You are on page 1of 16

Educational Research Review 15 (2015) 1–16

Contents lists available at ScienceDirect

Educational Research Review
j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / e d u r e v


Self-efficacy as a predictor of commitment to the teaching
profession: A meta-analysis
Steven Randall Chesnut *, Hansel Burley
Texas Tech University, Lubbock, Texas 79409-1071, USA



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


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 on
self-efficacy and commitment, this review systematically examines the effects found within
the literature and highlights important theoretical and methodological issues. A total of
33 qualified studies were included in the final analysis, including 16,122 preservice and
inservice teachers. Findings suggest that preservice and inservice teachers’ self-efficacy beliefs
influence their commitment to the teaching profession (ES = +0.32). However, these effects
vary 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 selfefficacy measure positively predicted the relationships between self-efficacy beliefs and
commitment to teaching. Implications for the measurement of self-efficacy and interpretation of preservice and inservice teacher self-efficacy beliefs are presented.
© 2015 Elsevier Ltd. All rights reserved.




Introduction .............................................................................................................................................................................................................................. 2
Teacher self-efficacy .................................................................................................................................................................................................. 2
Role of self-efficacy in career decisions ............................................................................................................................................................. 3
Commitment to the teaching profession .......................................................................................................................................................... 4
Measuring self-efficacy beliefs ............................................................................................................................................................................. 5
Rationale for current study .................................................................................................................................................................................... 5
Methods ..................................................................................................................................................................................................................................... 6
The search .................................................................................................................................................................................................................... 6
Inclusion ....................................................................................................................................................................................................................... 6
Coding ........................................................................................................................................................................................................................... 7
Effect size calculations and statistical analyses .............................................................................................................................................. 8
Limitations of current meta-analysis ................................................................................................................................................................. 8
Findings ...................................................................................................................................................................................................................................... 8
Discussion ............................................................................................................................................................................................................................... 12
References ............................................................................................................................................................................................................................... 14

Authors’ note: This research, conducted by Steven Randall Chesnut and Hansel Burley, was completed in the Department of Educational Psychology at
Texas Tech University: 3008 18th Street, Lubbock, Texas, 79409, United States of America.
* Corresponding author. Tel.: +1 405 928 8310.
E-mail address: (S.R. Chesnut).
1747-938X/© 2015 Elsevier Ltd. All rights reserved.

Throughout the years. & Valiante. we want to clarify that the line of research focusing on teaching efficacy as locus of control has been beneficial to research on teacher education. Schunk & Usher. These theoretical arguments provide false comfort when attempting to interpret self-efficacy measures aimed at improving teacher education and development. Klassen & Chiu. While most teacher self-efficacy researchers believe that self-efficacy beliefs can predict. 1982. over half of those who become teachers end up leaving the profession (Ingersoll. Subsequent studies continued to show strong predictive relationships between teachers’ efficacy beliefs and student outcomes (Armor et al. juxtaposed to things that were outside of their control. self-efficacy. While subsequent research went on to examine teacher efficacy from the perspective of locus of control. researchers have attempted to quantify (e. 1984. and leaving the profession have been suggested to originate from the direct and indirect influence of occupation-related beliefs (e. Teacher self-efficacy Teacher self-efficacy has come to be one of the most commonly examined factors believed to influence preservice and inservice teacher commitment. characterized by a high poverty. 2009). the Rand Corporation developed two items that sought to measure teachers’ beliefs as they concern their ability to influence student achievement. 2003. 2011). Bobbett. While the field has grown substantially and we as researchers have been able to learn quite a bit about the measurement of self-efficacy beliefs. Guskey.. 2009. Woolfolk Hoy & Davis. outcome expectations). Chesnut.g. 1981. 1990). some of the arguments posed about the characteristics of self-efficacy measures and their problems remain theoretical. p. Klassen. Wheatley. 1976. With this view of teacher efficacy. 2008. 2006. Additionally. 2010. Schunk & Pajares. The perspectives on efficacy and how it has been conceptualized and operationalized has evolved since the earliest recorded studies in the 1960s. Hartley. high minority student population (DeAngelis & Presley. One of the most controversial factors mentioned has been self-efficacy.1. self-efficacy is an individual’s belief about what he or she can do successfully (e. 1997. In essence. and the second focused on environmental factors as influential in student achievement..2 S. Bong. Tschannen-Moran & McMaster. 2011. remaining in the profession. 1986. Burley / Educational Research Review 15 (2015) 1–16 1..R. Capturing these beliefs can prove beneficial to researchers and teacher educators because they represent the underlying self-beliefs of teachers regarding what can be successfully done in the classroom. The problem with measures that looked teacher efficacy from this perspective is that they focused on the teachers’ beliefs that students’ changes were based upon things that they can do. 2006). burnout. Introduction The commitment that preservice and inservice teachers have to enter and remain in the teaching profession has been a construct of great interest among teacher educators (Chesnut & Cullen. interests. While much of this effort on the part of teacher educators and professional development specialists is rewarded with improved teacher performance. 2009. 2001. p.g. 2002. Before continuing. Teacher education programs and professional development training prepare future and current educators to meet the needs of their students through a variety of mastery and vicarious experiences (e. track. Pajares. 752). and the alignment of the prompted behaviors with the actual outcome measures. and willingness to adopt and implement reform efforts (Chesnut & Cullen. Prior studies that have investigated the psychometric properties of the relationships between self-efficacy beliefs and commitment have suggested that increased variability in the self-efficacy scale provides greater explanatory potential for the variation in commitment responses (e. 1997. they ignored a crucial aspect of Bandura’s (1977) interpretation: the actual teaching behaviors that lead to changes in student outcomes. an individual’s behaviors and performances (Bandura. It is beneficial to know that teachers who believe that they are responsible for . In his seminal work. 2014.g. This attrition is especially prevalent in urban schools. Olivier. 2011. & Hackett.g. H. 2006b. 3). & Ellett. More specifically. 2008. Tait. 2006a. Responses to these two items proved to be very powerful predictors of student achievement. 1977. the specificity of the items. 2006). measure) teacher attrition and explore the underlying reasons for teachers’ decisions to remain in or leave the profession. researchers pressed forward examining the influence of teachers’ perceptions of their influence in student learning and student outcomes. While Tschannen-Moran and Woolfolk Hoy (2001) provide a comprehensive history of teacher self-efficacy. Bandura (1997. The decisions teachers make regarding entrance into the profession. Brown. 2011). and even receiving distinguishing titles such as personal and general teaching efficacy (Gibson & Dembo. Tynjälä & Heikkinen. 2000. student achievement. Lent. Schlechty & Vance. 2010. the level of specificity of the self-efficacy items has also been found to increase the explained variability in certain outcomes. 2003). in its most general sense. and distal and proximal choice goals (Brown & Lent. and Ellett defined teacher self-efficacy as “individual’s beliefs in their capabilities to perform specific teaching tasks at a specified level of quality in a specified situation” (Dellinger. our purpose is to highlight the milestones and major events that lead up to our contemporary conundrum. 2008). 1988). 2011. 2001). Hoy & Woolfolk. some have failed to observe that connection (Chapman. Klassen & Chiu. Tschannen-Moran & Woolfolk Hoy. Bandura defined self-efficacy as “beliefs in one’s capabilities to organize and execute the courses of action required to produce given attainments” (Bandura. 1994. Usher & Pajares. 2014. 2007. to a large extent.. Based upon Rotter’s (1966) work in teacher beliefs. One of these items focused on the teacher as the influential factor in student achievement. 1984. 1981). In regard to teaching. Bobbett.g. 2006. teachers that believed they could influence student achievement were more likely to have students with higher achievement scores. 2006) and Bong (2006) have suggested that this lack of connection is due to weaknesses in the accuracy of the self-efficacy measures.. Dellinger. Friedman. Usher. Olivier. Siwatu & Chesnut. 2014). 1.

e. 2006.. As previously discussed. self-efficacy beliefs contribute to the explanation of the development of an individual’s career interests. for example. scales that utilize the context-specific aspect of self-efficacy by increasing the specificity of the situation in which teachers must be able to enact a behavior (e. the subsequently observed change is likely to be attributed to something that the individual did successfully. “use my students’ cultural background to help make learning meaningful” (Siwatu. As SCCT models suggest.. Self-efficacy also plays a fundamental role in Lent et al. but is rarely given consideration. motivate behaviors that will help individuals achieve their career-related goals. Chesnut. that all students can succeed. Highlighted in his 1986 publication. Jerald (2007) reviewed the research and reported that teachers with a stronger sense of self-efficacy had better levels of planning and organization. SCCT’s choice model (Lent et al. is that it does not provide the appropriate orientation to learn about teachers’ beliefs to engage in the behaviors needed to be successful. is an aspect of Bandura’s operationalization of self-efficacy that should enhance a scale’s ability to predict an outcome. 1997. These choice goals. and occupational related outcomes such as commitment (Chesnut & Cullen. Wyatt. This interest influences career-related goals of becoming a teacher. Bandura discussed the role of human agency as an individual’s beliefs in his or her abilities to enact change in an environment. Lent and colleagues (1994) used the choice model to explain how career-related goals can change as a result of positive or negative experiences related to pursuing a particular career. Betts. For example. after declaring an elementary education major. self-efficacy and outcome expectations). 2012). When Bandura (1977) began discussing the role of self-efficacy as perceptions of ability and its predictive role in human behavior. Klassen et al. & Gordon.g.2. and social persuasion. When reflecting on these experiences. a prospective teacher has high teaching self-efficacy and believes in the outcomes associated with being a teacher. Broadly speaking. Additionally.g. Lent et al. occupational and educational choices. the prospective teacher may engage in a wide variety of experiences in the classroom and in the field. leading to occupational choice goals (Brown & Lent. were more open to experimentation. Ultimately. preservice teachers may modify their career-related goals (e. Skaalvik & Skaalvik. However. Brown & Lent. influenced by occupation-related beliefs (i. and success in academic and career engagements (viz. and interactions (Bandura. 2012). Since the late 1990s. 1997). Should these experiences decrease self-efficacy beliefs.’s (1994) social cognitive career theory (SCCT). Lent et al. Tze. 1977. these beliefs influence preservice and inservice teachers’ decisions to pursue a career in teaching and remain in the profession (Brown & Lent. Commitment is a complex. In SCCT models. such as “plan activities that accommodate the range of individual differences among my students” (Dellinger et al.. multifaceted concept that has come to be identified as the psychological bond and identification of an individual to an organization . without necessarily stating. Burley / Educational Research Review 15 (2015) 1–16 3 student gains and that. have frequently shown stronger relationships between teacher self-efficacy and student performance and learning (Tschannen-Moran & Woolfolk Hoy.. 2008. before someone believes he/she can influence the environment. and referred students to special education less often. 2006. the prospective teacher may realize the complexity of teaching math in an urban school. This focus on the behaviors has not been forgotten since its conceptual and operational work in the late 1970s. he did not do so with intention of measuring an individual’s beliefs that they can enact change. observations. modeling by principals. Role of self-efficacy in career decisions Self-efficacy beliefs have a foundational role in social cognitive theory (Bandura.. H. following Bandura’s (1997) publication. 1997.. 2014. 1994) depicts a process by which an individual’s occupational-related goals influence his or her decision to pursue a career path.. Scales that utilize conceptually accurate items.R. 1986. Instead. In particular. 1997). the field needs more research in this area. This type of perspective can enhance orientations toward teacher education. teaching self-efficacy research has began to focus more of its attention on the behaviors that teachers need to be able to successfully engage in order to be effective instructors. these beliefs influence his or her interest in the teaching profession. Usher & Pajares..S. The problem with using the alternative lines of research. on a global level. goal of becoming a math teacher) or alter their choice of occupation. 2011. were less critical of students. his arguments centered on the existence of individuals’ beliefs in their ability to successfully engage and maintain a behavior that. but it has been frequently ignored or inappropriately cited to give credence to the other line of research. 1994). including the role of mastery experiences. and the inferences drawn from these incorrect operational definitions are of limited use in further research and teacher education (Klassen. However. instruments developed since 1997 have focused substantially more attention on effective teaching behaviors. were more resilient when classroom strategies did not go well. they can influence student development make for more effective instructors. he/she must be confident in his/her ability to execute and sustain the independent and combined behaviors necessary for that change (Bandura. 2001). 2007). such as in a second language classroom). preference for teaching in a suburban school rather than an urban school. in turn. self-efficacy beliefs are context-specific appraisals that can fluctuate based upon an individual’s interpretations of everyday engagements. 1994). For example. 1. under the guise of self-efficacy. 2001). Consequently. In fact. 1996). the choice model examines an individual’s commitment to an occupation. 2006. Lent & Brown. 2006). With the goal and the intentions of becoming a teacher. When an individual believes he or she is able to enact change in an environment. 2008). 2008). the prospective teacher will likely enroll in a traditional or alternative teacher preparation program and engage in the appropriate actions that will help him or her become a teacher. and “assist families in helping their children do well in school” (Tschannen-Moran & Woolfolk Hoy. would lead to them being successful agents in their environment. as envisioned and defined by Bandura (1977.. Bandura utilized agency as a term for locus of control. That is. occupational interests develop.

For example. Commitment to the teaching profession. Fives. A lack of commitment.3. This bias makes conceptual sense. Altmaier. such as leaving the profession. More specifically. including educational outcomes for students and avoidance of new work challenges (Rosenholtz. (1982) asserts that organizational commitment can be characterized by strong confidence in the organization’s goals. considered a form of motivation (Rosenholtz. researchers have focused primarily on the positive nature of the term.. if they talk to others negatively about the job. Teacher burnout has been examined nearly as much as teacher commitment. 1989). Maslach. might be evidenced by a weakened bond to an organization. depersonalization. workload). and can continue to be. 2007). with high motivation related to performance (Rosenholtz. as an occupation. is the psychological bond that an individual has with teaching. Teachers experiencing emotional burnout are likely to exhibit signs opposite to teachers who are highly committed to teaching. For example. 1977). influence policies at the workplace.R. & Nasser.g. but unique from. Motivation is broadly defined as the process by which goal-directed activities are undertaken and sustained (Schunk. A leading framework proposed by Mowday et al. Chan. willingness to exert considerable effort. In the literature surrounding preservice teacher commitment. Ware and Kitsantas (2011) described teacher commitment as the intention to stay in teaching. the sense of powerlessness this causes results in disassociation from the products of work. much like any other occupational commitment. depersonalization. 1987.. leading to poor performance (Ware & Kitsantas. or role as an educator. and as an institution (Chesnut & Cullen.4 S. & Olivarez. On the other hand. Coladarci. and whether they need increasing amounts of time to relax after work. someone who is committed might show a strong affiliation or connection to a career or organization. p. however. Evans & Tribble. Ultimately. and promoting the entrance of preservice teachers into the teaching profession. While some researchers have been known to look at intentions to leave the profession (e. organizational commitment is an employee’s affective bond to an organization (Mowday. accomplishment.. & Hogan. this commitment might be evidenced by extended hours at school. 1. 1993.g. Maslach & Pines. and stress (e. 160). a genuine level of care toward students and colleagues. While there are different subcategories of burnout (e. whether the job is draining. 1986) ask teachers if they feel that they perceive their daily tasks negatively. & Steers.g. self-efficacy. Emotional burnout has been defined as “the state of physical and emotional depletion resulting from the conditions of work” (Freudenberger. Jackson. For teachers.g. or colleagues that could give rise to a myriad of subsequent personal decisions and behaviors. Chwalisz. Hamman. 2007. Klassen et al. 2012). This depletion is the process of demotivation in which teachers reevaluate the intrinsic and extrinsic motivators. when viewing commitment with a negative frame of reference. 1989). 2008.. 1974. salary. many utilize psychological self-report measures that examine the amount of emotional exhaustion.. career. Conceptually similar to. with teachers who associate instructional outcomes to factors they control confronting new challenges with optimism (Rosenholtz. researchers focus on aspects that influence detachment. 2000). Kfir. They also reported that teachers were found to have a higher level of commitment when they have efficacy to garner the support of their principals. items from the commonly used Maslach Burnout Inventory (Maslach et al. Chesnut & Cullen. ultimately resulting in lower identification and psychological bond with the teaching profession. Chesnut.. burnout describes the attachment and identification of an individual to the profession. and control instruction in their classes (Ware & Kitsantas.. 1992. some researchers utilize questions that prompt teachers to indicate whether they plan to return to or enter teaching in the next year (e. Fresko.. 2014). 1992). Burley / Educational Research Review 15 (2015) 1–16 or occupation that has special meaning (Chan. 1982).. Others seek answers to how many years teachers plan to stay in the profession (e. items from common commitment scales ask teachers whether they identify and take problems with the teaching profession as theirs . researchers tend to focus primarily on the negative nature of the term. reduced personal accomplishment. Given this definition.g. 2009. Lau. on the other hand. 1986. Porter. longevity of teaching career. 1989). institution. Betoret. researchers focus on aspects of increasing psychological attachment. & Schwab. and exhibit fewer emotional reactions to events in the learning environment (Fimian & Blanton. and searching for new ways to reach students through professional development experiences. & Russell. 1986). as a role. H. value) and extrinsic motivators (e. Huberman & Vandenberghe. when viewing commitment with a positive frame of reference. Nie.. 2008. and a desire to stay with the organization.g. 2010). Bruinsma & Jansen. attachment to the job. 2014. When teachers feel that they cannot control the terms of their work. Commitment to the teaching profession From an organizational psychology perspective.g. as preservice teachers are less likely to have experienced enough negative events in the profession that would demotivate them from pursuing their career. Teacher commitment is related specifically to internal motivation. Firestone & Pennell. 1999). 2012). 2011). tend to focus on psychological self-report measures that seek to clarify how individuals perceive the value of the profession. In the literature surrounding inservice teacher commitment. the most commonly utilized is emotional burnout. reduce their desire for personal accomplishment. the combination of these elements conceptually defines burnout. teachers experiencing high levels of emotional burnout are more likely to depersonalize their students and colleagues. Most researchers. Commitment has been. depersonalization. commitment. Lim. More specifically. Research examining emotional burnout tends to focus on inservice teachers. 1989). Coladarci. Positive performance becomes self-rewarding while poor performance may result in alienation and disengagement. and ultimately withdrawal from the profession. In measurement.g. While the factors that influence this goal-directed behavior can be further subtyped into intrinsic and extrinsic motivators (Ryan & Deci. commitment should be interpreted as the goal-directed behavior that is influenced by both intrinsic (e. and perceptions of fit into the expectations that are held for those in the profession (e. 1992). 1997).

However. over.or underestimate abilities). 2012). 2011. if teaching has personal meaning. The present study will answer the following questions: 1. especially for conceptually inaccurate measures. Does the relationship between self-efficacy and commitment differ for preservice and inservice teachers? 3.5. In terms of research and responsive interventions for preservice and inservice teachers. as part of an effort to encourage accurate and appropriate self-efficacy measures for research and intervention. Being able to use self-efficacy beliefs to structure instruction and interventions is dependent upon knowledge of teachers’ self-efficacy beliefs. Bandura (1997) and Wheatley (2005) have suggested that self-efficacy instruments not only ignore the underlying factors that influence self-efficacy appraisals (e. attrition)? 4. 1. Instruments that claim to measure self-efficacy beliefs. As self-efficacy grows into an increasingly important factor of study in teacher commitment and burnout.g. While the visualization and conceptualization of this continuum are easy. so scales with just a few steps shrink quickly (Bandura. some researchers have investigated the fluctuations of commitment during teaching. sole interpretation of self-efficacy ratings may fail to portray accurately the self-efficacy beliefs and perceptions of the participants (Wheatley. 1. but may further promulgate the use of inaccurate measurement methods. Measuring self-efficacy beliefs In the appraisal of self-efficacy beliefs. 1997). Skinner (1996) suggested that in many instances. 2012). 2006). teachers who believe that they can make a difference in their environment through personal engagements tend to exhibit strong and adaptive self-efficacy beliefs. especially of an instrument that poorly measures self-efficacy. it is important to understand what is being measured in selfefficacy scales and to be confident in their accuracy. studies that investigate burnout are able to fill in information about lower and decreasing levels of motivation to remain in the profession. it remains unclear why researchers choose to examine burnout over commitment and how this selection influences the results that are reported.. stringent reviews of the relationship between teacher self-efficacy and commitment are absent in the literature..g. it is important that we gain an overview of the field and establish directions for future research. That is. For example. H. but that the construction of many famous instruments has been wrought with inaccurate conceptualizations and operational definitions (Klassen et al. Chesnut. At the negative end of the continuum. which could be biased if they are improperly measured. Interpreting self-efficacy scores and ratings is no easy task.S. With much of the focus on trying to increase teacher commitment and reduce teacher attrition (turnover). At the positive end of the continuum.g. Bong. Are self-efficacy beliefs positively related to teacher commitment to the profession? 2. performance. the prompted task. and if teaching would be a good career even if money were no issue. 2006). Additionally. Schunk & Pajares. interpreting instrument composite scores. Does the relationship between self-efficacy and commitment differ when measuring commitment with a positive orientation (e. Burley / Educational Research Review 15 (2015) 1–16 5 to solve. interpreting and drawing implications from these appraisals is a greater concern (Bandura.4. when weak and maladaptive. it is important to determine what perspective the respondent took and whether or not the items truly captured the nature of self-efficacy as Bandura (1977. 2006). This lack of variability may lead to inaccurate determinations of who has higher self-efficacy (Pajares et al. the spread of possible responses may bias outcomes.. 2008). 1997. Wyatt. 2001). individuals must be able to compare what they can do to what is being asked of them. 2011). these beliefs can be used to structure interventions to help prepare preservice and train inservice teachers (Siwatu & Chesnut. sources of information.R. teacher self-efficacy beliefs can also be the source of decreased confidence in perceived abilities and agency (Bandura. individuals tend to draw upon their self-knowledge. 2005. commitment. Wyatt. To date. Rationale for current study Our purpose in the present meta-analysis is to validate the above arguments made by self-efficacy researchers. or some other self construct perpetuate these misconceptions (Bong. Similarly. yet measure only locus of control. studies that investigate commitment are able to fill in information about higher and increasing levels of motivation to enter and continue in the teaching profession. Under the umbrella of motivation to enter and remain in the teaching profession can be viewed on a loose continuum. perspectives of task). 1997. This agency can greatly influence a teacher’s commitment to the profession (Bandura. Does the relationship between self-efficacy and commitment differ for teachers from diverse regions in the world? . researchers could conclude the strength of commitment and the onset of emotional burnout by examining a teacher’s perception of the impact that they can make on their own environment. Knoblauch & Woolfolk Hoy. It is important to understand what exactly is being measured in a self-efficacy scale. self-esteem. While it is likely that self-efficacy appraisals will be slightly inaccurate (e. 2001). 1997) envisioned. a scale that measures self-efficacy on a 5-point Likerttype scale does not offer as much variability as a scale based upon a 0–100 scale because respondents tend to avoid extreme positions. Without following up or interviewing participants within a temporally acceptable time frame (e. 2009). Basing decisions and interpretations of self-efficacy research upon studies that utilized inaccurate conceptualizations and operational definitions of self-efficacy beliefs might not only bias the outcomes of such processes. intention) compared to a negative orientation (e.. burnout.. may give way to inaccurate conclusions about the nature of selfefficacy and its role in achievement. and the strategies that might make engagement in the task a successful endeavor (Bandura. 2014).. or commitment. While weak and maladaptive self-efficacy beliefs can be detrimental to a teacher’s longevity (Klassen & Chiu..g. When trying to interpret high versus low ratings on individual items. In order to make an accurate appraisal.g.

g. turnover). the following inclusion criteria were established: 1. and commitment. Teacher Sense of Efficacy Scale. a. 2001)). A mixture of teacher self-efficacy and terminology related to commitment and burnout in studies that utilized popular instruments to measure self-efficacy (e.. burnout..6 S. Both published and unpublished studies. 5. Searches returned thousands of published articles. commitment. The search All of the studies in this meta-analysis came from keyword-guided searches from online journal and dissertation databases and from searches guided by citations in relevant studies. and turnover. The following search terms were utilized in the search for studies: 1. enter teaching). Does the relationship between self-efficacy and commitment differ for studies that accurately measure self-efficacy (according to Bandura. focusing on manuscripts. or negatively as burnout. Burley / Educational Research Review 15 (2015) 1–16 5.. for analysis. teachers. Inclusion To be included in this meta-analysis. Gibson & Dembo. Additionally. commitment.. significance. Tschannen-Moran & Woolfolk Hoy. ProQuest Dissertations and Theses. Of the studies that were returned in our searches. but not the other (e. and subgroup effects. job satisfaction). 2. Does the relationship between self-efficacy and commitment differ when measuring commitment with a positive orientation (e. 2010). burnout). attrition.. only those written in English were retained. 33 independent studies from 27 manuscripts were retained. retention. and dissertations from 1980 to 2013: ERIC. 3. The following databases were searched. 1984. Does the specificity of self-efficacy measures and accuracy of conceptualizations influence the effects of the relationship between self-efficacy and teacher commitment? 2. PsycINFO. More specifically. and conference presentations..g. Studies reported relationships between teacher self-efficacy and commitment that could be ultimately converted to effect sizes meaningful to a multiple regression meta-analysis. (2) locating all studies that might fit. 31 were published manuscripts and two were dissertation projects. 2. and Google Scholar to ensure coverage of studies left out in previous searches. self-efficacy). correlation coefficients to Fisher’s Z) and test Q and association parameters (presented as χ2 estimates). Dellinger et al. a majority of them were removed on the premise of a qualitative framework (e. (4) coding these studies based upon a code sheet into a database. Teacher Efficacy Scale. Fisher Z. 2012) were utilized to calculate effect sizes (e. SPSS (v. intention) compared to a negative orientation (e.g. were returned. and commitment. . 3.g.g.g.g.R. Of these 33 studies. Commitment could either be interpreted positively as intention. 1997) compared to those that inaccurately measure self-efficacy? 7.g. Those that utilized a quantitative framework were retained.2. retention. and appropriately weighted.. Studies were completed between 1980 and the present. (3) filtering the studies to keep only those that fit within the pre-established criteria.g. (5) calculating values that can be used in meta analyses (e. The procedures for this study followed the steps outlined by Lipsey and Wilson.. In the end. attrition. 4. and (6) conducting the statistical analyses testing for homogeneity. Broad searches of published and unpublished literature on teacher commitment and attrition were conducted using several strategies and databases. A mixture of teacher self-efficacy and terminology related to burnout (e. Multiple software suites were utilized to carry out this meta-analysis. and nearly 300 dissertations and conference proceedings on the topics of self-efficacy... but many of them were a function of noise within the search algorithms. studies were returned that mentioned the relationship between self-efficacy and commitment in the manuscript... Studies involved preservice and inservice teachers’ perceptions and confidence in ability (self-efficacy) as it relates to their subsequent commitment to the teaching profession. A mixture of teacher self-efficacy and terminology related to commitment (e. intention. dropout. 2. Rothstein. Studies were written in English. (2001): (1) establishing a reason for the study. Table 1 summarizes the studies retained for inclusion in the study. attrition)? 6. 2. Hong.1. Methods The current meta-analysis employed techniques proposed by Lipsey and Wilson (2001). & Cohen.g. including peer-reviewed journal articles. theses. dissertations. inverse variance weight). studies containing one of the components (e. withdrawal. H. Microsoft Excel and Comprehensive Meta Analysis (Borenstein. While international studies were specifically coded for nationality of sample. 2008. but focused the analysis on selfefficacy and some correlate of commitment (e. 22) was utilized to facilitate data entry into databases and running the multiple regression analysis.g. Teachers’ Efficacy Beliefs System. Chesnut.

344 SE measure Acc. or publication. America Europe N. Coding To examine the relationship between teacher self-efficacy beliefs and their commitment to the profession.350 Journal Journal Journal Dissertation Journal Journal Journal Journal Journal Journal Journal Journal Journal Journal Journal Journal Journal Journal Journal Journal Journal Journal Journal Journal Journal Journal 179 379 301 215 3715 364 49 247 724 558 379 203 211 394 608 595 294 1282 530 277 83 159 490 322 244 1213 Preservice Preservice Preservice Preservice Inservice Inservice Preservice Inservice Inservice Inservice Preservice Preservice Preservice Preservice Inservice Inservice Preservice Inservice Inservice Inservice Preservice Inservice Inservice Inservice Inservice Inservice N. America Asia Europe Asia Asia Europe Europe Europe N. TSES – Modified TSES – Original TSES – Modified Yes Yes Yes 3 3 3 TES TSES – Original TSES – Translated TSES – Original OSTES – Modified TES TSES Author Created TSES – Original TSES – Original TSES – Original TSES – Original TSE TSE FIT – Choice Not Described GSES – Shortened Author Created Schwarzer. and Bassler (1988) Madden-Szeszko et al. Hoover-Dempsey. 1999). Given that specificity was coded on a 5-point scale. 4. Items at level 3 will lack a clear domain. Chesnut. breadth of self-efficacy measures. Accuracy of self-efficacy items: Conceptually accurate or inaccurate (Bandura. 6.570 0. America N.610 −0. 5. Type of motivation to enter or remain in the profession: Commitment or burnout. I can affect students’ learning).160 0.403 −0. Accurate questionnaire items will reflect what a respondent can do.R.290 −0. studies that remained after the final selection screenings were coded. These items will not reflect intentionality. relationship) between self-efficacy and commitment. (2012) – 1 Klassen et al. . Evers. Author (year) Type Bruinsma and Jansen (2010) Chesnut and Cullen (2014) Rots. were placed in the database. H. Descriptive and substantive features. and Vermeulen (2007) Evans and Tribble (1986) Klassen and Chiu (2011) Rots and Aelterman (2009) Rocca (2005) Chan et al. While disagreements initially arose.310 0.310 0.420 −0.330 0. 1997. nationality of teacher sample. The study features utilized in the analyses were coded in the following way: 1. Type of study: Dissertation. GSES – General Self-Efficacy Scale (Wang. Specificity of items: Scale of 1–5 indicating global (1).g.g. Minute disagreements on the specificity of the self-efficacy were the most common. (2000) Punch and Tuettemann (1990) Louis (1998) Chwalisz et al.210 0.S. America Asia N.270 0. self-esteem. the discrepancies tended to arise when one author coded the study a 4 and the other a 5. and Tomic (2001) Chan (2002) Chan (2008) Evers.160 0. conference. (2012) – 2 Klassen et al. America Burnout Burnout Intent Burnout −0. 1984). or outcome expectancies (Bandura. behaviorally specific (3).310 −0. & Daytner.300 0. locus of control. 2006).. America Commit Commit Commit Intent Commit Commit Burnout Commit Burnout Burnout Commit Commit Commit Commit Burnout Burnout Commit Commit Burnout Burnout Burnout Burnout Burnout Burnout Burnout Burnout 0. Origin of data: Countries were coded and combined into continents.3. and Tomic (2002) Friedman (2003) Skaalvik and Skaalvik (2007) Brissie. Items at level 1 will have no clear domain or clear behavioral task. and may look like questions in measures of self-concept or locus of control (e. thesis. OSTES – Ohio State Teacher Efficacy Scale. America Australia N. 2001). (2012) – 3 Klassen et al.330 −0. the authors met to discuss and mutually decide upon a rating. accuracy of self-efficacy measures.421 0.. Aelterman. such as type of population. * Full scale could not be retrieved for coding. 2006). 2007 Betoret (2006) Betoret (2009) Brouwers and Tomic (2000) Klassen et al.430 Dissertation Journal Journal Journal 73 574 528 316 Inservice Inservice Inservice Inservice N. (2008) Coladarci (1992) Fives et al.450 0.350 0. Spec. 1999). TSE – Teacher Self-Efficacy (Schwarzer. and the correlation (e. Bong.340 −0. America Europe Intent Commit Commit ES (r) 0. 3. (2012) – 4 Schwarzer and Hallum (2008) – 1 Schwarzer and Hallum (2008) – 2 Watt and Richardson (2007) Chapman (1984) Shen (2009) Brouwers. 2. Type of population: Preservice or inservice teachers. the iterative process of meeting and realignment facilitated an interrater agreement of 93% on the codes for 10 random studies. 2006. America Europe Asia Asia Europe Europe Australia N.170 −0. Schmitz.160 0. In situations where there was a discrepancy in coding. Brouwers.480 0. The authors of this study trained with the coding scheme before manuscripts were coded for this study. 2. TES – Teacher Efficacy Scale (Gibson & Dembo. America N.230 0.310 −0.070 −0.168 −0.130 −0. behaviorally and contextually specific (5) self-efficacy items. America N.280 −0.322 −0. (1992) Journal Journal Journal N Sample Origin Outcome 198 209 209 Preservice Preservice Preservice Europe N. Burley / Educational Research Review 15 (2015) 1–16 7 Table 1 Studies included in the meta-analysis. Vlerick. America Europe Europe Europe N.350 0. 1999 Author Created Author Created Not Described Author Created Author Created No Yes Yes Yes Yes No Yes No No Yes Yes Yes Yes Yes Yes Yes No No No Yes Yes Yes Yes No Yes No 1 3 3 3 3 1 3 3 4 5 3 3 3 3 1 1 1 1 1 3 1 3 5 * 5 1 TES Author Created Author Created Author Created No No Yes No 1 1 2 1 TSES – Teacher Sense of Efficacy Scale (Tschannen-Moran & Woolfolk Hoy.

The large Q statistic value indicated that this collection of studies was quite heterogeneous (Q = 295. While this aspect of a meta-analysis may be its strength.05 level.122 preservice and inservice teachers from North America. H. However.. p < . The fail safe N is a calculation of the number of studies with little or no effect size that would be needed to be included in the analysis to reduce the overall correlation effect size to a non-significant level. the removal of this information could weigh heavily on the interpretations and directions of future research in this field. Table 2 reports the fixed and . it was first determined whether these effect sizes were independent of one another before utilizing both.. The χ2 threshold was determined by calculating degrees of freedom for each level of variance. Europe. As many of the studies on teacher self-efficacy and commitment were conducted from a qualitative framework. Lipsey & Wilson.g. Disaggregated degrees of freedom for each within category (QW) were calculated by kwithin − 1. Burley / Educational Research Review 15 (2015) 1–16 but will have a clear behavioral task (e. 33). Degrees of freedom for between variance (QB) were calculated by j − 1. Due to the collection procedures in meta-analyses. at the time of data collection. Chesnut. a study that utilized three different preservice teacher samples from different countries reported three different correlation coefficients between selfefficacy and commitment.R. 2006). 2001). the sample of included studies might be biased.. Therefore. 3. with j being equal to the number of categories in moderating variable(s). Asia. Findings A total of 33 studies qualified for inclusion in the analysis. Lipsey & Wilson. some confidence is warranted when evaluating this study’s outcomes. with k being equal to the number of studies in the sample (e. In order to account for this file drawer problem. The non-response of some of these individuals undoubtedly limited the scope and power of this study.g. I can provide realistic challenges in mixed ability classes). Failing to include these studies in the analyses biases the results in favor of those that were fortunate enough to be published (Rosenthal. Using the appropriate degrees of freedom. Effect size calculations and statistical analyses Since the relationship between self-efficacy and commitment is common among all studies in this review. the Q statistics were compared against corresponding χ2 thresholds for significance at the p < 0. Another limitation is that the researchers selected only those studies written in English. In cases were studies reported multiple effect sizes (e.67. While qualitative analyses offer a great deal of knowledge to any field of research. 1991). Meta-analyses are unique in that they offer a systematic investigation of the literature within field. For example. a retrospective review of 303 metaanalyses using the English language restriction found no systematic bias (Morrison et al. The 33 studies covered 16. fixed and random main effects were calculated. The overall weighted effect size (pooled correlation) was 0. df = 32.4. Items at level 5 will have a clear and relevant domain of functioning (Bandura.g.026 for random effects). they were coded as three different effect sizes. 2012). Summing the disaggregated degrees of freedom for each group results in the same degrees of freedom for the QWtotal. Significance of moderating variables was calculated by comparing the total (QT).317 for random effects (SE = 0. 2004.. The studies that fail to be published are stored away and typically never see the light of day. it is also its weakness.. according to Thornton and Lee (2000). 2004. 2001). I can control disruptive behaviors of students in my classroom). 2001). a fail safe N was calculated. with k being equal to the number of studies in the sample and j being equal to the number of categories in moderating variable(s). qualitative studies were removed from the analysis.g. Future reviews should meta-analyze the results of these qualitative studies so that the voices of these researchers will not be overlooked. within (QW). In situations where a measure has varying levels of specificity. they could not be quantified to the extent needed for inclusion. and Australia. Degrees of freedom for total variance (QT) were calculated by k − 1. Considering studies with a small. The reader should be aware that because of this limitation. Total degrees of freedom for within variance (QWtotal) were calculated by k − j. the Pearson correlation coefficient (r) was used as the effect size statistic. studies that remain unpublished are ultimately left out. but nonzero correlation effect size.318 for fixed effects and 0. 2012). Effect sizes were calculated using the Pearson r to Fisher z transformation (Lipsey & Wilson. Still. but provides a foundation upon which group level significance testing for homogeneity of variance can be calculated. Limitations of current meta-analysis Because this study required quantitative effect sizes. 2.8 S. Unfortunately.. The multiple regression was calculated using a weighted regression analysis in SPSS (v 22) and by then recalculating the standard errors and converting to z significance values (DeCoster. and between (QB) variances to a corresponding χ2 threshold (DeCoster. r). it would require 54 studies to lower the estimated correlation to a level that is no longer significant.5.001). with that domain being some aspect of teaching (e. all meta-analyses that do not review all studies are similarly biased. To answer the first research question. Second. The main effects and subgroup analyses for the subsequent research questions were conducted first using Microsoft Excel and then verified by using comprehensive meta analysis (Borenstein et al. ratings were averaged and rounded. 1979. Given that each of these correlations was independent from the other. yet failed to report correlational information. the researchers deemed translating the studies as cost and time prohibitive. with kwithin being equal to the number of studies in each category of the moderating variables. making possible the levels of 2 and 4.008 for fixed effects and 0. despite this limitation. 2. attempts were made to contact the authors of older studies that utilized measures of teacher self-efficacy and commitment.

The lack of significant difference between groups provides evidence upon which we can validate the decision to examine burnout and commitment as measuring a similar construct.001 Test of null 95% CI Z P Lower Upper Test of homogeneity Q-value df (Q) P-value 41.81 268.011 29. However. intention.001 random effects models for the main effect.001 .307 . Table 4 presents the summary results of the analysis by orientation. df = 18.331 . commitment.3).033 13.001 .001 <. df = 13.001 <.342 . Table 5 presents the summary results of the analysis by conceptual accuracy.001).348 . df = 1.3 17 19 .2 <.322 .294 .270 . k Fixed Random ES 33 33 . with accurate self-efficacy measures reporting an effect size of +0.001). p < .029 . df = 16.001 <. p < .413 . entrance). the studies were divided into two major categories: (1) positive orientation (e..346 . This suggests that the variance between cases within this study is larger than could be explained by sampling error.06.. the results of the random effects model provide a wider perspective. 2006) definition of self-efficacy items and manner in which items should be phrased.009 16.001 <.301 . Instead.026 Variance . This means that the difference in effect sizes between preservice and inservice teachers was not significant. df = 1.271 . Initial results from the fixed analysis suggested significance between group differences (QB = 21.010 .001 <.67 35. as did the variance within inservice teachers (QW = 268. Both models.018 .238 . and (2) negative orientation (e.17 164. With regard to the conceptual accuracy of the self-efficacy scale being utilized. results from the random effects model indicate no significant difference in effect size as a function of the orientation of the commitment measure (QB = 1. the studies were divided into two major categories: (1) conceptually accurate (n = 21) and (2) conceptually inaccurate (n = 12).16 1. Table 3 presents the summary results of the analysis by teacher level.334 .21 1.73 <. Positive: QW = 110. Results suggested significance between group differences (QB = 43.324 . The determination of conceptual self-efficacy was based upon Bandura’s (1997.82 <.001 =. With regard to the teacher level.05 <.05). df = 1.289 .6 14 19 .001 .82 31.255 .354 and inaccurate Table 3 By teacher level.67 32 <.001 .001).304 .65 25.05 <. attrition).267 .2).317 SE .001 =. Chesnut.001).308 .362 295.78 <.001 <.09 9.S. df = 1. p < 0. Burley / Educational Research Review 15 (2015) 1–16 9 Table 2 Overall effect sizes. p < .32 13 18 1 13 18 1 <.g. With regard to the orientation in which researchers measured the motivation to enter and remain in the teaching profession. however.21.001 <.17 164.63 12. Results suggested nonsignificance between group differences (QB = 1. p = . Model Fixed Random Group Preservice Inservice QB Preservice Inservice QB k ES SE Test of null 95% CI Z P Lower Upper Test of homogeneity Q df(Q) P 25. that difference was not great enough to establish significance.001 <.82 <. indicated heterogeneity within each of the two groups (Negative: QW = 164.g.21 0. This means that the differences in effect sizes between orientations of measures were significant. Results from the random effects model corroborated this finding of a non-significant difference between preservice and inservice teacher groups. the results from the random effects model do not corroborate this finding. .338 14 19 . p = .R. p < . In order to account for the variance in effect sizes. the studies were divided into two major categories: (1) preservice teachers (n = 14) and (2) inservice teachers (n = 19).023 . H. While inservice teachers had a higher pooled correlation between self-efficacy and commitment. Model Fixed Random Group Positive Negative QB Positive Negative QB Test of null 95% CI Z P Lower Upper Test of homogeneity Q df(Q) P 110. Given that there will always be more variance in the population than what can be accounted for in a sample.000 .16 21 110. burnout.038 10. The variance within preservice teachers indicated heterogeneity (QW = 25.16. yet at distinct ends of the same continuum. p < .81.330 .318 .376 17 19 .386 k ES SE Table 4 By orientation.001.356 .001 .06 16 18 1 16 18 1 <.259 . df = 18.94 <.81 268.001 <.268 . key methodological and sample characteristics were used to model some of the variation.008 .29 8.001 =.

1.336 . df = 20.333 self-efficacy measures reporting an effect size of +0.354 .001).001 . Overall.280 21 12 . p < 0. the conceptual accuracy of scales can influence the magnitude of the correlations found between self-efficacy and teacher commitment.56 21. Model Fixed Random Group Accurate Inaccurate QB Accurate Inaccurate QB k ES SE Test of null 95% CI Z P Lower Upper Test of homogeneity Q df(Q) P 130.001 <. Figure 1 demonstrates this heterogeneity. p < 0. Fig.026 .304 . similar to the variance within inaccurate studies (QW = 121.402 . Burley / Educational Research Review 15 (2015) 1–16 Table 5 By conceptual accuracy.44 20 11 1 20 11 1 <.009 .169 .001 <.253 . Comparison between conceptually accurate and inaccurate measures. the studies were divided into four major categories. .33 <.001 .903 <.256. df = 1.001 <. This indicates that scales employing an accurate measure of selfefficacy had a significantly higher correlation with commitment to the teaching profession than did studies that employed scales that inaccurately measured self-efficacy beliefs.233 . df = 11. (2) Europe (n = 11).256 .05).371 . Results from the random effects model corroborate this significant finding (QB = 4. (3) Asia (n = 8).R.67.62 130.67 4.001 <. H. and (4) Australia (n = 2).001).001 <.38 121. With regard to sample location.043 13.05 21 12 .10 S.38.354 .001 <.67 43.001 <. Chesnut.38 121. This is consistent with Bandura’s (1997) claim that self-efficacy beliefs can better predict outcomes when the beliefs are accurately measured. each describing the continent in which the participants were located: (1) North America (n = 12).88 5. p < 0.012 38. significant variance within each of these categories requires further examination. The variance within the accurate studies indicated heterogeneity (QW =130.44. however.

p < 0. Results suggest significance between group differences (QB = 106.171–.001 <. df = 10. .017) [.53 6 (3427) .160 .001 .046 .3123 (.57 10. Europe Asia Australia QB N. p < 0. p > 0.03 7.001). because it was a small sample. 2001).001 6 (5170) .014 .05) groups.014) [. Model Fixed Random Group k N.001 <.114 .10. Meta-analyses are used to find not just between group differences.28 38. df = 11.28 38.302 .2679 (.10 (5) P < 0.05 9 (3075) .301] 98. p < 0.001). p < 0.001 <. Amer. df = 5.1562 (. results indicate that there is significant between-group effect sizes with much within-group variance.05 df (Q) 6 P-value <0.001 <. Contrary to the previous analyses. further investigation was warranted.001).294–.68 29.g.413 . Amer.99 125.05 2 (852) ..034) [.57 (1) P > 0.001 2 (971) .234 .285–.259 . df = 3.033 23.23. A 2 × 2 comparison analysis was conducted by dividing orientation of measure (e.38 (8) P < 0.33 (1) P > 0.381 .05 2 (868) . p < 0.018) [. the further disaggregated groups began to demonstrate levels of within group homogeneity of variance.54 (1) P > 0.013 .001 . df = 1.60. Overall. further investigation was warranted.445] 7.29 (5) p < 0. Results from the random effects model corroborate these findings. Country of origin North America Europe Asia Australia QB # Cases (sample) Effect size (SE) 95% CI QW (df) P-value # Cases (sample) Effect size (SE) 95% CI QW (df) P-value # Cases (sample) Effect size (SE) 95% CI QW (df) P-value # Cases (sample) Effect size (SE) 95% CI QW (df) P-value Accurate conceptualization Inaccurate conceptualization 6 (1759) .56 (5) P > 0. p < 0.223] 0.S.001).001).001).3980 (. .326 P <. Europe (QW = 72.328 .01 <.377 .60 2.305] 2.2379 (.001 <. Results for the Australia group suggest homogeneous variance between the cases (QW = 2. positive.001 Cells are not optimal sizes for the analysis. and in the Inaccurate by North America group (QW = 98. Significant within group variance emerged in the Accurate by Europe group (QW = 64.414–.53. Table 7 Self-efficacy moderated by accuracy of conceptualization × origin of population.234–.57 56.18 11 10 7 1 3 11 10 7 1 3 12 11 8 2 .340] 20.99. Results suggested significance between group differences (QB = 56.001 <.429 .089–. H.305 .06 9.001 0 (0) N/A N/A N/A N/A Test of homogeneity Q-value 106. this information does make a strong argument about the forces behind the observed relationship between self-efficacy and teacher commitment.012 .032 [.17 <.486] 64.23 72.038 .23 72.01 Table 6 presents the summary of results of the analysis by location. p < 0.351–. Table 7 presents the summary of the results of the analysis.024) [. df = 5. p < 0. studies that accurately conceptualize and measure self-efficacy should be more consistent.306 . df = 8.305 .170 .282 .296 12 11 8 2 .05).001).001 <.33 24. Burley / Educational Research Review 15 (2015) 1–16 11 Table 6 By location. df = 6. As there was much variance within the groups in both the conceptually accurate and by orientation analyses.20 4.327 . however. as one might hope for with an ANOVA.404 .452 .96 6. but also homogeneity within groups indicating a goodness of fit (Lipsey & Wilson. Europe Asia Australia QB ES SE Test of null 95% CI Z P Lower Upper Test of homogeneity Q df(Q) 125.056 7. A 4 × 2 comparison analysis was conducted by dividing location effect sizes by the accuracy of self-efficacy conceptualization. in the Accurate by Asia group (QW = 20.237 .60 2. While it is logically understandable to have within group variance in studies that fail to accurately conceptualize self-efficacy as it relates to teacher commitment.4503 (.3575 (.421] 1. and Asia (QW = 38.223 .R.277 .57.038 . however. As there was much variance within the groups in both the conceptually accurate and by location analyses.034) [. Chesnut. this value should be interpreted with caution. df = 7. Within group differences suggested heterogeneous variance for the North America (QW = 125.

will provide only predictive power when the outcome is in the same context or domain. Discussion The findings of this meta-analysis indicate that preservice and inservice teachers’ self-efficacy beliefs are positively related to their commitment to the teaching profession (ES = +0. df = 6. For example. Alone. self-efficacy items tend to reference what outcomes an individual believes he/she can enact.202] 34. Chesnut. however. Selfefficacy measures that incorporate highly specific items. Dellinger et al.3. theoretically and pragmatically. they are still significant.g. Results suggest that instrument specificity and conceptual accuracy were both statistically significant predictors of the relationships between teacher self-efficacy and commitment to the teaching profession.351] 19. However. Why are the relationships between negatively oriented measures of commitment and inaccurately measured self-efficacy beliefs nearly as large as positively oriented measures of commitment and accurately measured self-efficacy beliefs? Are there characteristics of negatively oriented measures of commitment that better reference the identification and orientation toward the teaching profession? While the groups show decreased levels of within group variances as indicated by lower QW values.001). a global item might read.00 <0. On the global end of the spectrum.05 9 (3063) . commitment scales that utilized a negative orientation (e. Table 8 presents the summary of the results of the analysis.001 <0. a specific item might read. when accurate measures of self-efficacy were utilized.001 7 (3752) .001 Test of homogeneity Q-value df (Q) P-value 96. we are able to attain the amount of variance that self-efficacy explains in commitment.5). self-efficacy can account for 10% of the variance in the decision that teachers make to enter. Moderate: r = 0.330 (.435] 94.123–. This significant within group variance indicates that more moderators should be able to account for the variance. The predictors accounted for 22.309–. Large: r = 0. Table 9 summarizes the regression information. p < 0. H.25 2. This finding is statistically and practically significant.008 0.57 (4) p < .32).020 4. While these reflect the prior analyses. 4.015) [. On the specific end of the spectrum. self-efficacy items tend to not mention the accomplishment of an outcome.12 S.R. these relationships were the highest.375–. Cohen (1977. Results suggest significance between group differences (QB = 96. The specificity of self-efficacy measures has been a topic of major discussion within the field (Bandura. we utilized a multiple regression. 2006.99 (8) P < . Given the claims that contextually specific and conceptually accurate self-efficacy measures should more strongly predict the relationship with a contextualized outcome (in this case teacher commitment).013) [. As a moderator.” This item represents something that an individual believes he/she can accomplish. The specificity of a self-efficacy measure tends to span a global to specific continuum. 2006) suggests that global self-efficacy measures will suffer from predictive power when the outcome is specified or contextualized. Additionally. if half the population were categorized Table 9 Instrument specificity and conceptual accuracy as predictors of the relationship between self-efficacy and commitment. remain.284–.020) [. but focus more on the behaviors and contexts of an individual’s engagement. and leave the profession. “I can provide realistic challenges in mixed ability classes.38 3 <. “I can affect student learning. The values of the ES and r2 represent strong effects.040 0. 1997.034 0. further questions are raised. but ignores the small details of the actual behaviors that lead to student learning and the context in which student learning must occur. these potential moderators are currently unknown.001 Negative measure # Cases (sample) Effect size (SE) 95% CI QW (df) p-value # Cases (sample) Effect size (SE) 95% CI QW (df) P-value QB negative) effect sizes by the accuracy of self-efficacy conceptualization. By squaring this value. burnout) had higher relationships with teacher self-efficacy than positive measures of commitment..38. For example.9% of the variance in the relationship between self-efficacy and commitment.015) [. 2008). Variable B Adjusted SE of B Z (β replacement) P Instrument specificity Conceptual accuracy 0.405 (.163 (. Burley / Educational Research Review 15 (2015) 1–16 Table 8 Self-efficacy moderated by accuracy of conceptualization × measure orientation.30 (6) P < .05 ..44 (11) p > . Using the binomial effect size display (BESD). Measure of orientation Accurate conceptualization Inaccurate conceptualization Positive measure 12 (6941) . 1988) would classify this as a moderate effect size (Low: r = 0.” Bandura (1997. 2006.313 (. Bong.1.001 5 (2366) .342] 50.

Rosnow. Because of this.. they do suggest that there is a non-significant difference between the role that these beliefs play in commitment to the teaching profession between these two groups. 2006) and Bong (2006). 2006).. Bandura (1997. while the pooled correlation between inservice teachers’ self-efficacy beliefs and commitment to teaching was larger than preservice teachers’. what is it about negatively oriented measures of commitment that enhance the relationship? Are measures of burnout and emotionality more closely aligned with the processes of leaving the profession than psychological identification with the process of entering and remaining in the profession? Researchers would serve the field well by further analyzing the negative and positive orientations of commitment measures. When we examined the role of conceptually accurate and inaccurate self-efficacy measures by country. but it may also influence the likelihood of publication and the effectiveness of responsive interventions. another indication that the relationship between self-efficacy and teacher commitment is a strong one. the trend still poses an interesting question about how we interpret measures of commitment and burnout. the ES between self-efficacy and the extent to which teachers are committed to the profession hypothetically represents a gain of 32%. The confidence that preservice teachers have toward their ability to be successful in teaching practices significantly influences their commitment to the teaching profession. however. and Australia. 1997. Given that the correlation between negatively oriented measures of commitment and inaccurate conceptualizations of self-efficacy was nearly as strong as between positively oriented measures of commitment and accurate conceptualizations of self-efficacy. The primary issue surrounding the measurement and interpretation of self-efficacy beliefs began with the development of popular self-efficacy questionnaires created by Rotter (1966) and Gibson and Dembo (1984).. 2006) and Bong (2006).g. Schunk & Usher.g. we can conclude that measures that employ accurate self-efficacy conceptualizations will have a larger correlation with commitment than inaccurate conceptualizations.. Failing to utilize conceptually accurate self-efficacy measures will not only influence observed relationships with outcome variables. Are the differences related to the domain of this study (e. 2011). Alternatively. accuracy and specificity) in understanding the relationships that self-efficacy beliefs share with commitment to the teaching profession. Chesnut & Cullen. H. given the findings of this study. Bandura (1997. Chesnut. they will fail to account for much of the variation in the outcome. among others. & Rubin. While the results from this metaanalysis cannot corroborate or refute these claims. In addition to the need for adherence to conceptual accuracy. self-esteem. we suggest that researchers not overlook the value of selfefficacy beliefs in preservice teachers for engaging in teaching practices. Asia. 2005).S. this study confirms and corroborates the research suggesting that preservice and inservice teacher self-efficacy beliefs can be used to predict commitment to the teaching profession (e. 1997). conceptually accurate self-efficacy measures had significantly higher correlations with commitment than conceptually inaccurate measures. 2014. While questionable self-efficacy measures were created before and have been created since. The findings in this study indicate that conceptually accurate self-efficacy measures report significantly larger correlations with commitment to teaching than conceptually inaccurate measures. From the results. the correlations between self-efficacy beliefs and commitment to teaching were largest in Europe and smaller in magnitude for North America. we found that regardless of the country.R. A critical interest of educational researchers is trying to understand how their ideas. 2006) also suggested that self-efficacy instruments be as specific as necessary for the study. Previous studies on preservice and inservice teacher self-efficacy beliefs suggest that inservice teachers are likely to have more accurate appraisals than preservice teachers (e. In other words. have suggested that the inaccuracies within self-efficacy measures can be the reason that these purported measures of the construct fail to predict outcomes.g.. Unlike the findings in previous self-efficacy studies that downplay the link it might have with performance and choice outcomes.32 represents an improvement rate of 34% to 66% (Rosenthal.g. they will not provide much for generalization of results. When measuring self-efficacy beliefs. the problem that self-efficacy researchers have expressed is that the measure of self-efficacy is unique and not the same construct as locus of control. Klassen & Chiu. While the random effects model did not corroborate this finding. 2000). which many of these instruments unintentionally measure. an ES of . Further studies should examine the reason for these international discrepancies. Additionally. teacher commitment) or are they the result of some underlying social or cultural aspect that has yet to be reported? Many years after introducing self-efficacy. or self-concept. suggesting that if self-efficacy measures are too general. 2011). and constructs transfer to international populations with different cultures and socialization trends. Woolfolk Hoy & Spero. even if they seem to be something beyond the respondents’ repertoires. this question is expanded. theories. this study implements and examines the fundamental characteristics of self-efficacy measures (e. We can additionally conclude that negatively oriented measures of commitment will have a larger correlation with self-efficacy. when examining the 2 × 2 matrix of accuracy by orientation. the findings (from fixed effects) suggest that studies utilizing negatively oriented measures of commitment will report significantly larger effect sizes. it was not significantly different. Building upon these . Bandura began to specifically give attention to the measurement and interpretation of self-efficacy instruments (Bandura.g. if they are too specific. Burley / Educational Research Review 15 (2015) 1–16 13 as low commitment and the other as high commitment (with each new variable split at the median) and high self-efficacy and low self-efficacy. Other researchers have offered similar thoughts (e. In our study. it should be a primary concern for the researchers that they adhere to self-efficacy instruments as the construct is defined and exemplified by Bandura (1997. Consistent with the theoretical perspectives surrounding self-efficacy beliefs (Bandura. researchers should further investigate the differences in conceptualization and operationalization of commitment measures as positive and negative measures of commitment do not share common relationships with self-efficacy. According to the findings from this study. Therefore.

63(4). A. (1974). Meta-analysis notes. coping resources. Educational Psychology. R. Educational Psychology. behavioral) self-efficacy measures are more likely to culminate in higher relationships than those that are global and/or inaccurate. W. (1986). M. *Chapman. & Tomic.1. Englewood Cliffs. J. Olivier. *Betoret. D. 287–305). H. (2009). (2001).. & Lent... (2008). Pajares & T. Is the motivation to become a teacher related to preservice teachers’ intentions to remain in the profession? European Journal of Teacher Education. Journal of Applied Social Psychology. Brown. L. & Olivarez. E. Kfir. L. Santa Monica. M. conceptually accurate and specific (e.tate. *Evers. A. 916–934. Self-efficacy beliefs of adolescents (pp. W. Utilizing accurate and specific self-efficacy belief measures can enhance the ability for researchers and teacher educators to meet the needs of preservice and inservice teachers. A. doi:10. (1997). 13(4).. 157–165. & Nasser. J. Mirroring Bandura’s and other researchers’ recommendations.14 S. 159–165.010. & Ellett. Greenwich.. and coping with stress. 11(4). (2010)..1559-1816. 489–525. ERIC Document Reproduction Service No. W. Self-efficacy beliefs of adolescents (pp.. *Brissie. D.84. (2008). D. *Coladarci. The Journal of Educational Research. J. *Bruinsma. these beliefs can be used to structure teacher education and professional development experiences.1348/000709902158865. (2000). & Pennell. inappropriate usage of self-efficacy has influenced the development and implementation of incorrect interventions (Klassen et al. 23(6). (2007).1. 185–200.. Psychological Review. The Journal of Educational Research. B. M. burnout. Stressors.. and support during the student-teaching semester. A. Altmaier. (1977).. Bandura. and role problems among teacher trainees and first-year teachers. C. Bandura. E. Causal attributions.. doi:10. (1976). Pajares & T. DeCoster. A. situational contributors to teacher burnout. doi:10. Cohen. and psychological distress among prospective Chinese teachers in Hong Kong. Self-efficacy beliefs of adolescents (pp. Hillsday. many have inaccurately utilized the construct and misinterpreted the influence and the value of self-efficacy in predicting peoples’ behaviors. Individual. W. J. H.1177/0013124510380724. Bobbett. doi:10. 45(3).. Stress. Fimian. A longitudinal study of teacher burnout and perceived self-efficacy in classroom management. Social cognitive theory: an agentic perspective. Bandura. Chesnut. Review of Educational Research. American Educational Research Journal. D. Urdan (Eds. H. S.. Bandura. 22(5).. *Betoret. Pascal. doi:10. Bandura. Effects of self-efficacy. King. 33(2). C. 645–658. & Tribble. 84(2). Englewood. In F. F. W. & Tomic. V. Organizational and personal predictors of teacher commitment: the mediating role of teacher efficacy and identification with school. R-2007-LAUSD). . (2006).. and differential policies. 227–243. D. A. Self-efficacy: toward a unifying theory of behavioral change. A. doi:10. (1977).. (2012). A. Burley / Educational Research Review 15 (2015) 1–16 perspectives. D. S. Comprehensive meta-analysis: A computer program for research synthesis [computer software]. The Teacher Educator. (2004). and perceptions of future work environment on preservice teacher commitment. *Chwalisz. 21(3).1080/01443410701491833.R. M. B.. Teaching and Teacher Education. 751–766. 2011.1111/j.. et al. job stressors and burnout among Spanish primary and secondary school teachers: a structural equation approach.03. & Blanton. 191–215.02.. Hamman. Cohen. Journal of Social and Clinical Psychology. doi:10. Lim. The construct known as self-efficacy has come to be widely studied in many domains.. 43(5). NJ: Biostat. M.09.1080/ 00220973. W. (2006).1016/j. Education and Urban Society. *Evans.. 557–569.1037/0033-295X. E.. D. Cox. 116–132. (2006). social support.2006. H. 80. (1997). Perceived teaching problems. W. Teachers’ sense of efficacy and commitment to teaching.. J. (1988). 2012). M. 31(7). T. W.. J. *Chan. (2014). New York: Academic Press. Journal of Occupational Behavior.9943869.. 82(2).887168. 377–400. Firestone. Chesnut. Teaching and Teacher Education. Brouwers. (Report no.1016/s0742-051x(96)00037-6. (1987). Predicting teacher commitment. K. Y.). Dellinger.1146/annurev. doi:10..191. CT: Information Age.psych. Urdan (Eds. *Fives. 29(1). 1474–1491.013. (2008). The Journal of Experimental Education. 72(2). & Presley. 16(2).1080/08878730. G. British Journal of Educational Psychology. 45–68. Self-efficacy in eliciting social support and burnout among secondary-school teachers. CT: Information Age Publishing. Pajares & T. burnout. Dimensions of teacher self-efficacy among Chinese secondary school teachers in Hong Kong.. NJ: Prentice Hall..13. and commitment to teaching among preservice teachers.stat-help. Preparing adolescents to make career decisions. McDonnell. Stress.. S. the findings from this study suggest that the accuracy of self-efficacy measures and the specificity of the items will predict the magnitude of the correlation between self-efficacy beliefs and commitment to teaching. *Brouwers. Greenwich.g... R.html> Retrieved 28.H. Borenstein.3102/ Does burnout begin with student-teaching? Analyzing efficacy.1016/j. S. D. While it has a stronghold in education. Teaching and Teacher Education. & Cohen. Urdan (Eds. doi:10. 52. F. (1992).1080/0144341022000023635. Hoover-Dempsey. 201–223). (1986). Staff burn-out. Lau.3102/0002831208318259.tate. B.).tb02683. S.). <http://www.2. self-efficacy. ed). P. & Russell. 26(4). (2011). Annual Review of Psychology.52. D. F. In F. J. J. D. That is. J. Teacher retention: the test of a model. Educational Psychology. K. J. *Chan. doi:10. J. A. Journal of Social Issues. Educational Psychology. self-efficacy instruments should be conceptually accurate and as contextually and behaviorally specific to the intended outcome measure as possible. Guide for creating self-efficacy scales. W. doi:10.2001. 24(3). References *Indicates references used in the analysis. Statistical power analysis for the behavioral sciences (2nd ed.2307/40539578. P. working conditions..x. 597–630. N. 130 243. (2002). D. NJ: Erlbaum.. (1984). 429–438. self-efficacy. Self-efficacy. Nie. Freeman and Company. Social foundations of thought and action: A social cognitive theory. 81–85. (2006).). W. W. Rothstein. A. & Hogan.2014. Toward a more nuanced understanding of new teacher attrition. A. and burnout among secondary school teachers in Spain. 60(4). self-efficacy cognitions.. The knowledge of self-efficacy beliefs can provide much information to researchers and teacher educators.2007. Teaching and Teacher Education. M. D. J. Freudenberger. (2001). Conroy-Oseguera. F. school resources. & Tomic. CA: Rand Corporation. 8(2). 519–539. G. Burnout and self-efficacy: A study on teachers’ beliefs when implementing an innovative educational system in the Netherlands. Bong. Fresko.. A. contextual. 106–112.. & Jansen. W. 181–194. P. *Chan. Additionally. Using global self-efficacy scales will result in less impressive predictability and may negatively impact one’s analyses and study results. doi:10. Measuring teachers’ self-efficacy beliefs: development and use of the TEBS-self. DeAngelis.. doi:10. & Bassler. emotional intelligence. D. 598–626. T. Wyatt. A. self-efficacy. Asking the right question: How confident are you that you could successfully perform these tasks? In F.1992. J. 28(2). Teacher commitment. *Brouwers. & Cullen. 1–26. J. Y. Analysis of the school preferred reading programs in selected Los Angeles minority schools. doi:10. (1993). O. K. 239–253. (2002). For example. American Educational Research Journal. New York: W. Armor. Statistical power analysis for the behavioral sciences (Rev. 307–337). R. 323–337. D. (1992). Evers. doi:10. 49(2). 30(1). (1988). Self-efficacy: The exercise of control.

191–215.. Teacher self efficacy.1007/s10212-012-0166-x. and performance. C. & Aelterman. Hoy. 79–122. Stress and Health. Introduction: Burnout and the teaching profession. Maslach. absenteeism. 23(7).. 63(2).. Siwatu. Burley / Educational Research Review 15 (2015) 1–16 15 *Friedman.. *Klassen. C. Teacher efficacy. 80(1). Preservice teachers’ culturally responsive teaching self-efficacy and outcome expectancy beliefs. D. (2011). 138–144. S. H. Social cognitive approach to career development: an overview.. turnover. 103(5). E. Relationships between self-efficacy. 76. and teaching context. Rosenthal. Teacher efficacy research 1998–2009: signs of progress or unfulfilled promise? Educational Psychology Review. graduates’ teaching commitment and entrance into the teaching profession. doi:10... D. Practical meta-analysis. Self-determination theory and the facilitation of intrinsic motivation. Chesnut. *Louis. 68–78.. M. R. F. International Journal of Technology Assessment in Health Care. (2005). W. social support and stress coping strategies in Chinese primary and secondary school teachers. & Valiante. M. *Schwarzer. International handbook of research on teachers’ beliefs (pp. School Effectiveness and School Improvement.R. & Daytner. E. Handbook of self-regulation of learning and performance (pp. Differences in teachers’ perceptions of personal control of positive versus negative student learning outcomes. Schwarzer. Contemporary Educational Psychology. 342–350. R. I. Schunk. Wilson. affect. Retrieved December 1. & Usher. The file drawer problem and tolerance for null results. Pajares. G.1016/j.1017/ S0266462312000086. J. I. J. (1984). NY: Hofstra University. M. 1–27.. & Vance. et al. Polisena. Vandenberghe & M. The occupational commitment and intention to quit of practicing and pre-service teachers: influence of self-efficacy. W. R.. M. Teacher education.. R. Brown.. D.01. R. American Psychologist. M. W. H. and occupational commitment in four countries. D. Teacher efficacy: a construct validation.. Husereau. *Rots. 28. (2009). teacher efficacy. Jackson. American Educational Research Journal. 569–582. Assessing self-efficacy for self-regulated learning. Y. E. CA: Academic Press. M. 4. M. 282–297). 24(4). R. D. (1966). H. J. The effect of English-language restriction on systematic review-based meta-analyses: a systematic review of empirical studies. B. . Aelterman. Teaching and Teacher Education. 1289–1309. 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. C. & Woolfolk Hoy. (2010). & Hallum. (1982). G. K. Correlates of psychological distress among secondary school teachers.. 57. N. D. Schunk (Eds. M. doi:10. In B. M. White Plains. S. Seattle: University of Washington..00359. Lent. 63–69. Understanding and preventing teacher burnout: A sourcebook of international research and practice. The Journal of Educational Research.1080/ 00220670903383069. & Chiu. The burnout syndrome in the day care setting. Preservice teachers’ work stress. 9(1). 27. New York: Routledge. (2010). 543–556.. Rotter. Boston: Pearson. Ryan. doi:10. Effects on teachers’ self-efficacy and job satisfaction: teacher gender. E. & Vandenberghe.. Schlechty. Gibson. M. social development. Wong. Y. P. Schunk. A. & Wilson. 212–229). (2009). (2007). R. W. D. Huberman. W.1037/a0019237.. W. Siu... Lipsey. Fiander. S. Lent. Organizational linkages: the psychology of commitment. Rosenthal. Rosnow. R. Newbury Park. R. L. C. I. contextual influences. 114–129. (1979). R. S. K. Schunk. Maslach Burnout Inventory: Educator survey. T. T.). Social Psychology of Education. A. S. Center for the Study of Teaching and Policy. 32. (1988). M. Journal of Vocational Behavior. (1982).. R. 6(2). 7. (1990). (1999). 214–221. G. A. Fives & M. Knoblauch. In R. University of Florida. Klassen.1080/0141192900160405.. & Hackett. NY: Longman. 23(1). European Journal of Psychology of Education. Career Development Quarterly. & Chesnut. A. J.. & Gordon. 21–43. J. Hong. (2000). (2003). Journal of Educational Psychology. (1986).. D.. (2007). 35–54). Teaching and Teacher Education. (2011). Siwatu.. R.. (2010). European Journal of Psychology of Education.. 26(8). Moulton. T. V. (2012).. 45(1).... (1989). Applied Psychology. 33(4). Madden-Szeszko. Predicting preservice agriculture teachers’ intentions to teach utilizing person inputs. S. A. S. *Punch. Thousand Oaks. “Maybe I can teach those kids.1002/smi.. 86(3). Measurement and Evaluation in Counseling and Development. from http://userpage. K. (1999). Teachers’ workplace: The social organization of schools. D. Pre-service and beginning teachers’ professional identity and its relation to dropping out of the profession. doi:10. choice. 310– 322. Morrison. In K. 129–138. K. The Phi Delta Kappan. F. Psychological Bulletin. Y. Guskey. Self-efficacy theory.).. Gill (Eds. (2009). (1998). (1994). Wentzel & A. 152–171. & Brown. & Pajares.. R. 638–641. R. 55(1). L. Wongsri. Schmitz. S. Teacher education graduates’ entrance into the teaching profession: development and test of a model. 1530–1543. J. (1981).. 106–112. & Deci. H. 44–51. M. Perceived teacher self-efficacy as a predictor of job stress and burnout: mediation analyses. A.002. F. Handbook of motivation at school (pp. Learning theories: An educational perspective (6th ed. Contemporary Educational Psychology. and attitudes toward the implementation of instructional innovation. A. doi:10. Journal of Educational Psychology. Vlerick. (1981).76. (1996). R. Self-efficacy and burnout in teaching: the importance of interpersonal-relations efficacy. (2007). (2000). G. B. & Rubin. & Pines. et al. O. (2003). R.. (2011). Hannok. 44(4). S. Clark. (2008). (2000). 28(2). Maslach. Journal of Teacher Education. Klassen.. New York: Routledge.x. J. DC: Center for Comprehensive School Reform and Improvement. self-concept. doi:10. Cambridge. 25(2). & Schwab. & Dembo. & Chiu.1023/A:1024723124467. R. Mowday. F. Jerald. demographic variables and general self-efficacy (unpublished doctoral dissertation).. Measurement of responsibility teachers assume for academic successes and failures in the classrooms..). Ingersoll. 100–113. Guskey. 70–80. New teacher_se. Teaching and Teacher Education.1006/jvbe. P. 279–300. 2013.htm *Shen. and well-being.1007/s10648-010-9141-8. The career development of preservice and inservice teachers: Why teachers’ self-efficacy beliefs matter. K. E.. M.2008. K. Teacher stress: the mediating role of collective efficacy beliefs. Teaching and Teacher Education. and job stress. San Diego.. E. CA: Sage. R. R. V. & Steers. Cambridge: Cambridge University Press. Contrasts and effect sizes in behavioral research: A correlational approach. Porter. 23(5). 36(2). 1–28. 102(3). self-efficacy. 1086–1101. Klassen. Generalized expectancies for internal versus external control of reinforcement. doi:10. R. Meta-analytic procedures for social research. *Rocca. Variables contributing to teacher efficacy: An examination of burnout. In H. Toward a unifying social cognitive theory of career and academic interest. (2012). R. (2008). Child and Youth Care Forum. doi:10. Hartley. 24. Rosenthal. Tze. & Vermeulen. (1991). M. Do academically able teachers leave education? The North Carolina case.1027. and outcome expectations (Doctoral dissertation).1464-0597. G. and turnover. 6(3). doi:10.1037/0022-0663. D. (1990).. Socialization of student teachers. NY: Routledge.” The influence of contextual factors on student teachers’ efficacy beliefs. Guskey.1994. Teaching and Teacher Education.569.fu-berlin. Rosenholtz.. L. Psychological Monographs. R. (2001). doi:10.. (2014). & Tuettemann. (1977).4. S. *Rots. A. 741–756. (2012).). M. & Woolfolk. K. Wigfield (Eds. doi:10. Huberman (Eds. S. Response format in writing self-efficacy assessment: greater discrimination increases prediction. C. 453–471. L. R. S. Betts.cedpsych. job stress. O. Zimmerman & D.S.. 16(4).2011. Effects of teacher quality of work life in secondary schools on commitment and sense of efficacy. T. M.. *Klassen. Believing and achieving (issue brief). 166–179. 369–382. Hempstead. R. years of experience. (2001). British Educational Research Journal.). T. M. Washington. K. CA: Sage.1111/j. A. doi:10.1229. UK: Cambridge UP.

Usher. Craven. H. T. L. 68(3). W. Woolfolk Hoy. 21(7). and enabling human potential (pp. (2000). Journal of Clinical Epidemiology.. Greenwich. P. doi:10. (2000). 71(3). (2005). 207–216. Towards a re-conceptualization of teachers’ self-efficacy beliefs: tackling enduring problems with the quantitative research and moving on. Tschannen-Moran. 17. Ware. 783–805. Dimensions of teacher self-efficacy and relations with strain factors. M. Connecticut: Information Age Publishing. 944–956. (2006b). B. Motivational factors influencing teaching as a career choice: development and validation of the FIT-Choice scale. The potential benefits of teacher efficacy doubts for educational reform.99. E. 46(1). L. (2006a). G.tate. E. (2008). Usher. 34(1). K. 443–463. W. doi:10. S. Tschannen-Moran. Thornton. Teacher and collective efficacy beliefs as predictors of professional commitment. 35(3). The Journal of Educational Research. and teacher burnout. W. 104. F. E. Burley / Educational Research Review 15 (2015) 1–16 *Skaalvik.). 549.. 275–314. E. H. 18. F. Urdan (Eds. 5–22. (2001).003. In R. A. 167–202. & Skaalvik. M. Skaalvik. International Journal of Research & Method in Education. A. Wheatley. M. 21. Teaching and Teacher Education. and retention.03. H.2012. E. The Journal of Experimental Education. The differential antecedents of self-efficacy beliefs of novice and experienced teachers. P. 183–193. 31(2). A. M. Publication bias in meta-analysis: it causes and consequences. M. A. F. Changes in teacher efficacy during the early years of teaching: a comparison of four measures. Chesnut. 14(1). 75(3). K. CT: Information Age. 228–245.. (2012). Teaching and Teacher Education. & Richardson. Wyatt.1037/0022-0663. (2011).1016/j. H.3102/0002831208324517. A.611. & Spero. The structure and measurement of self-efficacy.3. (2006). 7–16. (2009). & Heikkinen. Tschannen-Moran.3200/JEXE.05. L. 110(2).1007/s11618-011-0175-6. W. H. Journal of Educational Research. 11–33. Resilience as a contributor to novice teacher success.cedpsych. Pajares & T.3. (2007). perceived collective teacher efficacy. Acta Scientiarum Naturalium. H. & Kitsantas.. Predicting teacher commitment using principal and teacher efficacy variables: an HLM approach. F. W.. Wheatley. (2007).. Teaching and Teacher Education.. Marsh.74205. (2011). 414–420. learning. 747–766. commitment. F. Contemporary Educational Psychology. Wheatley. (2008). doi:10.. 117–137). doi:10. Teacher self-efficacy and its influence on the achievement of adolescents. P. American Educational Research Journal. M. Journal of Research and Development in Education. M.16 S. Teacher Education Quarterly. & Pajares. Sources of academic and self-regulatory efficacy beliefs of entering middle school students.. L. Wang. & Woolfolk Hoy. P. 14–27..1016/j. & McMaster. 35(4). Positive teacher efficacy as an obstacle to educational reform.1177/ 0013164407308475. 125–141.75. *Watt. 99(3).167-202. & Davis. Tait. Self-efficacy beliefs of adolescents (pp. doi:10. & Kitsantas. Educational and Psychological Measurement. The Elementary School Journal. Universitatis Pekinensis. 611–625. & Pajares. A. (2002). S. (2007). The case for reconceptualizing teacher efficacy research. E.. Zeitschrift für Erziehungswissenschaft. (2005). R. (1996). A. .2006. In F. & Lee. (2008). & Skaalvik. (1999). Journal of Invitational Theory & Practice.. Beginning teachers’ transition from pre-service education to working life. L. & Pajares. doi:10. Self-processes.002. 23(6). (2007).2005.). E. Tynjälä.R. A. Self-efficacy for self-regulated learning. Inviting confidence in school: invitations as a critical source of the academic self-efficacy beliefs of entering middle school students. A. Sources of middle school students’ self-efficacy in mathematics: a qualitative investigation. Teaching and Teacher Education. 100.1080/1743727X. Usher. 303–310. 343–356. 53. 57–75. 37. & Woolfolk Hoy. Journal of Personality and Social Psychology. Teacher efficacy: capturing an elusive construct. Journal of Educational Psychology. K. McInerney (Eds. A guide to constructs of control. doi:10. F. 12. M.. L. Skinner.. Woolfolk Hoy. H. Ware. Self-efficacy: four professional development formats and their relationship to self-efficacy and implementation of a new teaching strategy. (2009). doi:10. 223–247). 166–189.. Teacher self-efficacy: conceptual analysis and relations with teacher burnout and perceived school context. Teaching and Teacher Education. & D. W. Usher.