 3Chan (2002) Journal 83 Preservice Asia Burnout −0.330 Schwarzer, 1999 Yes 1Chan (2008) Journal 159 Inservice Asia Burnout −0.290 Author Created Yes 3Evers, Brouwers, and Tomic (2002) Journal 490 Inservice Europe Burnout −0.610 Author Created Yes 5Friedman (2003) Journal 322 Inservice Europe Burnout −0.130 Not Described No *Skaalvik and Skaalvik (2007) Journal 244 Inservice Europe Burnout −0.310 Author Created Yes 5Brissie, Hoover-Dempsey, and Bassler

(1988)Journal 1213 Inservice N. America Burnout −0.430 Author Created No 1

Madden-Szeszko et al. (2000) Dissertation 73 Inservice N. America Burnout −0.160 TES No 1Punch and Tuettemann (1990) Journal 574 Inservice Australia Burnout 0.270 Author Created No 1Louis (1998) Journal 528 Inservice N. America Intent 0.420 Author Created Yes 2Chwalisz et al. (1992) Journal 316 Inservice N. America Burnout −0.344 Author Created No 1

TSES – Teacher Sense of Efficacy Scale (Tschannen-Moran & Woolfolk Hoy, 2001); OSTES – Ohio State Teacher Efficacy Scale.TES – Teacher Efficacy Scale (Gibson & Dembo, 1984); TSE – Teacher Self-Efficacy (Schwarzer, Schmitz, & Daytner, 1999).GSES – General Self-Efficacy Scale (Wang, 1999).

* Full scale could not be retrieved for coding.


but will have a clear behavioral task (e.g., I can control disruptive behaviors of students in my classroom). Items at level5 will have a clear and relevant domain of functioning (Bandura, 2006), with that domain being some aspect of teach-ing (e.g., I can provide realistic challenges in mixed ability classes). In situations where a measure has varying levels ofspecificity, ratings were averaged and rounded, making possible the levels of 2 and 4.

2.4. Effect size calculations and statistical analyses

Since the relationship between self-efficacy and commitment is common among all studies in this review, the Pearsoncorrelation coefficient (r) was used as the effect size statistic. Effect sizes were calculated using the Pearson r to Fisher ztransformation (Lipsey & Wilson, 2001). In cases were studies reported multiple effect sizes (e.g., r), it was first determinedwhether these effect sizes were independent of one another before utilizing both. For example, a study that utilized threedifferent preservice teacher samples from different countries reported three different correlation coefficients between self-efficacy and commitment. Given that each of these correlations was independent from the other, they were coded as threedifferent effect sizes. To answer the first research question, fixed and random main effects were calculated. The main effectsand subgroup analyses for the subsequent research questions were conducted first using Microsoft Excel and then verifiedby using comprehensive meta analysis (Borenstein et al., 2012). The multiple regression was calculated using a weightedregression analysis in SPSS (v 22) and by then recalculating the standard errors and converting to z significance values (DeCoster,2004; Lipsey & Wilson, 2001).

Significance of moderating variables was calculated by comparing the total (QT), within (QW), and between (QB) vari-ances to a corresponding χ2 threshold (DeCoster, 2004; Lipsey & Wilson, 2001). The χ2 threshold was determined by calculatingdegrees of freedom for each level of variance. Degrees of freedom for total variance (QT) were calculated by k − 1, with kbeing equal to the number of studies in the sample (e.g., 33). Total degrees of freedom for within variance (QWtotal) werecalculated by k − j, with k being equal to the number of studies in the sample and j being equal to the number of categoriesin moderating variable(s). Disaggregated degrees of freedom for each within category (QW) were calculated by kwithin − 1, withkwithin being equal to the number of studies in each category of the moderating variables. Summing the disaggregated degreesof freedom for each group results in the same degrees of freedom for the QWtotal, but provides a foundation upon which grouplevel significance testing for homogeneity of variance can be calculated. Degrees of freedom for between variance (QB) werecalculated by j − 1, with j being equal to the number of categories in moderating variable(s). Using the appropriate degreesof freedom, the Q statistics were compared against corresponding χ2 thresholds for significance at the p < 0.05 level.

2.5. Limitations of current meta-analysis

Because this study required quantitative effect sizes, qualitative studies were removed from the analysis. While quali-tative analyses offer a great deal of knowledge to any field of research, they could not be quantified to the extent neededfor inclusion. As many of the studies on teacher self-efficacy and commitment were conducted from a qualitative frame-work, the removal of this information could weigh heavily on the interpretations and directions of future research in thisfield. Future reviews should meta-analyze the results of these qualitative studies so that the voices of these researchers willnot be overlooked. Second, attempts were made to contact the authors of older studies that utilized measures of teacherself-efficacy and commitment, yet failed to report correlational information. The non-response of some of these individu-als undoubtedly limited the scope and power of this study.

Another limitation is that the researchers selected only those studies written in English. Unfortunately, at the time ofdata collection, the researchers deemed translating the studies as cost and time prohibitive. The reader should be awarethat because of this limitation, the sample of included studies might be biased. However, according to Thornton and Lee(2000), all meta-analyses that do not review all studies are similarly biased. Still, a retrospective review of 303 meta-analyses using the English language restriction found no systematic bias (Morrison et al., 2012). Therefore, despite this limitation,some confidence is warranted when evaluating this study’s outcomes.

Meta-analyses are unique in that they offer a systematic investigation of the literature within field. While this aspect ofa meta-analysis may be its strength, it is also its weakness. Due to the collection procedures in meta-analyses, studies thatremain unpublished are ultimately left out. Failing to include these studies in the analyses biases the results in favor of thosethat were fortunate enough to be published (Rosenthal, 1979, 1991). The studies that fail to be published are stored awayand typically never see the light of day. In order to account for this file drawer problem, a fail safe N was calculated. Thefail safe N is a calculation of the number of studies with little or no effect size that would be needed to be included in theanalysis to reduce the overall correlation effect size to a non-significant level. Considering studies with a small, but non-zero correlation effect size, it would require 54 studies to lower the estimated correlation to a level that is no longer significant.

3. Findings

A total of 33 studies qualified for inclusion in the analysis. The 33 studies covered 16,122 preservice and inservice teach-ers from North America, Europe, Asia, and Australia. The overall weighted effect size (pooled correlation) was 0.318 for fixedeffects and 0.317 for random effects (SE = 0.008 for fixed effects and 0.026 for random effects). The large Q statistic valueindicated that this collection of studies was quite heterogeneous (Q = 295.67, df = 32, p < .001). Table 2 reports the fixed and


random effects models for the main effect. This suggests that the variance between cases within this study is larger thancould be explained by sampling error. In order to account for the variance in effect sizes, key methodological and samplecharacteristics were used to model some of the variation.

With regard to the teacher level, the studies were divided into two major categories: (1) preservice teachers (n = 14) and(2) inservice teachers (n = 19). Table 3 presents the summary results of the analysis by teacher level. Results suggested non-significance between group differences (QB = 1.65, df = 1, p = .2). This means that the difference in effect sizes between preserviceand inservice teachers was not significant. The variance within preservice teachers indicated heterogeneity (QW = 25.81, df = 13,p < .05), as did the variance within inservice teachers (QW = 268.21, df = 18, p < .001). While inservice teachers had a higherpooled correlation between self-efficacy and commitment, that difference was not great enough to establish significance.Results from the random effects model corroborated this finding of a non-significant difference between preservice and inserviceteacher groups.

With regard to the orientation in which researchers measured the motivation to enter and remain in the teaching pro-fession, the studies were divided into two major categories: (1) positive orientation (e.g., commitment, intention, entrance),and (2) negative orientation (e.g., burnout, attrition). Table 4 presents the summary results of the analysis by orientation.Initial results from the fixed analysis suggested significance between group differences (QB = 21, df = 1, p < .001). This meansthat the differences in effect sizes between orientations of measures were significant. However, the results from the randomeffects model do not corroborate this finding. Instead, results from the random effects model indicate no significant differ-ence in effect size as a function of the orientation of the commitment measure (QB = 1.06, df = 1, p = .3). Given that therewill always be more variance in the population than what can be accounted for in a sample, the results of the random effectsmodel provide a wider perspective. Both models, however, indicated heterogeneity within each of the two groups (Nega-tive: QW = 164.16, df = 18, p < .001; Positive: QW = 110.17, df = 16, p < .001). The lack of significant difference between groupsprovides evidence upon which we can validate the decision to examine burnout and commitment as measuring a similarconstruct, yet at distinct ends of the same continuum.

With regard to the conceptual accuracy of the self-efficacy scale being utilized, the studies were divided into two majorcategories: (1) conceptually accurate (n = 21) and (2) conceptually inaccurate (n = 12). Table 5 presents the summary resultsof the analysis by conceptual accuracy. The determination of conceptual self-efficacy was based upon Bandura’s (1997, 2006)definition of self-efficacy items and manner in which items should be phrased. Results suggested significance between groupdifferences (QB = 43.62, df = 1, p < 0.001), with accurate self-efficacy measures reporting an effect size of +0.354 and inaccurate

Table 2Overall effect sizes.

k ES SE Variance Test of null 95% CI Test of homogeneity

Z P Lower Upper Q-value df (Q) P-value

Fixed 33 .318 .008 .000 41.63 <.001 .304 .331 295.67 32 <.001Random 33 .317 .026 .001 12.82 <.001 .271 .362

Table 3By teacher level.

Model Group k ES SE Test of null 95% CI Test of homogeneity

Z P Lower Upper Q df(Q) P

Fixed Preservice 14 .300 .018 16.67 <.001 .268 .330 25.81 13 <.05Inservice 19 .322 .009 35.78 <.001 .307 .338 268.21 18 <.001QB 1.65 1 =.2

Random Preservice 14 .301 .023 13.09 <.001 .255 .346 25.81 13 <.05Inservice 19 .324 .033 9.82 <.001 .259 .386 268.21 18 <.001QB 0.32 1 =.6

Table 4By orientation.



Fixed Positive 17 .289 .010 29.82 <.001 .270 .308 110.17 16 <.001Negative 19 .356 .011 31.73 <.001 .334 .376 164.16 18 <.001QB 21 1 <.001

Random Positive 17 .294 .029 10.29 <.001 .238 .348 110.17 16 <.001Negative 19 .342 .038 8.94 <.001 .267 .413 164.16 18 <.001QB 1.06 1 =.3


self-efficacy measures reporting an effect size of +0.256. This indicates that scales employing an accurate measure of self-efficacy had a significantly higher correlation with commitment to the teaching profession than did studies that employedscales that inaccurately measured self-efficacy beliefs. This is consistent with Bandura’s (1997) claim that self-efficacy beliefscan better predict outcomes when the beliefs are accurately measured. The variance within the accurate studies indicatedheterogeneity (QW =130.38, df = 20, p < 0.001), similar to the variance within inaccurate studies (QW = 121.67, df = 11, p < 0.001).Results from the random effects model corroborate this significant finding (QB = 4.44, df = 1, p < 0.05). Figure 1 demon-strates this heterogeneity. Overall, the conceptual accuracy of scales can influence the magnitude of the correlations foundbetween self-efficacy and teacher commitment; however, significant variance within each of these categories requires furtherexamination.

With regard to sample location, the studies were divided into four major categories, each describing the continent inwhich the participants were located: (1) North America (n = 12), (2) Europe (n = 11), (3) Asia (n = 8), and (4) Australia (n = 2).

Table 5By conceptual accuracy.



Fixed Accurate 21 .354 .009 38.56 <.001 .336 .371 130.38 20 <.001Inaccurate 12 .256 .012 21.33 <.001 .233 .280 121.67 11 <.001QB 43.62 1 <.001

Random Accurate 21 .354 .026 13.88 <.001 .304 .402 130.38 20 <.001Inaccurate 12 .253 .043 5.903 <.001 .169 .333 121.67 11 <.001QB 4.44 1 <.05

Fig. 1. Comparison between conceptually accurate and inaccurate measures.


Table 6 presents the summary of results of the analysis by location. Results suggested significance between group differ-ences (QB = 56.99, df = 3, p < 0.001). Within group differences suggested heterogeneous variance for the North America(QW = 125.23, df = 11, p < 0.001), Europe (QW = 72.28, df = 10, p < 0.001), and Asia (QW = 38.60, df = 7, p < 0.05) groups. Resultsfor the Australia group suggest homogeneous variance between the cases (QW = 2.57, df = 1, p > 0.05); however, because itwas a small sample, this value should be interpreted with caution. Results from the random effects model corroborate thesefindings. Overall, results indicate that there is significant between-group effect sizes with much within-group variance.

Meta-analyses are used to find not just between group differences, as one might hope for with an ANOVA, but also ho-mogeneity within groups indicating a goodness of fit (Lipsey & Wilson, 2001). As there was much variance within the groupsin both the conceptually accurate and by location analyses, further investigation was warranted. A 4 × 2 comparison anal-ysis was conducted by dividing location effect sizes by the accuracy of self-efficacy conceptualization. Table 7 presents thesummary of the results of the analysis. Results suggest significance between group differences (QB = 106.53, df = 6, p < 0.001).Contrary to the previous analyses, the further disaggregated groups began to demonstrate levels of within group homoge-neity of variance. Significant within group variance emerged in the Accurate by Europe group (QW = 64.38, df = 8, p < 0.001),in the Accurate by Asia group (QW = 20.29, df = 5, p < 0.001), and in the Inaccurate by North America group (QW = 98.10, df = 5,p < 0.001). While it is logically understandable to have within group variance in studies that fail to accurately conceptual-ize self-efficacy as it relates to teacher commitment, studies that accurately conceptualize and measure self-efficacy shouldbe more consistent.

As there was much variance within the groups in both the conceptually accurate and by orientation analyses, furtherinvestigation was warranted. A 2 × 2 comparison analysis was conducted by dividing orientation of measure (e.g., positive,

Table 6By location.



Fixed N. Amer. 12 .302 .013 23.68 <.001 .277 .327 125.23 11Europe 11 .404 .014 29.33 <.001 .377 .429 72.28 10Asia 8 .282 .012 24.03 <.001 .259 .305 38.60 7Australia 2 .234 .033 7.17 <.001 .170 .296 2.57 1QB 56.99 3 <.001

Random N. Amer. 12 .328 .046 7.06 <.001 .237 .413 125.23 11Europe 11 .381 .038 9.96 <.001 .306 .452 72.28 10Asia 8 .234 .038 6.20 <.001 .160 .305 38.60 7Australia 2 .223 .056 4.01 <.001 .114 .326 2.57 1QB 10.18 3 <.01

Table 7Self-efficacy moderated by accuracy of conceptualization × origin of population.

Country of origin Accurate conceptualization Inaccurate conceptualization

North America # Cases (sample) 6 (1759) 6 (3427)Effect size (SE) .3980 (.024) .2679 (.017)95% CI [.351–.445] [.234–.301]QW (df) 7.56 (5) 98.10 (5)P-value P > 0.05 P < 0.001

Europe # Cases (sample) 9 (3075) 2 (971)Effect size (SE) .4503 (.018) .3575 (.03295% CI [.414–.486] [.294–.421]QW (df) 64.38 (8) 1.54 (1)P-value P < 0.001 P > 0.05

Asia # Cases (sample) 6 (5170) 2 (852)Effect size (SE) .3123 (.014) .1562 (.034)95% CI [.285–.340] [.089–.223]QW (df) 20.29 (5) 0.33 (1)P-value p < 0.001 P > 0.05

Australia # Cases (sample) 0 (0) 2 (868)Effect size (SE) N/A .2379 (.034)95% CI N/A [.171–.305]QW (df) N/A 2.57 (1)P-value N/A P > 0.05

Test of homogeneityQ-value df (Q) P-value

QB 106.53 6 <0.001

Cells are not optimal sizes for the analysis; however, this information does make a strong argumentabout the forces behind the observed relationship between self-efficacy and teacher commitment.


negative) effect sizes by the accuracy of self-efficacy conceptualization. Table 8 presents the summary of the results of theanalysis. Results suggest significance between group differences (QB = 96.38, df = 6, p < 0.001). As a moderator, commitmentscales that utilized a negative orientation (e.g., burnout) had higher relationships with teacher self-efficacy than positivemeasures of commitment. Additionally, when accurate measures of self-efficacy were utilized, these relationships were thehighest. While these reflect the prior analyses, further questions are raised. Why are the relationships between negativelyoriented measures of commitment and inaccurately measured self-efficacy beliefs nearly as large as positively oriented mea-sures of commitment and accurately measured self-efficacy beliefs? Are there characteristics of negatively oriented measuresof commitment that better reference the identification and orientation toward the teaching profession? While the groupsshow decreased levels of within group variances as indicated by lower QW values, they are still significant. This significantwithin group variance indicates that more moderators should be able to account for the variance. However, theoreticallyand pragmatically, these potential moderators are currently unknown.

The specificity of self-efficacy measures has been a topic of major discussion within the field (Bandura, 1997, 2006; Bong,2006; Dellinger et al., 2008). The specificity of a self-efficacy measure tends to span a global to specific continuum. On theglobal end of the spectrum, self-efficacy items tend to reference what outcomes an individual believes he/she can enact.For example, a global item might read, “I can affect student learning.” This item represents something that an individualbelieves he/she can accomplish, but ignores the small details of the actual behaviors that lead to student learning and thecontext in which student learning must occur. On the specific end of the spectrum, self-efficacy items tend to not mentionthe accomplishment of an outcome, but focus more on the behaviors and contexts of an individual’s engagement. For example,a specific item might read, “I can provide realistic challenges in mixed ability classes.” Bandura (1997, 2006) suggests thatglobal self-efficacy measures will suffer from predictive power when the outcome is specified or contextualized. Self-efficacy measures that incorporate highly specific items, however, will provide only predictive power when the outcome isin the same context or domain.

Given the claims that contextually specific and conceptually accurate self-efficacy measures should more strongly predictthe relationship with a contextualized outcome (in this case teacher commitment), we utilized a multiple regression. Resultssuggest that instrument specificity and conceptual accuracy were both statistically significant predictors of the relation-ships between teacher self-efficacy and commitment to the teaching profession. The predictors accounted for 22.9% of thevariance in the relationship between self-efficacy and commitment. Table 9 summarizes the regression information.

4. Discussion

The findings of this meta-analysis indicate that preservice and inservice teachers’ self-efficacy beliefs are positivelyrelated to their commitment to the teaching profession (ES = +0.32). This finding is statistically and practically significant.Cohen (1977, 1988) would classify this as a moderate effect size (Low: r = 0.1; Moderate: r = 0.3; Large: r = 0.5). By squaringthis value, we are able to attain the amount of variance that self-efficacy explains in commitment. Alone, self-efficacy canaccount for 10% of the variance in the decision that teachers make to enter, remain, and leave the profession. The values ofthe ES and r2 represent strong effects. Using the binomial effect size display (BESD), if half the population were categorized

Table 8Self-efficacy moderated by accuracy of conceptualization × measure orientation.

Measure of orientation Accurate conceptualization Inaccurate conceptualization

Positive measure # Cases (sample) 12 (6941) 5 (2366)Effect size (SE) .330 (.013) .163 (.020)95% CI [.309–.351] [.123–.202]QW (df) 19.44 (11) 34.57 (4)p-value p > .05 p < .001

Negative measure # Cases (sample) 9 (3063) 7 (3752)Effect size (SE) .405 (.015) .313 (.015)95% CI [.375–.435] [.284–.342]QW (df) 94.99 (8) 50.30 (6)P-value P < .001 P < .001

Test of homogeneityQ-value df (Q) P-value

QB 96.38 3 <.001

Table 9Instrument specificity and conceptual accuracy as predictors of the relationship between self-efficacyand commitment.

Variable B Adjusted SE of B Z (β replacement) P

Instrument specificity 0.034 0.008 4.25 <0.001Conceptual accuracy 0.040 0.020 2.00 <0.05


as low commitment and the other as high commitment (with each new variable split at the median) and high self-efficacyand low self-efficacy, an ES of .32 represents an improvement rate of 34% to 66% (Rosenthal, Rosnow, & Rubin, 2000).Therefore, the ES between self-efficacy and the extent to which teachers are committed to the profession hypotheticallyrepresents a gain of 32%, another indication that the relationship between self-efficacy and teacher commitment is astrong one.

Consistent with the theoretical perspectives surrounding self-efficacy beliefs (Bandura, 1997), this study confirms andcorroborates the research suggesting that preservice and inservice teacher self-efficacy beliefs can be used to predict com-mitment to the teaching profession (e.g., Chesnut & Cullen, 2014; Klassen & Chiu, 2011). Unlike the findings in previousself-efficacy studies that downplay the link it might have with performance and choice outcomes, this study implementsand examines the fundamental characteristics of self-efficacy measures (e.g., accuracy and specificity) in understanding therelationships that self-efficacy beliefs share with commitment to the teaching profession.

Previous studies on preservice and inservice teacher self-efficacy beliefs suggest that inservice teachers are likely to havemore accurate appraisals than preservice teachers (e.g., Woolfolk Hoy & Spero, 2005). While the results from this meta-analysis cannot corroborate or refute these claims, they do suggest that there is a non-significant difference between therole that these beliefs play in commitment to the teaching profession between these two groups. In other words, while thepooled correlation between inservice teachers’ self-efficacy beliefs and commitment to teaching was larger than preserviceteachers’, it was not significantly different. Because of this, we suggest that researchers not overlook the value of self-efficacy beliefs in preservice teachers for engaging in teaching practices, even if they seem to be something beyond therespondents’ repertoires. The confidence that preservice teachers have toward their ability to be successful in teaching prac-tices significantly influences their commitment to the teaching profession.

A critical interest of educational researchers is trying to understand how their ideas, theories, and constructs transfer tointernational populations with different cultures and socialization trends. According to the findings from this study, the cor-relations between self-efficacy beliefs and commitment to teaching were largest in Europe and smaller in magnitude forNorth America, Asia, and Australia. When we examined the role of conceptually accurate and inaccurate self-efficacy mea-sures by country, we found that regardless of the country, conceptually accurate self-efficacy measures had significantlyhigher correlations with commitment than conceptually inaccurate measures. Further studies should examine the reasonfor these international discrepancies. Are the differences related to the domain of this study (e.g., teacher commitment) orare they the result of some underlying social or cultural aspect that has yet to be reported?

Many years after introducing self-efficacy, Bandura began to specifically give attention to the measurement andinterpretation of self-efficacy instruments (Bandura, 1997, 2006). The primary issue surrounding the measurement andinterpretation of self-efficacy beliefs began with the development of popular self-efficacy questionnaires created by Rotter(1966) and Gibson and Dembo (1984). While questionable self-efficacy measures were created before and have beencreated since, the problem that self-efficacy researchers have expressed is that the measure of self-efficacy is unique andnot the same construct as locus of control, self-esteem, or self-concept, which many of these instruments unintentionallymeasure. Bandura (1997, 2006) and Bong (2006), among others, have suggested that the inaccuracies within self-efficacymeasures can be the reason that these purported measures of the construct fail to predict outcomes. The findings in thisstudy indicate that conceptually accurate self-efficacy measures report significantly larger correlations with commitmentto teaching than conceptually inaccurate measures. When measuring self-efficacy beliefs, it should be a primary concernfor the researchers that they adhere to self-efficacy instruments as the construct is defined and exemplified by Bandura(1997, 2006) and Bong (2006). Failing to utilize conceptually accurate self-efficacy measures will not only influenceobserved relationships with outcome variables, but it may also influence the likelihood of publication and the effective-ness of responsive interventions.

Alternatively, given the findings of this study, researchers should further investigate the differences in conceptualiza-tion and operationalization of commitment measures as positive and negative measures of commitment do not sharecommon relationships with self-efficacy. In our study, the findings (from fixed effects) suggest that studies utilizing nega-tively oriented measures of commitment will report significantly larger effect sizes. While the random effects model didnot corroborate this finding, the trend still poses an interesting question about how we interpret measures of commitmentand burnout. Additionally, when examining the 2 × 2 matrix of accuracy by orientation, this question is expanded. Fromthe results, we can conclude that measures that employ accurate self-efficacy conceptualizations will have a larger corre-lation with commitment than inaccurate conceptualizations. We can additionally conclude that negatively oriented measuresof commitment will have a larger correlation with self-efficacy. Given that the correlation between negatively orientedmeasures of commitment and inaccurate conceptualizations of self-efficacy was nearly as strong as between positivelyoriented measures of commitment and accurate conceptualizations of self-efficacy, what is it about negatively orientedmeasures of commitment that enhance the relationship? Are measures of burnout and emotionality more closely alignedwith the processes of leaving the profession than psychological identification with the process of entering and remainingin the profession? Researchers would serve the field well by further analyzing the negative and positive orientations ofcommitment measures.

In addition to the need for adherence to conceptual accuracy, Bandura (1997, 2006) also suggested that self-efficacy in-struments be as specific as necessary for the study. Other researchers have offered similar thoughts (e.g., Schunk & Usher,2011), suggesting that if self-efficacy measures are too general, they will fail to account for much of the variation in theoutcome; however, if they are too specific, they will not provide much for generalization of results. Building upon these


perspectives, the findings from this study suggest that the accuracy of self-efficacy measures and the specificity of the itemswill predict the magnitude of the correlation between self-efficacy beliefs and commitment to teaching. That is, concep-tually accurate and specific (e.g., contextual, behavioral) self-efficacy measures are more likely to culminate in higherrelationships than those that are global and/or inaccurate. Mirroring Bandura’s and other researchers’ recommendations,self-efficacy instruments should be conceptually accurate and as contextually and behaviorally specific to the intended outcomemeasure as possible. Using global self-efficacy scales will result in less impressive predictability and may negatively impactone’s analyses and study results.

The construct known as self-efficacy has come to be widely studied in many domains. While it has a stronghold in ed-ucation, many have inaccurately utilized the construct and misinterpreted the influence and the value of self-efficacy inpredicting peoples’ behaviors. Additionally, inappropriate usage of self-efficacy has influenced the development and im-plementation of incorrect interventions (Klassen et al., 2011; Wyatt, 2012). The knowledge of self-efficacy beliefs can providemuch information to researchers and teacher educators. For example, these beliefs can be used to structure teacher edu-cation and professional development experiences. Utilizing accurate and specific self-efficacy belief measures can enhancethe ability for researchers and teacher educators to meet the needs of preservice and inservice teachers.

References

*Indicates references used in the analysis.Armor, D., Conroy-Oseguera, P., Cox, M., King, N., McDonnell, L., Pascal, A., et al. (1976). Analysis of the school preferred reading programs in selected Los

Angeles minority schools. (Report no. R-2007-LAUSD). Santa Monica, CA: Rand Corporation. ERIC Document Reproduction Service No. 130 243.Bandura, A. (1977). Self-efficacy: toward a unifying theory of behavioral change. Psychological Review, 84(2), 191–215. doi:10.1037/0033-295X.84.2.191.Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory. Englewood Cliffs, NJ: Prentice Hall.Bandura, A. (1997). Self-efficacy: The exercise of control. New York: W.H. Freeman and Company.Bandura, A. (2001). Social cognitive theory: an agentic perspective. Annual Review of Psychology, 52, 1–26. doi:10.1146/annurev.psych.52.1.1.Bandura, A. (2006). Guide for creating self-efficacy scales. In F. Pajares & T. Urdan (Eds.), Self-efficacy beliefs of adolescents (pp. 307–337). Greenwich, CT:

Information Age Publishing.*Betoret, F. D. (2006). Stressors, self-efficacy, coping resources, and burnout among secondary school teachers in Spain. Educational Psychology, 26(4), 519–539.*Betoret, F. D. (2009). Self-efficacy, school resources, job stressors and burnout among Spanish primary and secondary school teachers: a structural equation

approach. Educational Psychology, 29(1), 45–68.Bong, M. (2006). Asking the right question: How confident are you that you could successfully perform these tasks? In F. Pajares & T. Urdan (Eds.), Self-efficacy

beliefs of adolescents (pp. 287–305). Greenwich, CT: Information Age.Borenstein, M., Rothstein, H., & Cohen, J. (2012). Comprehensive meta-analysis: A computer program for research synthesis [computer software]. Englewood,

NJ: Biostat.*Brissie, J. S., Hoover-Dempsey, K. V., & Bassler, O. C. (1988). Individual, situational contributors to teacher burnout. The Journal of Educational Research,

82(2), 106–112. doi:10.2307/40539578.*Brouwers, A., Evers, W. J. G., & Tomic, W. (2001). Self-efficacy in eliciting social support and burnout among secondary-school teachers. Journal of Applied

Social Psychology, 31(7), 1474–1491. doi:10.1111/j.1559-1816.2001.tb02683.x.*Brouwers, A., & Tomic, W. (2000). A longitudinal study of teacher burnout and perceived self-efficacy in classroom management. Teaching and Teacher

Education, 16(2), 239–253.Brown, S. D., & Lent, R. W. (2006). Preparing adolescents to make career decisions. In F. Pajares & T. Urdan (Eds.), Self-efficacy beliefs of adolescents (pp.

201–223).*Bruinsma, M., & Jansen, E. P. (2010). Is the motivation to become a teacher related to preservice teachers’ intentions to remain in the profession? European

Journal of Teacher Education, 33(2), 185–200.*Chan, D. W. (2002). Stress, self-efficacy, social support, and psychological distress among prospective Chinese teachers in Hong Kong. Educational Psychology,

22(5), 557–569. doi:10.1080/0144341022000023635.*Chan, D. W. (2008). Dimensions of teacher self-efficacy among Chinese secondary school teachers in Hong Kong. Educational Psychology, 28(2), 181–194.

doi:10.1080/01443410701491833.*Chan, W. Y., Lau, S., Nie, Y., Lim, S., & Hogan, D. (2008). Organizational and personal predictors of teacher commitment: the mediating role of teacher

efficacy and identification with school. American Educational Research Journal, 45(3), 597–630. doi:10.3102/0002831208318259.*Chapman, D. W. (1984). Teacher retention: the test of a model. American Educational Research Journal, 21(3), 645–658. doi:10.3102/00028312021003645.Chesnut, S. R., & Cullen, T. A. (2014). Effects of self-efficacy, emotional intelligence, and perceptions of future work environment on preservice teacher

commitment. The Teacher Educator, 49(2), 116–132. doi:10.1080/08878730.2014.887168.*Chwalisz, K., Altmaier, E. M., & Russell, D. W. (1992). Causal attributions, self-efficacy cognitions, and coping with stress. Journal of Social and Clinical

Psychology, 11(4), 377–400.Cohen, J. (1977). Statistical power analysis for the behavioral sciences (Rev. ed). New York: Academic Press.Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsday, NJ: Erlbaum.*Coladarci, T. (1992). Teachers’ sense of efficacy and commitment to teaching. The Journal of Experimental Education, 60(4), 323–337. doi:10.1080/

00220973.1992.9943869.DeAngelis, K. J., & Presley, J. B. (2011). Toward a more nuanced understanding of new teacher attrition. Education and Urban Society, 43(5), 598–626.

doi:10.1177/0013124510380724.DeCoster, J. (2004). Meta-analysis notes. <http://www.stat-help.com/notes.html> Retrieved 28.09.13.Dellinger, A. B., Bobbett, J. J., Olivier, D. F., & Ellett, C. D. (2008). Measuring teachers’ self-efficacy beliefs: development and use of the TEBS-self. Teaching

and Teacher Education, 24(3), 751–766. doi:10.1016/j.tate.2007.02.010.*Evans, E. D., & Tribble, M. (1986). Perceived teaching problems, self-efficacy, and commitment to teaching among preservice teachers. The Journal of Educational

Research, 80, 81–85.*Evers, W. J. G., Brouwers, A., & Tomic, W. (2002). Burnout and self-efficacy: A study on teachers’ beliefs when implementing an innovative educational

system in the Netherlands. British Journal of Educational Psychology, 72(2), 227–243. doi:10.1348/000709902158865.Fimian, M. J., & Blanton, L. P. (1987). Stress, burnout, and role problems among teacher trainees and first-year teachers. Journal of Occupational Behavior,

8(2), 157–165.Firestone, W. A., & Pennell, J. R. (1993). Teacher commitment, working conditions, and differential policies. Review of Educational Research, 63(4), 489–525.*Fives, H., Hamman, D., & Olivarez, A. (2007). Does burnout begin with student-teaching? Analyzing efficacy, burnout, and support during the student-teaching

semester. Teaching and Teacher Education, 23(6), 916–934. doi:10.1016/j.tate.2006.03.013.Fresko, B., Kfir, D., & Nasser, F. (1997). Predicting teacher commitment. Teaching and Teacher Education, 13(4), 429–438. doi:10.1016/s0742-051x(96)00037-6.Freudenberger, H. J. (1974). Staff burn-out. Journal of Social Issues, 30(1), 159–165.


http://refhub.elsevier.com/S1747-938X(15)00004-4/sr0010




















http://www.stat-help.com/notes.html








*Friedman, I. A. (2003). Self-efficacy and burnout in teaching: the importance of interpersonal-relations efficacy. Social Psychology of Education, 6(3), 191–215.doi:10.1023/A:1024723124467.

Gibson, S., & Dembo, M. (1984). Teacher efficacy: a construct validation. Journal of Educational Psychology, 76, 569–582. doi:10.1037/0022-0663.76.4.569.Guskey, T. R. (1981). Measurement of responsibility teachers assume for academic successes and failures in the classrooms. Journal of Teacher Education,

32, 44–51.Guskey, T. R. (1982). Differences in teachers’ perceptions of personal control of positive versus negative student learning outcomes. Contemporary Educational

Psychology, 7, 70–80.Guskey, T. R. (1988). Teacher efficacy, self-concept, and attitudes toward the implementation of instructional innovation. Teaching and Teacher Education,

4, 63–69.Hong, J. Y. (2010). Pre-service and beginning teachers’ professional identity and its relation to dropping out of the profession. Teaching and Teacher Education,

26(8), 1530–1543.Hoy, W. K., & Woolfolk, A. E. (1990). Socialization of student teachers. American Educational Research Journal, 27, 279–300.Huberman, M., & Vandenberghe, R. (1999). Introduction: Burnout and the teaching profession. In R. Vandenberghe & M. Huberman (Eds.), Understanding

and preventing teacher burnout: A sourcebook of international research and practice. Cambridge, UK: Cambridge UP.Ingersoll, R. (2003). Is there really a teacher shortage? A report co-sponsored by the center for the study of teaching and policy and the center for policy

research in education. Seattle: University of Washington, Center for the Study of Teaching and Policy.Jerald, C. D. (2007). Believing and achieving (issue brief). Washington, DC: Center for Comprehensive School Reform and Improvement.Klassen, R. M. (2010). Teacher stress: the mediating role of collective efficacy beliefs. The Journal of Educational Research, 103(5), 342–350. doi:10.1080/

00220670903383069.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(3), 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(2), 114–129. doi:10.1016/j.cedpsych.2011.01.002.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(1), 21–43. doi:10.1007/s10648-010-9141-8.*Klassen, R., Wilson, E., Siu, A. F. Y., Hannok, W., Wong, M. W., Wongsri, N., et al. (2012). Preservice teachers’ work stress, self-efficacy, and occupational

commitment in four countries. European Journal of Psychology of Education, 28, 1289–1309. doi:10.1007/s10212-012-0166-x.Knoblauch, D., & Woolfolk Hoy, A. (2008). “Maybe I can teach those kids.” The influence of contextual factors on student teachers’ efficacy beliefs. Teaching

and Teacher Education, 24, 166–179.Lent, R. W., & Brown, S. D. (1996). Social cognitive approach to career development: an overview. Career Development Quarterly, 44(4), 310–

322.Lent, R. W., Brown, S. D., & Hackett, G. (1994). Toward a unifying social cognitive theory of career and academic interest, choice, and performance. Journal

of Vocational Behavior, 45(1), 79–122. doi:10.1006/jvbe.1994.1027.Lipsey, M., & Wilson, D. (2001). Practical meta-analysis. Thousand Oaks, CA: Sage.*Louis, K. S. (1998). Effects of teacher quality of work life in secondary schools on commitment and sense of efficacy. School Effectiveness and School

Improvement, 9(1), 1–27.Madden-Szeszko, G. (2000). Variables contributing to teacher efficacy: An examination of burnout, affect, demographic variables and general self-efficacy

(unpublished doctoral dissertation). Hempstead, NY: Hofstra University.Maslach, C., Jackson, S., & Schwab, R. (1986). Maslach Burnout Inventory: Educator survey.Maslach, C., & Pines, A. (1977). The burnout syndrome in the day care setting. Child and Youth Care Forum, 6(2), 100–113.Morrison, A., Polisena, J., Husereau, D., Moulton, K., Clark, M., Fiander, M., et al. (2012). The effect of English-language restriction on systematic review-based

meta-analyses: a systematic review of empirical studies. International Journal of Technology Assessment in Health Care, 28(2), 138–144. doi:10.1017/S0266462312000086.

Mowday, R. T., Porter, L. W., & Steers, R. (1982). Organizational linkages: the psychology of commitment, absenteeism, and turnover. turnover. San Diego,CA: Academic Press.

Pajares, F., Hartley, J., & Valiante, G. (2001). Response format in writing self-efficacy assessment: greater discrimination increases prediction. Measurementand Evaluation in Counseling and Development, 33(4), 214–221.

*Punch, K. F., & Tuettemann, E. (1990). Correlates of psychological distress among secondary school teachers. British Educational Research Journal, 16(4),369–382. doi:10.1080/0141192900160405.

*Rocca, S. J. (2005). Predicting preservice agriculture teachers’ intentions to teach utilizing person inputs, contextual influences, teacher efficacy, and outcomeexpectations (Doctoral dissertation). University of Florida.

Rosenholtz, S. J. (1989). Teachers’ workplace: The social organization of schools. White Plains, NY: Longman.Rosenthal, R. (1979). The file drawer problem and tolerance for null results. Psychological Bulletin, 86(3), 638–641.Rosenthal, R. (1991). Meta-analytic procedures for social research. Newbury Park, CA: Sage.Rosenthal, R., Rosnow, R. L., & Rubin, D. B. (2000). Contrasts and effect sizes in behavioral research: A correlational approach. Cambridge: Cambridge University

Press.*Rots, I., & Aelterman, A. (2009). Teacher education graduates’ entrance into the teaching profession: development and test of a model. European Journal

of Psychology of Education, 24(4), 453–471.*Rots, I., Aelterman, A., Vlerick, P., & Vermeulen, K. (2007). Teacher education, graduates’ teaching commitment and entrance into the teaching profession.

Teaching and Teacher Education, 23(5), 543–556.Rotter, J. B. (1966). Generalized expectancies for internal versus external control of reinforcement. Psychological Monographs, 80(1), 1–28.Ryan, R. M., & Deci, E. L. (2000). Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being. American Psychologist,

55(1), 68–78.Schlechty, P. C., & Vance, V. S. (1981). Do academically able teachers leave education? The North Carolina case. The Phi Delta Kappan, 63(2), 106–112.Schunk, D. (2012). Learning theories: An educational perspective (6th ed.). Boston: Pearson.Schunk, D. H., & Pajares, F. (2009). Self-efficacy theory. In K. R. Wentzel & A. Wigfield (Eds.), Handbook of motivation at school (pp. 35–54). New York, NY:

Routledge.Schunk, D. H., & Usher, E. L. (2011). Assessing self-efficacy for self-regulated learning. In B. J. Zimmerman & D. H. Schunk (Eds.), Handbook of self-regulation

of learning and performance (pp. 282–297). New York: Routledge.*Schwarzer, R., & Hallum, S. (2008). Perceived teacher self-efficacy as a predictor of job stress and burnout: mediation analyses. Applied Psychology, 57,

152–171. doi:10.1111/j.1464-0597.2008.00359.x.Schwarzer, R., Schmitz, G. S., & Daytner, G. T. (1999). Teacher self efficacy. Retrieved December 1, 2013, from http://userpage.fu-berlin.de/%7Ehealth/

teacher_se.htm*Shen, Y. E. (2009). Relationships between self-efficacy, social support and stress coping strategies in Chinese primary and secondary school teachers. Stress

and Health, 25(2), 129–138. doi:10.1002/smi.1229.Siwatu, K. O. (2007). Preservice teachers’ culturally responsive teaching self-efficacy and outcome expectancy beliefs. Teaching and Teacher Education,

23(7), 1086–1101.Siwatu, K. O., & Chesnut, S. R. (2014). The career development of preservice and inservice teachers: Why teachers’ self-efficacy beliefs matter. In H. Fives

& M. Gill (Eds.), International handbook of research on teachers’ beliefs (pp. 212–229). New York: Routledge.























































http://userpage.fu-berlin.de/%7Ehealth/teacher_se.htm

http://userpage.fu-berlin.de/%7Ehealth/teacher_se.htm





*Skaalvik, E. M., & Skaalvik, S. (2007). Dimensions of teacher self-efficacy and relations with strain factors, perceived collective teacher efficacy, and teacherburnout. Journal of Educational Psychology, 99(3), 611–625. doi:10.1037/0022-0663.99.3.611.

Skaalvik, E. M., & Skaalvik, S. (2008). Teacher self-efficacy: conceptual analysis and relations with teacher burnout and perceived school context. In R. Craven,H. W. Marsh, & D. McInerney (Eds.), Self-processes, learning, and enabling human potential (pp. 223–247). Connecticut: Information Age Publishing.

Skinner, E. A. (1996). A guide to constructs of control. Journal of Personality and Social Psychology, 71(3), 549.Tait, M. (2008). Resilience as a contributor to novice teacher success, commitment, and retention. Teacher Education Quarterly, 35(4), 57–75.Thornton, A., & Lee, P. (2000). Publication bias in meta-analysis: it causes and consequences. Journal of Clinical Epidemiology, 53, 207–216.Tschannen-Moran, M., & McMaster, P. (2009). Self-efficacy: four professional development formats and their relationship to self-efficacy and implementation

of a new teaching strategy. The Elementary School Journal, 110(2), 228–245.Tschannen-Moran, M., & Woolfolk Hoy, A. W. (2001). Teacher efficacy: capturing an elusive construct. Teaching and Teacher Education, 17, 783–805.Tschannen-Moran, M., & Woolfolk Hoy, A. W. (2007). The differential antecedents of self-efficacy beliefs of novice and experienced teachers. Teaching and

Teacher Education, 23(6), 944–956. doi:10.1016/j.tate.2006.05.003.Tynjälä, P., & Heikkinen, H. L. T. (2011). Beginning teachers’ transition from pre-service education to working life. Zeitschrift für Erziehungswissenschaft,

14(1), 11–33. doi:10.1007/s11618-011-0175-6.Usher, E. L. (2009). Sources of middle school students’ self-efficacy in mathematics: a qualitative investigation. American Educational Research Journal,

46(1), 275–314. doi:10.3102/0002831208324517.Usher, E. L., & Pajares, F. (2006a). Sources of academic and self-regulatory efficacy beliefs of entering middle school students. Contemporary Educational

Psychology, 31(2), 125–141. doi:10.1016/j.cedpsych.2005.03.002.Usher, E. L., & Pajares, F. (2006b). Inviting confidence in school: invitations as a critical source of the academic self-efficacy beliefs of entering middle school

students. Journal of Invitational Theory & Practice, 12, 7–16.Usher, E. L., & Pajares, F. (2008). Self-efficacy for self-regulated learning. Educational and Psychological Measurement, 68(3), 443–463. doi:10.1177/

0013164407308475.Wang, L. (1999). The structure and measurement of self-efficacy. Acta Scientiarum Naturalium, Universitatis Pekinensis, 35(3), 414–420.Ware, H. W., & Kitsantas, A. (2007). Teacher and collective efficacy beliefs as predictors of professional commitment. Journal of Educational Research, 100,

303–310.Ware, H. W., & Kitsantas, A. (2011). Predicting teacher commitment using principal and teacher efficacy variables: an HLM approach. The Journal of Educational

Research, 104, 183–193.*Watt, H. M. G., & Richardson, P. W. (2007). Motivational factors influencing teaching as a career choice: development and validation of the FIT-Choice

scale. The Journal of Experimental Education, 75(3), 167–202. doi:10.3200/JEXE.75.3.167-202.Wheatley, K. F. (2000). Positive teacher efficacy as an obstacle to educational reform. Journal of Research and Development in Education, 34(1), 14–27.Wheatley, K. F. (2002). The potential benefits of teacher efficacy doubts for educational reform. Teaching and Teacher Education, 18, 5–22.Wheatley, K. F. (2005). The case for reconceptualizing teacher efficacy research. Teaching and Teacher Education, 21(7), 747–766.Woolfolk Hoy, A., & Davis, H. A. (2006). Teacher self-efficacy and its influence on the achievement of adolescents. In F. Pajares & T. Urdan (Eds.), Self-efficacy

beliefs of adolescents (pp. 117–137). Greenwich, CT: Information Age.Woolfolk Hoy, A., & 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.Wyatt, M. (2012). Towards a re-conceptualization of teachers’ self-efficacy beliefs: tackling enduring problems with the quantitative research and moving

on. International Journal of Research & Method in Education, 37, 166–189. doi:10.1080/1743727X.2012.74205.




































Documents

Self-efficacy as a Predictor of Commitment to the Teaching Profession_a Meta-Analysis