lied

oh

B-30

stric

E Til

Studies in Educational Evaluation 39 (2013) 169179

stiga

con

ence

s u

od

th a

ren

Contents lists available at SciVerse ScienceDirect

Studies in Educati

jo ur n al ho mep ag e: wwwIntroduction

At present, institutes for teacher education put effort insupporting their student teachers in developing the knowledge,skills and competences required of them. In the development ofthese competences, researchers in educational settings areincreasingly drawing attention to the role student perceptionsand beliefs play in the learning process. In particular self-efcacy,as a key element of social cognitive theory, appears to be asignicant variable in student learning and development (see e.g.,Pajares, 2006; or for a review, see Van Dinther, Dochy, & Segers,2011). Concerning the educational eld, considerable research hasbeen conducted with regard to the relevance of teacher self-efcacy and the development of teacher self-efcacy measures(Tschannen-Moran & Woolfolk Hoy, 2001; Woolfolk Hoy & Davis,2006). However, existing teacher self-efcacy measures are mostlyconcerned with graduated teachers working in the educationaleld, lacking the optimal level of task and context specicitybecause they do not take into account student teacher competencedevelopment and student teacher self-efcacy development.

According to Bandura (1997) and Woolfolk Hoy and Burke-Spero (2005), teacher self-efcacy may be most malleableduring teacher preparation and the rst years of teaching.However, teacher educational institutes pay scarce attention tostudent teacher self-efcacy and research to explore the develop-ment of student teacher self-efcacy is limited.

Taking into account students incipient developmental stage ofteacher competences and teacher self-efcacy, this study intendsto investigate the construct validity and predictive validity of aself-efcacy measure which is developed for predictive anddiagnostic purposes for rst year student teachers in compe-tence-based education.

Teachers sense of efcacy

As a key element of social cognitive theory, self-efcacy appearsto be a signicant variable in diverse domains of humanfunctioning (Pajares, 1996; Schunk, 1995, 2003). Self-efcacyrefers to beliefs in ones capabilities to organize and execute thecourses of action required to produce given attainments (Bandura,1997, p. 3). Within the educational eld, the meaning and measureof teachers sense of efcacy has been the focus of many researchstudies. Teacher self-efcacy is usually dened as the extent towhich the teacher believes he or she has the capacity to affect student

performance (Berman, McLaughlin, Bass, Pauly, & Zellman, 1977, p.137) or as their belief in their ability to have a positive effect onstudent learning (Ashton, 1985, p. 142).

Competence-based education

Evaluation methods

Logistic regression analysis revealed that the measure succeeds in predicting students rst-year

outcomes and delivered evidence for the diagnostic value of the scale.

2013 Elsevier Ltd. All rights reserved.

* Corresponding author at: Department of Pedagogical Studies, Fontys University

of Applied Sciences, P.O. Box 347, 5600 AH Eindhoven, The Netherlands.

Tel.: +31 8778 78810; fax: +31 8778 76588.

E-mail addresses: M.vandinther@fontys.nl (M. van Dinther),

Filip.Dochy@ppw.kuleuven.be (F. Dochy), M.Segers@maastrichtuniversity.nl

(M. Segers), J.Braeken@uvt.nl (J. Braeken).

0191-491X/$ see front matter 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.stueduc.2013.05.001The construct validity and predictive vastudent teachers in competence-based

Mart van Dinther a,*, Filip Dochy a, Mien Segers b, JaDepartment of Educational Sciences, Katholieke Universiteit Leuven, P.O. Box 03772, b School of Business and Economics, Maastricht University, P.O. Box 616, 6200MD MaacDepartment of Methodology and Statistics, Tilburg University, P.O. Box 90153, 5000L

A R T I C L E I N F O

Article history:

Received 12 December 2012

Received in revised form 13 May 2013

Accepted 15 May 2013

Keywords:

Student evaluation

Student teacher self-efcacy measure

A B S T R A C T

This study intends to inve

and diagnostic purposes

delivered converging evid

and the bi-factor model a

loadings of the bifactor m

enter the programme wi

experiences a further diffedity of a self-efcacy measure forucation

an Braeken c

00 Leuven, Belgium

ht, The Netherlands

burg, The Netherlands

te the validity of a self-efcacy measure which is developed for predictive

cerning student teachers in competence-based education. CFA results

for the multidimensionality of the student teacher self-efcacy construct

nderlying structure, reecting a teacher competence framework. Factor

el evidenced the theoretical assumption that incipient student teachers

global undifferentiated sense of teacher self-efcacy, having teaching

tiation takes place to a partly differentiated sense of teacher self-efcacy.

onal Evaluation

.e lsev ier . c om / s t u ed u c

M. van Dinther et al. / Studies in Educational Evaluation 39 (2013) 169179170The notion that teachers beliefs about their capabilities asteachers are of consequence, dates from Rotters social learningtheory (1966). According to this conceptual base, teachers sense ofefcacy was viewed as the extent to which teachers believewhether the reinforcement of their teaching activities lies withintheir own control (internal) or outside their control and within theinuence of the environment (external).

The second conceptual base originated from Banduras work(1977) and identied teacher self-efcacy as a type of self-efcacyamong several other types. The meaning of teacher self-efcacy asa type of self-efcacy regarding student achievement andmotivation has been investigated in several studies (WoolfolkHoy & Davis, 2006). Several researchers found signicant relationsbetween teacher sense of efcacy and student achievement. Wemention some examples. Ashton and Webb (1986) demonstratedthat students generally learn more from teachers with a high senseof efcacy than from teachers with a low sense of efcacy. Otherresearchers showed that students guided by high self-efcaciousteachers achieved higher in subjects such as mathematics (Muijs &Reynolds, 2001; Ross, 1992, 1998) and reading (Ross, 1992, 1998)than did students guided by low self-efcacious teachers. Othersconnected teacher self-efcacy with student motivation (Midgley,Feldlaufer, & Eccles, 1989) and students interest in and attitudetowards school (Woolfolk, Rosoff, & Hoy, 1990). Referencingteacher behaviour, research has pointed out that teachers with ahigh sense of self-efcacy differ from those with low sense of self-efcacy in their teaching behaviour regarding issues such asclassroom management, instruction, teacher feedback. Research-ers as Chacon (2005), Woolfolk and Hoy (1990) and Woolfolk et al.(1990) suggest that teacher efcacy is related to teacher classroommanagement. High efcacy teachers incline to less controlling andmore humanistic behaviour in handling their students. Highefcacious teachers apt to divide the class for small groupinstruction and direct teaching (Gibson & Dembo, 1984; Muijs &Reynolds, 2001), spend more time in interactive instruction(Smylie, 1988), demonstrate higher levels of planning andorganisation (Allinder, 1994), and demonstrate more enthusiasmin their teaching (Allinder, 1994; Guskey, 1984) than do their lowefcacious colleagues. Ashton, Webb, and Doda (1983) foundsignicant relations between teacher self-efcacy and interactionsbetween teacher and students, and student accomplishments.High efcacy teachers focused more on high standards, instruction,student task behaviour and a supportive climate, than do lowefcacy teachers. Gibson and Dembo (1984) and Dembo andGibson (1985) investigated the inuence of teacher efcacy onacademic focus and teacher feedback. Their results revealed thathigh efcacy teachers were more effective in leading students tocorrect responses by means of questioning than were low efcacyteachers. High efcacious teachers are less critical to and spentmore time in working with and monitoring students who exhibitedlearning difculties (Ashton & Webb, 1986; Gibson & Dembo,1984), those teachers perceive all students as teachable (Soodak &Podell, 1993, 1996).

Considering this substantial amount of research ndings,pointing to the central role of teacher self-efcacy plays inteaching competence and teacher effectiveness, it seems relevantfor teacher educational institutes to pay attention to studentsdeveloping self-efcacy within the learning process.

Measuring teachers sense of efcacy

During the last three decades several researchers haveattempted to measure teacher self-efcacy, resulting in short,general measures as well as long, detailed ones. Although the studyof teacher self-efcacy started with RAND researchers notion,dating from Rotters social learning theory; in particular theconceptual base originating from Banduras social cognitive theory(1977, 1997) gave rise to the development of several teacher self-efcacy measures.

According to this Bandura tradition, the Gibson and Dembo(1984) Teacher Efcacy Scale (TES) is the most used instrument.They developed a two-factor instrument, to measure twoconstructs of social cognitive theory, self-efcacy and outcomeexpectancy. One factor, conceptualized as Personal TeachingEfcacy, refers to self-efcacy. The second factor, conceptualizedas General Teaching Efcacy, refers to outcome expectancy, whichis the individuals appraisal of the likely consequences of executedactions. However, continued research on this two-factor instru-ment revealed inconsistencies and factor loadings appeared to benot always consistent across studies (see e.g., Anderson, Greene, &Loewen, 1988; Hoy & Woodfolk, 1993; Soodak & Podell, 1993). Atrst, factor analyses conrmed the two-factor instrument. Lateron, in continued research building on Gibson and Dembos two-factor solution, researchers introduced other factor solutions.Woolfolk and Hoy (1990) maintained Gibson and Dembos GeneralTeaching Efcacy dimension but broke the Personal TeachingEfcacy dimension into two factors, namely teachers sense ofpersonal accountability concerning positive and negative studentlearning outcomes. Soodak and Podell (1996) also argued for athree-factorial solution but proposed an alternative interpretationof the two factors that, according to Woolfolk and Hoy (1990),comprise Personal Teaching Efcacy. Results of their principalcomponents analysis revealed that these two factors were notdifferentiated by positive and negative student learning outcomesbut by Banduras self-efcacy and outcome expectations. Inaddition to this Emmer and Hickman (1991) argued that thePersonal Teaching Efcacy dimension reects two differentefcacy beliefs, teaching and classroom management. Results oftheir principal component analysis conrmed this three-factorsolution. Lin and Gorrell (1998) mentioned a four-factor solutionand labelled the factors as: professional knowledge, effectiveteaching, guiding difcult children and home environment.However, they gave no a priori theoretical arguments that makethis four-factor solution plausible. Brouwers and Tomic (2003)noticed that most researchers who studied the factorial validity ofthe TES only used the statistical technique principal componentsanalysis, which provides no information about the overall t of thefactorial models. They tested different factorial models asproposed by several above-mentioned researchers on theoreticalgrounds. The results of their conrmatory factor analyses deliveredevidence for a four-factor model that signicantly tted the databetter than the other model, although its t did not reach therecommended criterion of adequately tted models. They men-tioned the following reasons why the TES did not demonstrate anadequate factorial model t. Firstly, the item content in bothsubscales reects two different constructs, namely knowing howto teach and being condent about teaching. Secondly, the GeneralTeaching Efcacy subscale reects different reference points, someitems refer to teachers in general and other items refer to theindividual teacher. Deemer and Minke (1999) extensively exam-ined the TES and found that the items of the Personal TeacherEfcacy Scale were valid indicators of teaching efcacy, howeverthey questioned the validity of the General Teaching Efcacy Scale.Removing item wording confounds, they argued for a one-factorsolution, indicating a global Personal Teacher Efcacy dimension.

Considering the above-mentioned teacher self-efcacy mea-surement research, the underlying structure of teacher efcacymeasures resulted in different factor solutions. Some researchersargued for the one-factor solution (Deemer & Minke, 1999). In aone-factor model the covariance among items is explained by onecommon factor (Reise, Morizot, & Hays, 2007). The one-factormodel suggests that in the perception of teachers a global

repr

ror t

M. van Dinther et al. / Studies in Educational Evaluation 39 (2013) 169179 171self-efcacy belief counts and each item is considered to be anindicator of that common factor. According to this model a furtherdifferentiation in more specic self-efcacy aspects would not beworthwhile (denoted by model A in Fig. 1). Other researchersfound evidence for a multi-factor solution (Brouwers & Tomic,2003; Gibson & Dembo, 1984; Tschannen-Moran, Woolfolk Hoy, &Hoy, 1998). In a multi-factor model (denoted as model B in Fig. 1)the covariance among items is explained by several factors, andthese factors are correlated (Reise, Morizot, & Hays, 2007). Thismulti-factor model suggests that in teacher perception there existsa differentiation between several (two or more) teacher self-efcacy aspects, such as instructional self-efcacy and disciplinaryself-efcacy, in which each item is considered to be an indicator ofone specic teacher self-efcacy aspect.

Fig. 1. Path diagrammes of four possible teacher self-efcacy models. Note. Circles arrows represent factor loadings; double-headed arrows represent correlations. ErMore recently, advances in instrumentation make it possible toinvestigate more complex structures such as so-called higher orsecond order factor models (Henson, 2001). In a second-ordermodel (denoted as model C in Fig. 1) items load on rst-orderfactors and rst-order factors load on second-order factors(Rindskopf & Rose, 1988). Due to the persistent measurementproblems, Tschannen-Moran and Woolfolk Hoy (2001) developeda new measure of teacher self-efcacy, the Teachers Sense ofEfcacy Scale (TSES), and labelled three factors: efcacy for studentengagement, efcacy for instructional strategies and efcacy forclassroom management. The TSES goes beyond previous measuresbecause it captures a wider range of teaching tasks. Testing theTSES, Tschannen-Moran and Woolfolk Hoy (2001) conducted asecond-order analysis as the three subscales showed moderatecorrelations. The results demonstrated that the earlier found threefactors collapsed into one strong factor with factor loadingsranging from .74 to .84. According to Tschannen-Moran andWoolfolk Hoy (2001), the appearance of this second-order factorand the moderate correlations between the subscales suggest thatTSES total score as well as the three subscale scores can becalculated. A second-order structure suggests that in teachersperception the three mentioned specic teacher efcacy beliefscontribute to and cluster together in one factor, which refers to amore global self-efcacy belief.

Summarizing the history of teacher self-efcacy measurementresearch, several teacher self-efcacy measures have beendeveloped with mixed psychometric results and different factorsolutions. The discussion centred on two connected issues. Therst issue is related to the theoretical nature of the self-efcacyconstruct (Bandura, 1977, 1997). According to social cognitivetheory (Bandura, 1997) self-efcacy beliefs can vary alongdomain-specic activities and tasks and this implies the challengeof nding the optimal level of specicity for measurement.The second issue refers to the different factor solutions from theprimary instruments aiming to measure teacher self-efcacy. Apossible reason for an inadequate t of these primary instrumentsmay have been due to the employment of a global measure ofteacher self-efcay rather than a context and task specicmeasure. The STES (Tschannen-Moran & Woolfolk Hoy, 2001) ischaracterised by a higher level of task specicity as previous

esent latent factors; squares represent manifest observed variables; single-headed

erms have been omitted for clarity of presentation.measures and demonstrates adequate validity and reliability.However, this measure is concerned with graduate teachersworking in the educational eld. Therefore, the STES is notsuitable for educational purposes because its level of contextspecicity does not take into account student teachers compe-tence development.

According to Bandura (1997) and Woolfolk Hoy and Burke-Spero (2005), teacher self-efcacy may be most malleable duringteacher preparation and the rst years of teaching. However,limited research has explored the development of student teacherself-efcacy and little is known about the way in which incipientstudent teachers self-efcacy developes in relation to experiencedteachers sense of efcacy. According to Eccles, Wigeld, andSchiefele (1998) students school experiences help shape their self-efcacy beliefs and with cognitive development students create amore differentiated view of self-efcacy. With reference to thetarget group of this study, rst-year student teachers in compe-tence-based education, it is plausible that this group enter the rst-year programme with a more global undifferentiated sense ofteacher efcacy. As students have more diverse teaching experi-ences, a differentiation takes place from a broad understanding to amore ne-grained sense of teacher efcacy (Schunk & Meece,2006). In conclusion, there is a need for a teacher self-efcacymeasure that takes into account student teacher competencedevelopment and students incipient developmental stage ofteacher self-efcacy.

Teacher education nowadays: the use of teacher competences

The context for this study is an institute for competence-basedteacher education. Although competence-based approaches with-in teacher education are not new, this approach has emerged sincethe late nineties of the last century, more and more as a leadingparadigm for innovation within higher (teacher) education (Dochy& Nickmans, 2005). A competence can be viewed as an integratedset of related knowledge, skills and attitudes, which enables thestudent to perform professional tasks (in accordance with e.g.Parry, 1996 and Lizzio & Wilson, 2004). Hence, competence-based

in teacher education. This Dutch Association (2009) developed andvalidated a framework for elementary teacher competences inclose collaboration with a large representation of the professionalgroup of teachers in the eld (Dietze, Jansma, & Riezenbosch,2000). For developing this teacher competence framework, fourdifferent roles were distinguished which are characteristic of theteaching profession. These roles are: the interpersonal role, thepedagogical role, the role of expert in subject matter and teachingmethods and the organisational role (Fig. 2, rst column). A teacherperforms these roles within four different situations, alsocharacteristic of the teaching profession. These situations are:

ag

L

pe

M. van Dinther et al. / Studies in Educational Evaluation 39 (2013) 169179172teacher education emphasises the development of competences inrelation to authentic professional situations, instead of merely theacquisition of isolated knowledge, skills and attitudes.

In the late nineties of the last century researchers and teachereducational institutes in several European countries developed, incollaboration with the work eld and other educational institutesin the same occupational domain, teaching competences studentteachers need to acquire for qualication (Struyven & De Meyst,2010). We mention some examples. Tigelaar, Dolmans, Wolfhagen,and Van der Vleuten (2004) developed and validated a frameworkof teaching competences in higher education, containing thedomains: person as teacher, expert on content knowledge,facilitator of learning processes, organizer and scholar/lifelonglearner. Kovacs-Cerovic (2006) mention in their report: skills andknowledge regarding pedagogy and psychology, subject knowl-edge and subject didactics. The Scottisch Ofce (1998) and Zgaga(2006) refer to four areas of competence: subject matter andcontent of teaching; classroom competences; school and theeducation system; and values and attributes related to profession-alism. Fives and Buehl (2008) proposed a framework consisting ofteaching knowledge (e.g. children knowledge and contentknowledge), teaching abilities (e.g. classroom management), skills(e.g. cognitive and communication skills) and qualities (e.g.enthusiasm and dedication). Pantic and Wubbels (2010) investi-gated perceptions of teachers and teacher educators in order toidentify areas of expertise that make up a competent teacher andidentied the four components: values and child rearing;understanding of and contribution to the educational system;subject knowledge, pedagogy and curriculum; self-evaluation andprofessional development.

Next to this, changes in European Union policy together with anincreased interest in teachers and teacher education, resulted inconcensus about the competences teachers need to acquire to meetthe challenges of their role within education nowadays (Fredriks-son, 2003). To support policy makers at a national or regional level,the European Commission set out common European principles forteaching competences and qualications, and recommendationsconcerning the key competences of teachers (European Commis-sion, 2004, 2005).

Dutch schools and teachers use teacher competences, which aredeveloped by the Dutch Association for the professional qualitiesof teachers (2009). Dutch institutes for competence-based teachereducation apply these teacher-derived competences (Storey, 2006)

Contexts

Roles

With students With colle

Interper sonal INT

COPeda gogical PEDSubjec t k now led ge and methodological

SKM

Organisaonal ORG

Fig. 2. Teacher comworking wih students, working with colleagues, working with theschool environment and working with him/herself (Fig. 2, rstrow). A cross-tabulation of these four professional roles andprofessional situations generates a framework for the descriptionof seven teaching competence aspects which are essential for theteaching profession.

The seven competence aspects can be described as follows: ainterpersonally competent teacher (Fig. 2: INT) demonstratesleadership and creates a friendly and cooperative atmosphere,stimulating an open communication and encouraging studentsautonomy. A teacher who is pedagogically competent (Fig. 2: PED)offers students a safe learning environment, within which they canmake choices, and he/she stimulates their social-emotional andmoral development. A teacher who is competent in subjectknowledge and methodology (Fig. 2: SKM), has thoroughknowledge of subject matter and the ability to use teachingmethods effectively. A teacher is organisationally competent(Fig. 2: ORG) when he is able to create a well-organised andtask-oriented learning environment within which students canlearn. A teacher who is competent in collaborating with hiscolleagues (Fig. 2: COL), contributes to the school climate, tocollaboration with colleagues, to the school organisation and to theimprovement of the school. A teacher who is competent incollaborating with the schools environment (Fig. 2: ENV),communicates carefully and responsibly, with students parentsand with colleagues of institutions his school collaborates with. Ateacher who is competent in terms of reection and development(Fig. 2: REF) reects regularly on his/her professional views andcompetence development, keeps his/her professional ability up todate and improves it.

The resulting framework, serving as a teaching standard, closelyresembles the teacher competencies from the above-mentionedinternational studies in the eld of teacher education. Dutchinstitutes for competence-based teacher education apply theelementary aspects of teacher competence (Fig. 2) by deninglevels of prociency for each competence aspect, in terms ofcompetence criteria that a teacherstudent has to achieve givenhis/her specic phase in the study programme. Determiningappropriate competence criteria to assess student competencedevelopment, Dutch teacher institutes use level variables such as:extent of independence, extent of responsibility, extent of task andsituation complexity and extent of transfer (see e.g., Spencer &Spencer, 1993). Considering the purpose of this study, the

ues With the schools environm ent

With him/he rse lf

ENV REF

tence framework.

M. van Dinther et al. / Studies in Educational Evaluation 39 (2013) 169179 173mentioned competence criteria are of great importance for thecontent validity of the student teacher self-efcacy measure.

In addition to the use of competence proles in the curriculumwhich serve as a standard that has to be achieved at the end of theeducational process, competence-based teacher education ischaracterised by the following features: realistic teaching tasksconnected with the vocational practice, the centrality withinteacher education of students competence development, theincreasing responsibility of students for their own learning, theassessments that are aimed at levels of teaching competences, theaddressing of students as starting teachers, the systematicinvolvement of vocational practice, and the functioning of schoolas a learning organisation (Ritzen & Kosters, 2002). This paradigmchange from traditional into competence-based education isconnected with a shift from a testing culture to an assessmentculture (Dochy, Segers, & De Rijdt, 2002) including new modes ofassessment of student learning. These new modes of assessmentstrongly emphasise the integration of assessment and instructionand focus on assessment of the learning process in addition to thatof its products. This new view on assessment is represented by thenotion of assessment as a tool for learning (Black & William, 1998;Gielen, Dochy, & Dierick, 2003) and stresses the diagnostic orformative use of evaluation methods with which studentscompetence development can be monitored and guided. Self-evaluation and self-reection leading to planned competencedevelopment form part of the assessment for learning view(Pollard, Collins, Simco, Swafels, Warin, & Warwick, 2005; Schon,1987). According to social cognitive theory (Bandura, 1986, 1997;Pajares, 2006), self-efcacy plays a predicting role in relation tostudents achievements and student teachers sense of efcacy canbe seen as an indicator for competence development. Incorporat-ing the social-cognitive tradition, with respect to self-efcacy, inup-to-date competence-based teacher education with its emphasison the diagnostic use of evaluation methods with which studentscompetence development can be monitored and guided, there is aneed for a self-efcacy instrument that is suitable for educationalpurposes concerning student teachers in competence-basededucation.

Methods

In this section we rst outline the conceptual framework andreport on the construction and pretesting of the item pool. Afterthat we describe participants and procedures for the validationprocess, where we consider both construct and predictive validity.

Conceptual framework

Self-efcacy is specic in relation to domains, contexts andtasks. As a consequence, self-efcacy measures should be tailoredto the specic domain which is the object of assessment(Zimmerman & Cleary, 2006). The context for this study is aprimary teacher educational programme and the broad constructdomain intended to be measured is self-efcacy with respect toteacher competences. The target population in this study arestudent-teachers in the rst year of the bachelor programme.

Bandura (1997) states that to use the power of the self-efcacyconstruct to explain and predict human functioning, the to-be-assessed beliefs have to match the target of prediction. Hence, self-efcacy items have to be related to the performance outcomes thatare meant to be predicted (Bong, 2006). The targets of predictionare the student teacher results on a competence-based (pass/fail)assessment at the end of the rst year. In particular, the conceptualframework of the Association for the professional qualities ofteachers (2009) introduced earlier is used as a teaching standard.Hence, the focus is on the required competence level for the initialphase in the rst year of the bachelor programme. This includes sixcompetences (see Fig. 2), but excludes the competence forcollaboration with the schools environment, as the latter is notyet relevant for rst-year students. Note that in the Dutch settingthis conceptual framework is in fact legally established in theProfessions in Education Act (Wet BIO). This conceptual compe-tence framework will be exported to a corresponding conceptualteacher-efcacy framework and used as blueprint for the itemconstruction.

Construction of the item pool

A self-efcacy measure should accurately reect the constructinvolved. The construct in question is self-efcacy which refers toperceived capability to perform a task, rather than to psychologicaltraits. Hence, the different competence criteria in the conceptualframework were reformulated into self-efcacy items according tothe standard guidelines of Bandura (2006a, 2006b). For example,one of the criteria for the pedagogical competence is as follows: Idemonstrate my interest in every child, which was reformulatedas How much condence do you have regarding the following: Idemonstrate my interest in every child?. Students were instructedto rate the strength of their condence in executing the requiredactivities by using a 0100 point scale (cannot do at all to highlycertain can do).

In line with the conceptual teacher competence framework, thisprocedure resulted in the construction of 44 items divided acrosssix aspects of teacher self-efcacy. The general instrument designfollows Banduras perspective that multi-faceted measures arefavoured over too general self-efcacy measures, because particu-larized domain-related measures surpass global measures inexplanatory and predictive power.

Pretesting of the initial item pool

We pretested and evaluated the initial item pool in two ways, indepth and in general. For more detailed feedback on item wordingand content, all items were screened on readability, familiarity,and content validity by means of a think-aloud procedure byindividuals of the target population (i.e., teacher students, n = 5)and by content-matter experts (i.e., experienced teachers, n = 3).Furthermore, data on the 44 items were collected from 108 rstyear student teachers, enrolled in a Dutch elementary teachereducation programme. Besides lling in the questionnaire, allparticipants were also asked to put a cross against the items theydid not recognize and/or understand. To make sure that all itemsmeasured at least to some extent the same broad self-efcacyconstruct, using exploratory factor analysis (EFA) individual itemswere screened based upon their communality within a one-factormodel.

As a result of this initial pretesting step, three items werereworded and 13 items with loadings less than the cut-off score of.40 (i.e. barely 16% common variance within one item) wereremoved. Table 1 visualizes the initial item pool and the resultingrened item pool of 31 items. The item pool blue print is structuredaccording to the underlying teacher competence framework.Further validation steps will focus on how well the items conformto this theoretical structure.

Note that the model with one common factor already explained49% of the total variance in the item pool. However, inspection ofthe residual correlations between the items showed 48% non-redundant residuals with absolute values greater than 0.05.According to Preacher and MacCallum (2003) in factor analysis,a factors success is not determined by how much variance itexplains because the model is not intended to explain optimalamounts of variance. A factors success is gauged by how well it

remediation decisions. For that reason we conducted reliabilityanalyses for the teacher-efcacy measure as a whole as well as forthe subscales correspond to each of the six aspects. We managed toacquire at the end of the rst year programme, the results of therst year evaluation for 138 out of 301 students that earlierparticipated in the validation phase. This rst year evaluation is acompetence-based assessment in accordance with the conceptualframework. Students obtained a pass or fail score on each of the sixprimary teacher competences. To evaluate whether the teacherefcacy measure had predictive value towards the end-of-year

M. van Dinther et al. / Studies in Educational Evaluation 39 (2013) 169179174helps the researcher understand the sources of common varianceunderlying the observed data. The results of the EFA demonstratethat the one-factor model is not sufcient given the amount ofresidual correlations and we suspect a more complex underlyingfactor structure (cf. literature review). Due to the small subject toitem ratio (see e.g., Kass & Tinsley, 1979; Comrey & Lee, 1992), nofurther exploration of the factor structure was conducted on thesmall pretest sample to safeguard against possible artefacts inanalyses that attempt to reveal a more detailed factor structurewith more common factors. The information provided by thelimited sample is not sufcient to guarantee unbiased results forthese more complex multi-factor models.

Construct validity

Given the mixed psychometric results and questioned factorsolutions of previous teacher self-efcacy measures and also thefunction of our instrument regarding student teachers incompetence-based education, we wanted to shed some morelight on the possible complex multi-factorial underlying structureof our student teacher efcacy measure. According to Reise,Morizot, and Hays (2007) the so-called bi-factor or general-specic model is a valuable alternative for exploring factorsolution questions relating to broad constructs with heteroge-neous indicators. A bi-factor model (denoted as model D in Fig. 1)consists of one general factor as in model A (see Fig. 1), plus agroup of specic factors as in model B (see Fig. 1). It is one varietyof hierarchy model in which some factors are more general thanothers (Rindskopf & Rose, 1988). The difference between asecond-order model (denoted as model C in Fig. 1) and a bi-factormodel is that the general factor is not a super-ordinate factor, butlies on the same conceptual level as the more specic factors. A bi-factor model makes it possible to measure how much of the itemvariance is due to the general factor and how much to the specicfactors (Chen, West & Sousa, 2006; Reise, Morizot & Hays, 2007). Inthe perception of incipient student teachers the general factor is aglobal not differentiated sense of teacher efcacy, the specicfactors represent already existing more specic teacher self-

Table 1Initial and rened item pool.

Competence aspect Initial

item pool

N items

Rened

item pool

N items

Interpersonal competence 5 3

Pedagogical competence 6 4

Subject knowledge and methodological competence 11 7

Organizational competence 8 7

Competence for collaboration with colleagues 6 4

Competence for reection and development 8 6

Total 44 31efcacy aspects. To further examine construct validity from theperspective of this differentiation hypothesis, we conductedconrmatory factor analysis (CFA). Data were collected from anew and larger sample of 301 rst year student teachers, enrolledin a Dutch primary school teacher education programme. At theend of the rst year programme, but preceding the rst yearassessment, these students were asked by their teachercoach toll in the questionnaire.

Predictive validity

According to Sinharay, Puhan and Haberman (2010), the use ofdiagnostic scores have to meet psychometric quality in terms ofhigh reliability and validity to minimize incorrect instructional andcompetence assessments we conducted logistic regression anal-yses predicting the pass/fail outcome on each of the sixcompetences based upon the students scores on the teacherefcacy measure. For practical reasons for 163 out of the 301students, the results of the rst year evaluation were not available.Scores on teacher efcacy subscales, correlations, and standarddeviations were largely comparable between the two groups. Notethat for all samples considered in this study, approximately 90% ofthe participants were female students and 10% were malestudents.

Results

Construct validity: In search for the underlying structure of thestudent teacher self-efcacy measure

Conrmatory factor analysis was used to t the four discussedteacher self-efcacy models (the one-factor, the multi-factor, thesecond-order factor, and the bi-factor model) (see Fig. 1) to thedata. Factor analysis assumes that a small set of latent unobserv-able variables function as common causes of the manifestobservable items. As such, the items function as indicators ofthese common causes and the latent variables or factors explainthe correlations between the item scores and reect the constructof interest, in this case the teacher self-efcacy construct. Themodels differ in the specication of these common causes and theircorresponding theoretical basis.

To put these four competing factor models to the test, a modelcomparison approach was followed based upon four commonly-accepted goodness-of-t statistics used in structural equationmodelling: the models chi-square x2 as absolute measure of t,the root mean square error of approximation (RMSEA), theBayesian information criterion (BIC) that penalizes model t interms of model complexity, and the comparative t index (CFI) thatoffsets a model to the null model (i.e., no correlation between allitems).

The goodness-of-t indices for all models are shown in Table 2.The multi-factor model, the second-order and the bi-factor modelall show a better t than the one-factor model. This result supportsthe multidimensionality of the teacher self-efcacy construct. Thebest tting model in an absolute sense (i.e., lowest chi-square x2) isthe bi-factor model, which is logical given that it is the mostcomplex model of the four. Yet when balancing model t andcomplexity using the other t statistics, all still converges towardsthe bi-factor model. Therefore, we reject the three alternativemodels and theories, in favour of the bi-factor model.

Table 2Summary of Goodness-of-Fit Indexes for the four teacher self-efcacy models.

Model Parameters CFI RMSEA BIC x2

OM 62 0.784 0.093 1863.548 1512.877

MM 77 0.877 0.072 1470.217 1034.706

M2 68 0.867 0.074 1474.586 1089.979

MB 108 0.918 0.061 1410.712 799.865

Note: OM: one-factor model, MM: multi-factor model, M2: second-order model,

MB: bi-factor model

The bi-factor model

The factor loadings of the bi-factor model can provideadditional insight in the structure of the student teacher self-efcacy scale and can point to potential implications for its use inpractice. In Fig. 3 we provide an overview of the thirty-one itemsand their factor loadings on the general factor and their specicfactor, within the bi-factor model.

Items systematically differed in their loadings on the generalfactor and the specic factor. Three distinct item types can berecognized: items that have a substantive higher loading on thespecic factor than on the general factor (a difference of 0.20 ormore), items that have a substantive higher loading on the generalfactor than on the specic factor (a difference of 0.20 or more), anditems having no substantial difference in loadings on the generaland the specic factor (a difference less than 0.20). (1) Items with asubstantive higher loading on the specic factor are generallyitems concerning concrete behaviour within specic situations, forexample I remain calm in unexpected situations; (2) Items thatload substantively higher on the general factor are itemsreferencing cognitive activities such as I list the characteristicsof childrens social behaviour; (3) Items with no substantiallydifferent loadings on the general and the specic factor, are often acombination of practical and cognitive activities or tasks, forexample Im aware of differences in cultural backgrounds betweenmy colleagues.

The potential implication of this is that a student teacherssense of efcacy is partly differentiated, consisting of a general partand specic parts. The general part refers to a broad commoncognitive belief largely determined by indicators concerningcognitive activities. The specic parts refer to more practicalbeliefs largely determined by indicators referencing performingconcrete behaviour within specic situations. To provide furtherinsight into this potential two-folded interpretation, we deter-mined the variance explained by the bi-factor model and split it upinto variance explained by the general factor and a part that isexplained by the specic factors together. The bi-factor modelexplained 54.2% of the total variance, with the general factorexplaining 22% and the six specic factors together explaining32.2%. These specic factors themselves cannot be split up becausethey are correlated.

Scale and subscale reliability

M. van Dinther et al. / Studies in Educational Evaluation 39 (2013) 169179 175Fig. 3. Results of the conrmatory factor analysis for the bi-factor model. Note.Standardized coefcients are reported. Errors are omitted for reasons of clarity.In order to check if the student teacher self-efcacy scale leadsto consistent measurement results of the construct of interest, weconducted a reliability analysis for the scale as a whole and for eachof its 6 sub-scales (see Table 3). The internal consistency of thescale as a whole is high (Cronbach a = .957). Taking into accountthe small number of items in some sub-scales, all 6 sub-scales alsodemonstrated high internal consistency (i.e., alpha coefcientsranging from .736 for a 3-item sub-scale up to .887 for a 6-itemsub-scale). The resulting student teacher self-efcacy scale can beseen in the appendix.

Predictive validity and diagnostic implications

To examine the predictive validity of the student teacher self-efcacy measure, we looked at how good student teachers efcacyduring the rst year programme can explain outcomes on thecompetence-based evaluation at the end of the year. There are sixbinary (i.e., pass/fail) outcomes, one for each competence aspect,and also the student teacher self-efcacy measure was constructedwith these six aspects in mind. When considering the prediction ofa specic competence, say SKM, we can consider four options onhow to use the information provided by the teacher efcacymeasure.

Baseline. Firstly, we can choose to not use it, and rely only onbaseline information on the average passing rate for the subjectknowledge and methodological (SKM) aspect regardless of thestudent teacher efcacy. The aspect SKM is, with a passingpercentage of only 54% (see baseline column in Tables 4 and 5), infact the competence that is the toughest to obtain, followed by thereection and development (REF) aspect and pedagogical (PED)

Table 3The Cronbach alpha coefcients for the student teacher self-efcacy scale and its 6

subscales.

Subscale Example item n items Cronbach a

How much condence do you

have regarding the following:

31 .957

Interpersonal I stimulate positive behaviour 3 .736

Pedagogical I demonstrate my interest in

every child

4 .784

Subject

knowledge

I use varied learning activities 7 .856

Organisational I remain calm in unexpected

situations

7 .882

Collaboration I am open to advice from

colleagues

4 .804

Reection/

development

I critically reect on my

learning process

6 .887

en

the student teacher self-efcacy measure for their corresponding competence.

Competence Passing %

Y Baseline M SD M M + SDINT 86 76 88 95

PED 70 48 74 90

SKM 54 7 58 96

ORG 86 75 91 97

COL 86 73 91 98

REF 67 39 72 91

M. van Dinther et al. / Studies in Educational Evaluation 39 (2013) 169179176aspect with 67% and 70% (see baseline columns in Tables 4 and 5),whereas the percentage of students that passes the other threecompetence aspects is high (i.e., 86% each).

Specic. Secondly, we can use the one-to-one mapping between

Table 5Logistic regression assessing the predictive validity of the full-scale score of the stud

Student Teacher Efcacy Full scale

M SD b0 (se) b1 (se) OR

STE 79 9 2.41 (.37) .16 (.04)*** 1.17

STE 79 9 1.07 (.22) .13 (.03)*** 1.13

STE 79 9 .18 (.24) .28 (.05)*** 1.32

STE 79 9 2.38 (.36) .14 (.03)*** 1.15

STE 79 9 2.32 (.35) .14 (.03)*** 1.15

STE 79 9 .92 (.22) .13 (.03)*** 1.14

Note: STE = average (INT, PED, SKM, ORG, COL, REF)

Predicted passing % based upon logistic regression:

Pr(Y = pass) = 1/(1 + exp([b0 + b1(STEM)]))with Odds ratio effect size OR = exp(b1).

Table 4Logistic regression assessing the predictive validity of the specic subscale scores of

Student Teacher Efcacy Specic subscale

X M SD b0 (se) b1 (se) OR

INT 79 12 2.03 (.29) .07 (.02)*** 1.07

PED 76 11 1.05 (.21) .10 (.02)*** 1.11

SKM 79 10 .33 (.26) .31 (.05)*** 1.36

ORG 82 10 2.34 (.35) .13 (.03)*** 1.14

COL 82 11 2.34 (.35) .13 (.03)*** 1.14

REF 76 13 .94 (.22) .11 (.02)*** 1.12

Note: Predicted passing % based upon logistic regression:

Pr(Y = pass) = 1/(1 + exp([b0 + b1(X M)]))with Odds ratio effect size OR = exp(b1).competence and student teacher efcacy, and predict the SKMcompetence using information provided by the SKM efcacysubscale. Hence, we tted a logistic regression in which eachcompetence was predicted by the student teacher efcacy score onthe corresponding aspect-specic subscale. This approach stressesmostly the specic element (and partially the general element) ofthe teacher efcacy measure as a predictor. Each specic subscalewas able to signicantly improve upon the baseline prediction ofits corresponding competence (see Table 4). As expected, effectsare always positive, indicating that higher teacher efcacy goesalong with higher chances to pass each of the 6 competences. Effectsizes vary between odds ratio effects of OR = 1.07 for INT toOR = 1.36 for SKM. These results are re-assuring for the predictivevalue of the student teacher efcacy subscales for assessingcompetence development. Even in the presence of near ceiling-effects for 3 out of 6 competences (baseline probabilities of 86%),student teacher efcacy still provides added value in competenceprediction.

Full. Thirdly, we can make use of the broader context and useinformation provided by the full efcacy scale to predict the SKMcompetence. Hence, we tted a logistic regression in which eachcompetence was predicted by the student teacher efcacy score onthe full scale (i.e., comprising all six subscales). This combines boththe general as well the specic element of the teacher efcacymeasure into one predictor score. For each competence, the full-scale score signicantly improved upon the baseline prediction(see Table 5). As expected, effects are always positive, indicatingthat higher teacher efcacy goes along with higher chances to passeach of the six competences. Effect sizes are of similar magnitudeas for the specic subscales and vary between odds ratio effectsof OR = 1.13 for PED to OR = 1.32 for SKM. These results arere-assuring for the predictive value of the student teacher efcacyscale as a whole for assessing competence development.

Aspecic. A last alternative only makes use of non-aspectspecic information and can only borrow information from the

t teacher self-efcacy measure for the six competences.

Competence Passing %

Y Baseline M-SD M M + SD

INT 86 72 92 98

PED 70 47 74 90

SKM 54 8 54 94

ORG 86 74 92 98

COL 86 73 91 97

REF 67 42 71 90general element of the teacher efcacy measure, while mostlyignoring the specic part. Hence, we tted a logistic regression inwhich each competence was predicted by the scores on all teacherefcacy subscales except for the one corresponding to thecompetence.

Table 6 summarizes the model t results of these four logisticregressions by providing posterior model weights based upon thebayesian information criterion (BIC) (Hastie, Tibshirani, & Fried-man, 2009). The higher the weight, the more relative evidencethere is in the data in favour of this model. Notice that modelweights can be interpreted as proportions because they sum up to1 across models for the same competence. Reassuring for thepredictive validity of the teacher self-efcacy measure is thatthe baseline model that ignores its information is ruled outcompletely. Furthermore, the non-aspect-specic model also doesnot receive any support, which implies that general efcacyinformation alone will not provide an accurate competenceprediction. We nd two aspects, INT and PED for which the fullscale context is preferred when making competence predictions,

Table 6Posterior model weights of the four competing logistic regressions predicting each

of the six competences.

INT PED SKM ORG COL REF

1. Baseline .00 .00 .00 .00 .00 .00

2. Specic subscale .00 .17 .98 .36 .81 .91

3. Full scale 1.00 .82 .02 .64 .19 .08

4. Non-aspect-specic subscales .00 .01 .00 .00 .00 .00

Note: Higher weights correspond to more relative evidence in favour of the models

use for prediction.

M. van Dinther et al. / Studies in Educational Evaluation 39 (2013) 169179 177two aspects COL and REF for which specic information issufcient, and mixed support for the SKM aspect. Hence, also inthe predictive validity results, the bi-factor structure of the studentteacher efcacy measures surfaces.

The practical relevance of these results can best be illustrated interms of probabilities (see passing % in Tables 3 and 4). A studentwho gives a score that is equal to the average score on subjectknowledge teacher self-efcacy (SKM = 79) has a 58% chance inobtaining this competence. A student who writes down a score thatis one standard deviation above the average score in this sample(SKM = 88), has a 96% chance and we can almost be certain that s/hepasses this competence. For a student who takes down a score that isone standard deviation below the average score in this sample(SKM = 69), with a 7% chance we can almost be sure that he or shefails this competence. Teacher educators can use the results of thesubscales for diagnostic purposes, although for INT and PED teacherself-efcacy it is advisable to rely on the full-scale information (cf.model weights in Table 6). With regard to students who demonstratesubscale scores that are clearly below the average score, a diagnosticinterview and/or consultation can be useful. Tracing the items thatlower the score can help students competence development byfocusing on the proper learning activities.

Conclusions and discussion

Reviewing teacher self-efcacy measurement research revealedthat during the last three decades several teacher self-efcacymeasures have been developed with mixed psychometric resultsand different factor solutions. It is hard to pinpoint the main reasonfor these mixed psychometric results, but a likely candidate mightbe differences in heterogeneity of the settings and experience ofthe teachers in the different samples. This study focuses on rst-year student teachers in competence-based education and takesinto account students incipient developmental stage of teachercompetences and teacher self-efcacy. The purpose of this studywas to investigate the construct validity and predictive validity of aself-efcacy measure which is developed for predictive anddiagnostic purposes for this target group.

To investigate the construct validity of the self-efcacy measurewe conducted conrmatory factor analysis to provide furtherinsight in the underlying structure of the measure. The four distinctfactor models were compared based upon 4 commonly acceptedgoodness-of-t statistics used in structural equation modeling.These analyses delivered evidence for the multidimensionality ofthe student teacher self-efcacy construct, reecting the underly-ing competence criteria for student teachers. Our results alsorevealed that the three multifactorial models (multi-factor-,second-order and bi-factor model) demonstrate a better t thanone-factor models. These results conrm, with regard to previousteacher self-efcacy measurement research (see e.g. Gibson &Dembo, 1984; Tschannen-Moran & Woolfolk Hoy, 2001), themultidimensionality of the teacher self-efcacy construct ingeneral. Within these results the multi-factor model demonstratesa slightly better t than the second-order model. Although theSTES (Tschannen-Moran & Woolfolk Hoy, 2001) meets the Banduracriterion (1997) of task specicity, our results do not conrmTschannen-Moran and Woolfolk Hoy (2001) ndings pointing atthe second-order model as best tting model.

Furthermore the conrmatory factor analyses delivered con-verging evidence for our differentiation hypothesis regarding thebi-factor model. The bifactormodel explained 54.2% of the totalvariance, within which a general factor explained 22% and sixspecic factors together explained 32.2%, proved to be the besttting model for our purpose. The factor loadings of the bi-factormodel provided additional insight in the structure of the studentteacher self-efcacy scale, a potential implication that a student isincipient developmental stage of teacher self-efcacy is partlydifferentiated, consisting of a general part and specic parts. Thegeneral part refers to a general common cognitive belief largelydetermined by indicators concerning cognitive activities. Thespecic parts refer to specic practical beliefs largely determinedby indicators referencing performing concrete behaviour withinspecic situations. For a further interpretation of these results, it isnecessary to involve the context for this study which is the rstyear of a competence-based teacher education programme.Students who enter this rst year have an early idea of teachingand teaching competences, which tends to be more global orgeneral in nature. This early global concept is based on priorknowledge, teaching experiences drawn from their student roleand in general no or very limited teaching experiences as a teacher.Competence-based teacher education nowadays provides studentteachers with realistic teaching experiences from the rst year ofthe programme. Incipient student teachers, encountering newteaching experiences, interpret these experiences and create a newand better understanding of the teaching practice and requiredteaching competences. In line with Schunk and Meece (2006) whostate that students school experiences help shape their self-efcacy beliefs, we argue that the development of teachercompetences matches the development of rst year studentteachers self-efcacy. This implies, according to the theoreticalassumption of Eccles, Wigeld, and Schiefele (1998), that rst-yearstudent teachers enter the rst-year programme with a moreglobal undifferentiated sense of teacher efcacy. As students havemore teaching experiences a differentiation takes place from abroad understanding to a partly differentiated sense of efcacy,nally leading to a more ne-grained sense of teacher efcacy.

The results of the reliability analysis revealed a high internalconsistency for the scale as a whole and taking into account thesmall number of items in some sub-scales, all six sub-scales alsodemonstrated high internal consistency. Furthermore, the resultsof the logistic regression analyses revealed that the student teacherself-efcacy subscales as well as the student teacher self-efcacyas a whole succeeds in making an accurate prediction of thestudents rst-year outcome on all of the six aspects. These resultsare in line with and conrm other empirical and theoreticalresearch ndings that point at the predicting role of the self-efcacy construct in relation to students achievements, as statedby social cognitive theory (Bandura, 1997). Logistic regressionresults also reveal that general efcacy information alone does notprovide a accurate competence prediction. These results are in linewith social cognitive theory (Bandura, 1977, 1997) as well asearlier teacher self-efcacy measurement research (see e.g.Brouwers & Tomic, 2003), pointing at the inappropriateness ofglobal measures with one-factor solutions.

The summarized construct and predictive validity resultsreveal that we succeeded in developing a student teacher self-efcacy measure (see Appendix) that meets psychometricrequirements in terms of reliability and validity. In concreteterms, the student teacher self-efcacy measure, reects theunderlying competence criteria, and consequently takes intoaccount student teachers stage of competence development.With regard to our literature review, results show that our studentteacher self-efcacy measure outperforms existing teacher self-efcacy measures such as the STES (Tschannen-Moran andWoolfolk Hoy, 2001), because it meets the optimal level of taskas well as context specicity.

The practical result of this study, a method for measuringstudent teachers developing self-efcacy, can be used as amonitoring system for tracking student teachers competencedevelopment during the educational programme in a non-threatening way. The implication of this is that teacher educatorsuse lled-in questionnaires to analyze students answers at three

M. van Dinther et al. / Studies in Educational Evaluation 39 (2013) 169179178levels: the scale level, the subscale level and the item level.Referencing the scale level, students with a low overall sense ofefcacy can be detected in an early stage of the programme. Usingthe subscale level the supervision of students can be targeted at thespecic competence aspects on which they feel less efcacious.According to the item level, inspection of the factor loadings, as aresult of the factor analysis, revealed three distinct groups of items.At the item level, the supervision of students can be targeted at thetype of items on which they feel less efcacious such as concretebehaviour within specic teaching situations or cognitive activitiesas part of the teacher educational programme.

According to Bandura (1997) self-efcacy is most pliable at anearly stage of the learning process, but once self-efcacy beliefshave been solidly set, it would take a certain shock to cause arecalibration. Woolfolk Hoy and Burke-Spero (2005) mentionedthe so-called reality shock as a possible explanation for the declinein sense of efcacy when novice teachers enter the eld and face allthe role demands and the complexity of the teaching task. Ascompetence-based teacher education nowadays addresses stu-dents as starting teachers and provides them with realisticteaching experiences from the beginning, there is the opportunityto prevent a future reality shock. However a teaching experience initself is not automatically a mastery experience, which is in turnthe main source for the establishment of a rm sense of teacherefcacy. This implicates that, to provide incipient student teacherswith mastery teaching experiences, teacher educators have to tunethe authenticity level of the teaching experience, the structure ofthe situation and the supervision of the student teachers to thecomplexity of the teaching task and to the students competencedevelopmental level.

Referencing the risk of overcondence, the measure has to beused with some caution during the rst months of the educationalprogramme. Although modest overcondence is posited to promoteachievement, some student teachers can be overcondent, that is asignicant incongruence between student teacher self-efcacy andsubsequent accomplishments, which can obscure students weak-nesses (Klassen, 2006). However when student teachers enter thevocational practice and gain teaching experiences, their overlyoptimistic self-efcacy beliefs tend to recalibrate.

However, due to the limited response rate, the results of thelogistic regression analyses-concerning predicitive validity anddiagnostic implications must be interpreted with caution. Inorder to conrm our results and to gain more insight into thediagnostic implications of the student teacher efcacy subscales,further research is needed on a larger scale. Next to this furtherinvestigation it is necessary to gain insight in the diagnostic use ofthe student teacher efcacy subscales within the practice ofcompetence-based teacher education. Finally, in this article wefocused on rst-year student teachers concerning students with anincipient developmental stage of teacher efcacy. As a conse-quence, new research is needed to invesigate if and how a furtherdifferentiation of student teacher efcacy takes place during theirfurther competence development.

Appendix A

Student teacher self-efcacy measure

Rate your degree of condence by recording a number from 0 to

100 using the scale given below

0 10 20 30 40 50 60 70 80 90 100

Cannot Moderately Highly certain

Do at all Can do Can do

Condence

How condent are you that you can do the following: (0100)A.1. Interpersonal competence

I see what happens in the classroom.

I stimulate positive behaviour.

I make aspects of group processes explicit in the classroom.

A.2. Pedagogical competence

I demonstrate my interest in every child.

I have knowledge of childrens view of their world.

I act on the basis of the 3 psychological basic needs.

I list the characteristics of childrens social behaviour.

A.3. Subject knowledge and methodological competence

I observe purposefully.

I am acquainted with the learning domains within primary

education.

I have a thorough knowledge of the learning content of my class.

I spot differences of level in my classroom.

I instruct my class clearly.

I use varied learning activities.

I motivate children.

A.4. Organisational competence

I adjust my activities to t in with the group planning.

I monitor time during learning activities.

I oversee the children during learning activities accommodate

the learning environment to learning activities.

I provide learning material on time.

I keep records of students.

I remain calm in unexpected situations.

A.5. Competence for collaboration with colleagues

I keep to my agreements.

I am aware of differences in cultural backgrounds of my colleagues.

I am open to advice from colleagues.

I handle private information about children and colleagues

with care.

A.6. Competence for reection and development

I give my opinion on education.

I ask feedback from others to enable my development.

I critically reect on my learning process.

I use theory to analyze my practical experiences.

I word learning goals.

I consult theory while solving problems.

References

Allinder, R. M. (1994). The relations between efcacy and the instructional practices ofspecial education teachers and consultants. Teacher Education and Special Educa-tion, 17, 8695.

Anderson, R., Greene, M., & Loewen, P. (1988). Relationships among teachers andstudents thinking skills, sense of efcacy, and student achievement. Alberta Journalof Educational Research, 34(2), 148165.

Ashton, P. T. (1985). Motivation and teachers sense of efcacy. In C. Ames & R. Ames(Eds.), Research on motivation in education: Vol. 2. The classroom milieu (pp. 141174). Orlando, FL: Academic Press.

Ashton, P. T., & Webb, R. B. (1986). Making a difference: Teachers sense of efcacy andstudent achievement. New York: Longman.

Ashton, P. T., Webb, R., & Doda, N. (1983). A study of teachers sense of efcacy. Finalreport (Research Report). Washington, DC: National Institute of Education (ERIC,Document Reproduction Service No. ED 231833).

Association for Professional Qualities of Teachers. (2009). Professions in Education ActConsulted online on http://www.lerarenweb.nl.

M. van Dinther et al. / Studies in Educational Evaluation 39 (2013) 169179 179Bandura, A. (1977). Self-efcacy: Toward a unifying theory of behavioral change.Psychological Review, 84, 191215.

Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory.Englewood Cliffs, NJ: Prentice Hall.

Bandura, A. (1997). Self-efcacy: The exercise of control. New York: WH Freeman andCompany.

Bandura, A. (2006a). Adolescent development from an agentic perspective. In F. Pajares& T. Urdan (Eds.), Self-efcacy beliefs of adolescents (pp. 143). Greenwich, CT:Information Age Publishing.

Bandura, A. (2006b). Guide for creating self-efcacy scales. In F. Pajares & T. Urdan(Eds.), Self-efcacy beliefs of adolescents (pp. 307337). Greenwich, CT: InformationAge Publishing.

Berman, P., McLaughlin, M., Bass, G., Pauly, E., & Zellman, G. (1977). Federal programssupporting educational change: Vol VII. Factors affecting implementation and contin-uation (Rep. No. R-1589/7-HEW). Santa Monica, CA: RAND (ERIC Document Repro-duction Service No. 140432).

Black, P., & William, D. (1998). Assessment and classroom learning. Assessment inEducation, 5(1), 774.

Bong, M. (2006). Asking the right question. In F. Pajares & T. Urdan (Eds.), Self-efcacybeliefs of adolescents (pp. 287305). Greenwich, CT: Information Age Publishing.

Brouwers, A., & Tomic, W. (2003). A test of the factorial validity of the teacher efcacyscale. Research in Education, 69, 6779.

Chacon, C. T. (2005). Teachers perceived efcacy among English as a foreign languageteachers in middle schools in Venezuela. Teaching and Teacher Education, 21, 257272.

Chen, F. F., West, S. G., & Sousa, K. H. (2006). A comparison of bifactor and second-ordermodels of quality of life. Multivariate Behavioral Research, 41, 189225.

Comrey, A. L., & Lee, H. B. (1992). A rst course in factor analysis. Hillsdale, NJ: LawrenceErlbaum Associates.

Deemer, S. A., & Minke, K. M. (1999). An investigation of the factor structure of theteacher efcacy scale. Journal of Educational Research, 93, 310.

Dembo, M., & Gibson, S. (1985). Teachers sense of efcacy: An important factor inschool improvement. Elementary School Journal, 86(2), 173184.

Dietze, A., Jansma, F., & Riezebosch, A. (2000). Een kijkkader voor competenties voor detweedegraads lerarenopleidingen [A framework of competencies for secondary gradeteacher education]. Utrecht: Programma Management Educatief Partnerschap.

Dochy, F., & Nickmans, G. (2005). Competentiegericht opleiden en toetsen. Theorie enpraktijk van exibel leren. [Competence-based instruction and assessment. Theoriesand practice of exible learning]. Utrecht: Lemma.

Dochy, F., Segers, M., & De Rijdt, C. (2002). Nieuwe ontwikkeling: De assessmentcul-tuur. [New developments: The assessmentculture]. In F. Dochy, L. Heylen, & H. Vande Mosselaer (Eds.), Assessment in onderwijs. Nieuwe toetsvormen en examinering instudentgericht en competentiegericht onderwijs (pp. 1126). Utrecht: Lemma.

Eccles, J. S., Wigeld, A., & Schiefele, U. (1998). Motivation to succeed. In Eisenberg, N.(Ed.). Handbook of child psychology: Social, emotional, and personality developmentpp. 10171095). (Vol. 3New York: Wiley.

Emmer, E. T., & Hickman, J. (1991). Teacher efcacy in classroom management anddiscipline. Educational and Psychological Measurement, 51, 755765.

European Commission. (2004). Commission staff working paper: Progress towards thecommon objectives in education and training. Indicators and benchmarks. Brussels:European Commission.

European Commission. (2005). Testing Conference on the Common European principlesfor Teacher Competences and Qualications 20th21st June 2005. Brussels: EuropeanCommission.

Fives, H., & Buehl, M. M. (2008). What do teachers believe? Developing a framework forexamining beliefs about teachers knowledge and ability. Contemporary Education-al Psychology, 33(2), 134176.

Fredriksson, U. (2003). Changes of education policies within the European union in thelight of globalisation. European Educational Research Journal, 2(4), 522546.

Gibson, S., & Dembo, M. (1984). Teacher efcacy: A construct validation. Journal ofEducational Psychology, 76(4), 569582.

Gielen, S., Dochy, F., & Dierick, S. (2003). Evaluating the consequential validity of newmodes of assessment: The inuence of assessment on learning, including pre-,post-and true assessment effects. In M. Segers, F. Dochy, & E. Cascallar (Eds.),Optimising new modes of assessment: In search of qualities and standards (pp. 3754).Dordrecht: Kluwer Academic Publishers.

Guskey, T. R. (1984). The inuence of change in instructional effectiveness upon theaffective characteristics of teachers. American Educational Research Journal, 21,245259.

Hastie, T., Tibshirani, R., & Friedman, J. (2009). The elements of statistical learning: Datamining, inference and prediction. New York: Springer.

Henson, R. K. (2001). The effects of participation in teacher research on teacher efcacy.Teaching and Teacher Education, 17, 819836.

Hoy, W. K., & Woodfolk, A. E. (1993). Teachers sense of efcacy and the organiza-tional health of schools. The Elementary School Journal, 93, 356372.

Kass, R. A., & Tinsley, H. E. A. (1979). Factor analysis. Journal of Leisure research, 11,120138.

Klassen, R. M. (2006). Too much condence? The self-efcacy of adolescents withlearning disabilities. In F. Pajares & T. Urdan (Eds.), Self-efcacy beliefs of adolescents(pp. 181200). Greenwich, CT: Information Age Publishing.

Kovacs-Cerovic, T. (2006). National Report Serbia. In P. Zgaga (Ed.), The prospects ofteacher education in South-East Europe (pp. 487526). Ljubljana: University ofLjubljana.

Lin, H., & Gorrell, J. (1998). Pre-service teachers efcacy beliefs in Taiwan. Journal ofResearch and Development in Education, 32, 1725.

Lizzio, A., & Wilson, K. (2004). Action learning in higher education; an investigation of itspotential to develop professional capability. Studies in Higher Education, 29, 469488.Midgley, C., Feldlaufer, H., & Eccles, J. S. (1989). Change in teachers efcacy and studentself and task related beliefs in mathematics during the transition to junior highschool. Journal of Educational Psychology, 81, 247258.

Muijs, D., & Reynolds, D. (2001). Teachers beliefs and behaviours: What really matters.Journal of Classroom Interaction, 37, 315.

Pajares, F. (1996). Self-efcacy beliefs in academic settings. Review of EducationalResearch, 66(4), 543578.

Pajares, F. (2006). Self-efcacy during childhood and adolescence: Implications forteachers and parents. In F. Pajares & T. Urdan (Eds.), Self-efcacy beliefs of adoles-cents (pp. 339367). Greenwich, CT: Information Age Publishing.

Pantic, N., & Wubbels, T. (2010). Teacher competencies as a basis for teacher education Views of Serbian teachers and teacher educators. Teaching and Teacher Education,26, 694703.

Parry, S. R. (1996). The quest for competence. Training Magazine, 8, 4856.Pollard, A., Collins, J., Simco, N., Swafeld, S., Warin, J., & Warwick, P. (2005). Reective

teaching. London: Continuum.Preacher, K. J., & MacCallum, R. C. (2003). Repairing tom swifts electric factor analysis

machine. Understanding Statistics, 2, 1343.Reise, S. P., Morizot, J., & Hays, R. D. (2007). The role of the bifactor model in resolving

dimensionality issues in health outcomes measures. Quality of Life research, 16,1931.

Rindskopf, D., & Rose, T. (1988). Some theory and applications of conrmatory second-order models of quality of life. Multivariate Behavioral Research, 23, 5167.

Ritzen, M., & Kosters, J. (2002). Mogelijke functies van een portfolio binnen eencompetentiegestuurd curriculum [Possible functions of a portfolio within a com-petence-based curriculum]. Tijdschrift Onderzoek van Onderwijs, 31(1), 37.

Ross, J. A. (1992). Teacher efcacy and the effect of coaching on student achievement.Canadian Journal of Education, 17(1), 5165.

Ross, J. A. (1998). The antecedents and consequences of teacher efcacy. In Brophy, J.(Ed.). Advances in research on teaching pp. 4973). (7Greenwich, CT: JAI Press.

Rotter, B. J. (1966). Generalized expectancies for internal versus external controlreinforcement. Psychological Monographs80 (whole no. 609).

Schon, D. (1987). Educating the reective practitioner. San Francisco: Jossey Bass.Schunk, D. H. (1995). Self-efcacy and education and instruction. In J. E. Maddux (Ed.),

Self-efcacy, adaptation and adjustment: Theory, research and application (pp. 281303). New York: Plenum Press.

Schunk, D. H. (2003). Self-efcacy for reading and writing: inuence of modeling, goalsetting and self-evaluation. Reading and Writing Quarterly: Overcoming LearningDifculties, 19(2), 159172.

Schunk, D., & Meece, J. L. (2006). Self-efcacy development in adolescence. In F. Pajares& T. Urdan (Eds.), Self-efcacy beliefs of adolescents (pp. 7196). Greenwich, CT:Information Age Publishing.

Sinharay, S., Puhan, G., & Haberman, J. (2010). Reporting diagnostic scores in educa-tional testing: Temptations, pitfalls and some solutions. Multivariate BehavioralResearch, 45, 553573.

Smylie, M. A. (1988). The enhancement function of staff development: Organizationaland psychological antecedents to individual teacher change. American EducationalResearch Journal, 25, 130.

Soodak, L., & Podell, D. (1993). Teacher efcacy and student problem as factors inspecial education referral. The Journal of Special Education, 27(1), 6681.

Soodak, L., & Podell, D. (1996). Teacher efcacy: Toward the understanding of amultifaceted construct. Teaching and Teacher Education, 12(4), 401411.

Spencer, L., & Spencer, S. (1993). Competence at work: Models for superior performance.New York: Wiley.

Storey, A. (2006). The search for teacher standard: a nationwide experiment in theNetherlands. Journal of Education Policy, 21(2), 215234.

Struyven, K., & De Meyst, M. (2010). Competence-based teacher education: Illusionor reality? An assessment of the implementation status in Flanders fromteachers and students points of view. Teaching and Teacher Education, 26,14951510.

The Scottisch Ofce. (1998). Guidelines for initial teacher education courses in ScotlandRetrieved from http://www.scotland.gov.uk/library/documents-w3/git-00.gtm.

Tschannen-Moran, M., Woolfolk Hoy, A., & Hoy, W. (1998). Teacher efcacy: Itsmeaning and measure. Review of Education Research, 68(2), 202248.

Tschannen-Moran, M., & Woolfolk Hoy, A. (2001). Teacher efcacy: Capturing anelusive construct. Teaching and Teacher education, 17, 783805.

Tigelaar, D., Dolmans, D., Wolfhagen, I., & Van der Vleuten, C. (2004). The developmentand validation of a framework for teaching competencies in higher education.Higher Education, 48, 253268.

Van Dinther, M., Dochy, F., & Segers, M. (2011). Factors affecting students self-efcacyin higher education. Educational Research Review, 6(2), 95108.

Woolfolk, A. E., & Hoy, W. K. (1990). Prospective teachers sense of efcacy and beliefsabout control. Journal of Educational Psychology, 82, 8191.

Woolfolk Hoy, A., & Burke-Spero, R. (2005). Changes in teacher efcacy during the earlyyears of teaching: A comparison of four measures. Teaching and Teacher Education,21, 343356.

Woolfolk Hoy, A., & Davis, H. A. (2006). Teacher self-efcacy and its inuence on theachievement of adolescents. In F. Pajares & T. Urdan (Eds.), Self-efcacy beliefs ofadolescents (pp. 117137). Greenwich, CT: Information Age Publishing.

Woolfolk, A. E., Rosoff, B., & Hoy, W. K. (1990). Teachers sense of efcacy andtheir beliefs about managing students. Teaching and Teacher Education, 6, 137148.

Zgaga, P. (2006). The prospects of teacher education in South-East Europe. Ljubljana:University of Ljubljana.

Zimmerman, B. J., & Cleary, T. J. (2006). Adolescents development of personal agency.In F. Pajares & T. Urdan (Eds.), Self-efcacy beliefs of adolescents (pp. 4569).Greenwich, CT: Information Age Publishing.

The construct validity and predictive validity of a self-efficacy measure for student teachers in competence-based educationIntroductionTeachers sense of efficacyMeasuring teachers sense of efficacyTeacher education nowadays: the use of teacher competences

MethodsConceptual frameworkConstruction of the item poolPretesting of the initial item poolConstruct validityPredictive validity

ResultsThe bi-factor modelScale and subscale reliabilityPredictive validity and diagnostic implications

Conclusions and discussionAppendix AInterpersonal competencePedagogical competenceSubject knowledge and methodological competenceOrganisational competenceCompetence for collaboration with colleaguesCompetence for reflection and development

References