sources of self‐efficacy: four professional development formats and their relationship to...

19
Sources of Self‐Efficacy: Four Professional Development Formats and Their Relationship to Self‐Efficacy and Implementation of a New Teaching Strategy Author(s): By Megan Tschannen‐Moran and Peggy McMaster Source: The Elementary School Journal, Vol. 110, No. 2 (December 2009), pp. 228-245 Published by: The University of Chicago Press Stable URL: http://www.jstor.org/stable/10.1086/605771 . Accessed: 21/05/2014 00:58 Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp . JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected]. . The University of Chicago Press is collaborating with JSTOR to digitize, preserve and extend access to The Elementary School Journal. http://www.jstor.org This content downloaded from 195.78.108.157 on Wed, 21 May 2014 00:58:59 AM All use subject to JSTOR Terms and Conditions

Upload: peggy

Post on 08-Jan-2017

212 views

Category:

Documents


0 download

TRANSCRIPT

Sources of Self‐Efficacy: Four Professional Development Formats and Their Relationship toSelf‐Efficacy and Implementation of a New Teaching StrategyAuthor(s): By Megan Tschannen‐Moran and Peggy McMasterSource: The Elementary School Journal, Vol. 110, No. 2 (December 2009), pp. 228-245Published by: The University of Chicago PressStable URL: http://www.jstor.org/stable/10.1086/605771 .

Accessed: 21/05/2014 00:58

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact [email protected].

.

The University of Chicago Press is collaborating with JSTOR to digitize, preserve and extend access to TheElementary School Journal.

http://www.jstor.org

This content downloaded from 195.78.108.157 on Wed, 21 May 2014 00:58:59 AMAll use subject to JSTOR Terms and Conditions

Sources ofSelf-Efficacy: FourProfessionalDevelopment Formatsand TheirRelationship toSelf-Efficacy andImplementation of aNew TeachingStrategy

Megan Tschannen-MoranCollege of William and Mary

Peggy McMasterWest Virginia ASCD

Abstract

This quasi-experimental study tested the po-tency of different sources of self-efficacy beliefs.Respondents were primary teachers (N � 93) in9 schools who completed surveys of their self-efficacy beliefs and level of implementation of anew teaching strategy for beginning readers be-fore and after participating in 1 of 4 formats ofprofessional development presenting the sameteaching strategy with increasing levels ofefficacy-relevant input. Results indicated thatthe professional development format that sup-ported mastery experiences through follow-upcoaching had the strongest effect on self-efficacybeliefs for reading instruction as well as for im-plementation of the new strategy. A substantialproportion of the teachers who participated informats that included a demonstration with lo-cal students and a planning and practice session,but no follow-up coaching, experienced a de-crease in their self-efficacy for reading instruc-tion.

Self-efficacy, the belief in one’s abilities toaccomplish desired outcomes, powerfullyaffects people’s behavior, motivation, and,ultimately, their success or failure (Ban-dura, 1997). Without self-efficacy, peopledo not expend effort in endeavors becausethey perceive their efforts will be futile.Teacher self-efficacy is a teacher’s per-ceived capability to impart knowledge andto influence student behavior, even that ofunmotivated or challenging students (Gus-key & Passaro, 1994). A growing body ofempirical evidence supports Bandura’s(1977) theory that teachers’ self-efficacy be-liefs are related to the effort they invest inteaching, the goals they set, their persis-tence when things do not go smoothly, andtheir resilience in the face of setbacks.Teachers’ self-efficacy has been linked to

The Elementary School JournalVolume 110, Number 2© 2009 by The University of Chicago. All rights reserved.0013-5984/2009/11002-0006$10.00

This content downloaded from 195.78.108.157 on Wed, 21 May 2014 00:58:59 AMAll use subject to JSTOR Terms and Conditions

their behavior in the classroom and the im-plementation of instructional change (Ash-ton & Webb, 1986; Guskey, 1986; Haney,Wang, Keil, & Zoffel, 2007; McKinney, Sex-ton, & Meyerson, 1999; Timperley & Phil-lips, 2003). Teachers’ self-efficacy has alsobeen related to student outcomes such asstudents’ self-efficacy beliefs and studentengagement, motivation, and achievement(Anderson, Greene, & Loewen, 1988; Midg-ley, Feldlaufer, & Eccles, 1989; Ross, 1992;Shahid & Thompson, 2001). The presentstudy examined the relationship betweenteacher self-efficacy and the format of pro-fessional development in a new skill andthe implementation of that skill in teachingreading. Teacher self-efficacy was mea-sured before and after each of four differentworkshop formats to reveal the extent towhich each format was related to increasedteacher self-efficacy. The study also exam-ined program implementation in relation tothe professional development format andteacher self-efficacy.

The Development of TeacherSelf-EfficacySelf-efficacy beliefs influence thought pat-terns and emotions that enable goal-directed actions in situations where peoplebelieve they can exercise some control. Self-efficacy is a future-oriented belief about thelevel of competence a person expects he orshe will display in a given situation. Giventhe pivotal role of self-efficacy beliefs inunderstanding human behavior, it is im-portant to understand how these beliefs areformed. Bandura (1997) postulated thatteachers make judgments of their self-efficacy based on the verbal encouragementof important others such as colleagues, su-pervisors, and administrators (verbal per-suasion), the success or failure of otherteachers who serve as models (vicariousexperiences), perceptions of past experi-ences of teaching (mastery experiences),and the level of emotional and physiologi-cal arousal experienced as they anticipate

and practice teaching. These beliefs arespecific to particular teaching contexts;therefore, teachers form perceptions abouttheir personal capabilities in light of therequirements of a particular teaching task(Tschannen-Moran, Woolfolk Hoy, & Hoy,1998). As a motivational construct, self-efficacy influences teachers’ effort and per-sistence that then affects performance, inturn becoming a new source of efficacy in-formation. The cyclical nature of behaviorinfluencing self-efficacy, and thus new be-haviors, forms a self-reinforcing cycle ofeither success or failure that tends to be-come quite stable unless a jarring experi-ence provokes a reassessment. Professionaldevelopment formats that emphasize vari-ous sources of teacher self-efficacy form thebasis of this study.

Verbal PersuasionVerbal persuasion involves verbal input

from others, such as colleagues, supervi-sors, and administrators, that serves tostrengthen a person’s belief that he or shepossesses the capability to achieve a de-sired level of performance. Bandura (1997)noted that “it is easier to sustain a sense ofefficacy, especially in times of difficulty, ifsignificant others express faith in one’s ca-pabilities than if they convey doubts” (p.101). Verbal persuasion may be limited inits power to create enduring increases inself-efficacy, but it can bolster self-change ifthe positive appraisal promotes greater ef-fort in the development of skills that sub-sequently lead to a stronger sense of effi-cacy. In schools, teachers often receiveverbal persuasion in the form of profes-sional development workshops that pro-vide knowledge of a new strategy as wellas persuasive claims about its usefulness;unfortunately, this format allows little in-put from teachers (Stein & Wang, 1988).Teachers may also receive verbal persua-sion in the form of specific feedback or en-couragement from a supervisor or col-league designed to convince them that they

SOURCES OF SELF-EFFICACY 229

This content downloaded from 195.78.108.157 on Wed, 21 May 2014 00:58:59 AMAll use subject to JSTOR Terms and Conditions

can successfully implement a new teachingstrategy. Verbal persuasion alone may notbe a powerful source of self-efficacy; how-ever, in partnership with other sources ofefficacy, it may provide teachers the en-couragement necessary to expend effort to-ward realistic goals aimed at strengtheningtheir teaching skills.

Vicarious ExperienceA second source of self-efficacy is that

of observing another person successfullyperform the action that one is contemplat-ing. Because teaching lacks absolute mea-sures of adequacy, teachers must appraisetheir capabilities in relation to the perfor-mance of others (Bandura, 1997). The ob-server has the opportunity to appraise hisor her own capabilities because the modelprovides a standard and this can help theobserver set goals for his or her own teach-ing. The greater the assumed similarity be-tween the observer and the model, themore persuasive will be the belief that onepossesses the capabilities to master compa-rable activities. When an observer watchesa successful teaching exchange, he or sheis more likely to see the teaching task asmanageable. Likewise, when the teachingmodel fails despite strong efforts, the ob-server may judge the teaching task to beout of reach. People actively seek proficientmodels who demonstrate the competenciesto which they aspire. Competent modelstransmit knowledge and teach observerseffective skills and strategies for manag-ing task demands through their behaviorand by revealing their thinking about thetask at hand. Some professional develop-ment models provide vicarious experiencesthrough videos of the skill or strategy inaction. If the vicarious experience is limitedto watching the presenter, it may be onlyminimally effective at increasing teachingskill (Joyce & Showers, 1988). However, aspart of a comprehensive developmental ex-perience, observing a proficient perfor-

mance of the skill to be learned can providevaluable information and insight.

Mastery ExperiencesThe most influential source of efficacy

information is personal mastery experi-ences because they provide the most au-thentic evidence of whether one can masterwhatever it takes to succeed in a particularfield or endeavor (Bandura, 1997). Suc-cesses build a robust belief in one’s efficacy,especially when success is achieved early inlearning with few setbacks. Self-efficacy be-liefs may be diminished when success isachieved through extensive external assis-tance, after considerable effort, or on a taskperceived as easy or unimportant. Failuresthat cannot be attributed to a lack of effortor to external events are likely to have adeleterious effect on self-efficacy beliefs.This has important implications for teacherprofessional development. The actual useof the new knowledge presented in a pro-fessional development workshop has beenshown to contribute to changes in teacherself-efficacy, whereas simple exposure tothe material did not (Ross, 1994). Teacherself-efficacy is a dynamic construct that iscyclical in nature. The proficiency of a per-formance creates a new mastery experiencethat serves as a new source of self-efficacythat either confirms or disrupts existingself-efficacy beliefs. Over time, the processstabilizes and a relatively enduring set ofefficacy beliefs are established that tend tobe resistant to change (Bandura, 1997). Thepreservice period, in particular, tends to bea time of learning to teach that is marked bymajor changes in teacher self-efficacy (Ross,1994; Woolfolk Hoy & Burke-Spero, 2005).

Physiological and Affective StatesWhen judging their own capabilities,

people rely partly on information conveyedby physiological and emotional states (Ban-dura, 1997). A person’s level of arousal,whether perceived positively as anticipa-tion or negatively as anxiety, can influence

230 THE ELEMENTARY SCHOOL JOURNAL

DECEMBER 2009

This content downloaded from 195.78.108.157 on Wed, 21 May 2014 00:58:59 AMAll use subject to JSTOR Terms and Conditions

his or her self-efficacy beliefs. Arousal, suchas elevated heart and respiratory rate, in-creased perspiration, or trembling hands,may have enabling or debilitating effects,depending upon whether the situation isperceived as a challenge or a threat (Gre-goire, 2003). Moderate levels of arousalwhen perceived as a challenge can improveperformance by focusing attention and en-ergy on the task, whereas high levels ofarousal perceived as a threat might inter-fere with making the best use of one’s skillsand capabilities. In the professional devel-opment context, exposure to new knowl-edge and teaching strategies may evokearousal in the form of interest and curios-ity. Initial training experiences may causenervous anticipation for a teacher, espe-cially if the teacher is to be observed andthe performance critiqued. But trying out anew strategy in a supportive workshopsetting where encouragement and assis-tance are available can also help reduce thefear of trying it with a room full of stu-dents. With the ease that comes throughcontinued training and skill development,successfully implemented lessons createfeelings of accomplishment, pride, and ex-hilaration.

Analysis of the Teaching TaskIn assessing their beliefs about their

teaching capability in a particular context,teachers will make two interrelated judg-ments: an assessment of their personalteaching competence and an assessment ofthe assumed requirements of an antici-pated teaching task (Tschannen-Moran etal., 1998). Judgments of personal compe-tence are those a teacher makes about his orher capabilities based on an assessment ofinternal strengths and deficits. The assess-ment of the teaching task may include stu-dent factors such as their current level ofachievement, motivation, or socioeconomicstatus, and contextual factors such as theinstructional resources available, the prob-lems or possibilities posed by the physical

space, and the attitudes of other teachersand administrators, including the level oftrust, collective efficacy, and school climate.Self-efficacy beliefs can vary within teach-ers depending upon the subject area, tracklevel of students, and whether they areteaching outside their field of expertise(Raudenbush, Rowan, & Cheong, 1992;Ross, Cousins, & Gadalla, 1996; Ross, Cous-ins, Gadalla, & Hannay, 1999; Tschannen-Moran & Johnson, 2004). In addition,Tschannen-Moran and Woolfolk Hoy (2007)found that contextual variables such asavailable resources, verbal persuasion inthe form of collegial and community sup-port, and mastery experiences contributedmore to the self-efficacy beliefs of noviceteachers than career teachers who had awealth of mastery experiences on which tobase their self-perceptions.

Teacher Self-Efficacy and theImplementation of New InstructionalStrategiesOne of the most interesting and importantreasons for scholars and school leaders topay attention to teachers’ self-efficacy is therole it plays in teachers’ implementation ofnew teaching strategies presented throughprofessional development. In his model ofteacher change, Guskey (1986) hypothe-sized that the majority of instructional im-provement programs fail because they donot take into account what motivates teach-ers to engage in professional developmentand the process by which change in teach-ers typically takes place. Researchers exam-ining teacher attitudes toward the imple-mentation of new instructional practiceshave frequently found teachers’ self-efficacyto be among the most powerful influenceson receptivity to change (Guskey, 1988;Poole & Okeafor, 1989; Shahid & Thomp-son, 2001; Smylie, 1988). An efficacy-basedchange model of in-service teachers as theywere introduced to whole-language teach-ing strategies was tested by McKinney et al.(1999). As participants moved through the

SOURCES OF SELF-EFFICACY 231

This content downloaded from 195.78.108.157 on Wed, 21 May 2014 00:58:59 AMAll use subject to JSTOR Terms and Conditions

three stages of initiation, implementation,and refinement, a relationship betweentheir self-efficacy and the concerns they ex-pressed was observed. Participants withlower self-efficacy beliefs expressed con-cerns characteristic of those in an earlystage of change, focusing on the impact itwould have on them. Participants withhigher self-efficacy turned their attention tohow the new strategies might affect theirstudents and their school and how theymight work to refine teaching practices andrelationships to better fit within their con-texts—concerns that are typical of laterstages of change. Participants with thehighest self-efficacy tended to view the in-novation as both important and possible.Similarly, Scribner (1999) observed thatteachers’ level of efficacy influenced theirresponse to professional development suchthat high self-efficacy teachers were “op-portunistic in their approach to profes-sional learning and they sought knowledgethrough their involvement in activities thatoften were not overtly professional devel-opment opportunities” (p. 220) while lowself-efficacy teachers were unable or un-willing to engage in the reforms because ofa “perceived disconnection between thepurposes of the efforts and their own needsas professionals” (p. 221). Interventions de-signed specifically to address teachers’ self-efficacy beliefs as they encountered newinstructional processes have been found toraise teachers’ self-efficacy and their imple-mentation of problem-based instructionalpractices in science (Haney et al., 2007). Inaddition, such interventions have beenfound to raise the self-efficacy for class-room management among mathematicsteachers when compared to a control group(Ross & Bruce, 2007).

There seems to be a complex interplaybetween teachers’ self-efficacy beliefs andtheir knowledge and skills to implement anew instructional strategy. In a study of theconditions required to achieve change ininstructional practices in reading, Timper-ley and Phillips (2003) found that change

involved new domain knowledge that re-quired “redefining the reading task andhow to teach it” (p. 627). When teachersimplemented the new methods and wit-nessed unanticipated changes in theirstudents’ achievement, their self-efficacybeliefs were enhanced. Timperley and Phil-lips (2003) proposed that “the change pro-cess is likely to be an iterative rather than asequential one, where changes in beliefs,actions or outcomes are both shaped by,and built on, each other” (p. 630). The find-ings of the complex interplay of newknowledge, changes in students’ achieve-ment, and teachers’ feelings of self-efficacyled Timperley and Phillips to assert thatteacher professional development needs tosimultaneously address teachers’ beliefs aswell as the improvement in their practices.Similarly, when Stein and Wang (1988) setout to identify factors related to teachers’commitment to acquire and consistentlyuse the knowledge and skills necessary forthe successful implementation of a newmodel of reading-skill development, theyfound a “pattern consisting of improve-ment in teachers’ actual expertise in pro-gram implementation, followed by in-creases in their perceptions of self-efficacyfor implementing the program” (p. 181).

The cultivation of self-efficacy beliefsamong teachers has not been found to bea straightforward process of incrementalgains. A number of researchers have foundan “implementation dip” in self-efficacy asteachers began to implement a change ini-tiative (Ross, 1994; Stein & Wang, 1988;Woolfolk Hoy & Burke-Spero, 2005). Theseself-efficacy beliefs tended to rebound forteachers who were able to successfully im-plement the change initiative. There is evi-dence that teachers’ calibration of the levelof their content knowledge is not especiallyaccurate, with teachers tending to overesti-mate their knowledge and skills (Cunning-ham, Perry, Stanovich, & Stanovich, 2004).Thus, their self-efficacy beliefs may bebased on an overestimation of their skills,which may negatively affect self-efficacy

232 THE ELEMENTARY SCHOOL JOURNAL

DECEMBER 2009

This content downloaded from 195.78.108.157 on Wed, 21 May 2014 00:58:59 AMAll use subject to JSTOR Terms and Conditions

beliefs in the face of evidence that theirknowledge and skills were not as strong asthey believed. The detrimental effect onself-efficacy beliefs as teachers attempt topractice a new teaching strategy may beameliorated by additional support duringthe crucial early stages of change. In exam-ining the effect of various components ofprofessional development models, Joyceand Showers (1988) found a jump in effectsizes when practice feedback was added toinformation, theory, and demonstration;there was further increase in effect sizewhen coaching to support implementationwas added to the other components. Re-searchers have suggested that professionaldevelopment programs that aim to supportteachers’ ongoing utilization of new knowl-edge regarding effective practice need todevelop a delivery system characterized bythe provision of continued support andfollow-up after initial training (Guskey,1989; Stein & Wang, 1988).

In a challenge to the view that “more isbetter” when it comes to self-efficacy be-liefs, Wheatley (2002) has proposed thatdoubts about one’s efficacy can sometimesbe beneficial in that uncertainty or doubt iscrucial for the teacher reflection that leadsto new insights. Wheatley challenged Ban-dura’s (1997) claim that it is difficult for aperson to achieve while fighting self-doubt,stating instead that it is difficult for teach-ers to learn and improve without experi-encing self-efficacy doubts. This disequilib-rium and uncertainty may come aboutfrom a challenge to teachers’ beliefs abouttheir existing practices. Wheatley sug-gested that factors such as follow-up coach-ing might moderate the debilitating influ-ence that teacher self-efficacy doubts mayhave on teachers, resulting in improvedpractice. Gregoire (2003) has proposed amodel of teacher conceptual change basedon whether teachers, when presented withan instructional reform initiative, appraiseit as either a threat or a challenge. In thismodel, teachers who believe that they arealready implementing the reform will as-

sess that they are not implicated in thechanges being proposed and will processthe new content superficially. Teacherswho do feel implicated by the reforms pre-sented will experience stress and discom-fort. Those with low self-efficacy are pre-dicted to respond to the reform initiative asa threat, leading to an avoidance intentionand superficial belief change. On the otherhand, teachers with high self-efficacy whoperceive that they have the resources, time,and support necessary to implement theproposed changes would likely interpretthe reform as a challenge and consequentlyengage in more systematic (and thus effort-ful) processing of the information pre-sented. This model explicates the mecha-nisms through which teachers’ self-efficacymediates their response to instructionalchange.

The present study was designed to ex-plore the impact that four professional de-velopment formats with varying levels ofself-efficacy-relevant input would have onteachers’ self-efficacy for teaching. It waspredicted that as the richness of the sourcesof self-efficacy increased, there would be anincrease in teacher self-efficacy and that thelevel of implementation of the new readingstrategy would consequently increase.

MethodThis quasi-experimental study examinedthe role that the format of professional de-velopment played in increasing teacherself-efficacy and implementation of a newinstructional strategy. Four professionaldevelopment formats that emphasized var-ious sources of self-efficacy beliefs postu-lated by Bandura (1997) were utilized. Im-plementation of the newly taught strategyin relation to professional development for-mat and teacher self-efficacy was also ex-amined.

DesignThis study utilized a quasi-experimental

quantitative design. Using cluster sampling,

SOURCES OF SELF-EFFICACY 233

This content downloaded from 195.78.108.157 on Wed, 21 May 2014 00:58:59 AMAll use subject to JSTOR Terms and Conditions

nine schools from five different public schoolsystems in varied settings were placed intoone of four treatment groups. The participat-ing schools were socioeconomically diverse,representing all four quartiles of free/re-duced lunch participation in the state, withone school in the highest quartile, three in thesecond and third quartiles, and two in thelowest quartile as identified by free/reducedlunch participation. There was also diversityof context, with four rural schools, four sub-urban schools, and one urban school. Fourschools received the professional develop-ment in the spring, and five the following fall.The four schools that were available for treat-ment and data collection in the spring wererandomly assigned to the four treatmentgroups, and the five that were available in thefall were likewise randomly assigned to eachof the four groups with the exception thattwo schools were selected for TreatmentGroup 1 to balance small numbers in thatgroup from the spring. A cluster samplingdesign was selected because the professionaldevelopment treatment was offered at theschool level. To have teachers who receiveddifferent treatment formats in the sameschools and on the same grade-level teamswould have introduced a potential cross-contamination of treatment effects.

The Tucker Signing Strategies for Read-ing was selected as the teaching strategy tobe taught because the strategy is simple inscope and short-term in implementation(Tucker, 2001). The strategies address aspecific area of beginning reading—that ofmatching letters to sounds and decodingwords using hand signals to provide a con-ceptual bridge between the written symboland the sound associated with that letter.This method has demonstrated compellingresults with struggling readers (Cole &Majd, 2003; McMaster, 2003). All treatmentdelivery was done by the same trainer whowas trained by the author of the Tuckerstrategies.

In an attempt to differentiate betweenthe various sources of self-efficacy, thisstudy separated the training into four dif-

ferent formats, taking an additive approachto including new or deeper sources ofefficacy-relevant input. All participatingteachers were administered surveys assess-ing their self-efficacy beliefs as well as prioruse of the target instructional strategy atthe beginning of the initial workshop.

Treatment 1 (information). All schoolsreceived a 3-hour workshop entitled theTucker Signing Strategies for Readingstructured in alignment with the Tuckertraining manual. The standard one-timeworkshop format was used as the baseline,with verbal persuasion as the identifiedsource of self-efficacy beliefs. Treatment 1schools received the workshop alone. Us-ing a lecture format, each of the 44 handgestures was presented and demonstratedas participants followed in their own man-uals. Any questions were answered andverbal descriptions were given of the pre-senter’s previous use of the strategies. Par-ticipants were invited to use the gestures todecode words as outlined in the manual.

Treatment 2 (information � modeling).Treatment 2 added vicarious experience, asecond source of self-efficacy, as the pre-senter modeled the use of the Tucker handcues with local students selected by theteachers in the participating school, typi-cally second or third graders who were notyet reading. In this treatment, approxi-mately 20 minutes of the 3-hour workshopwere devoted to a demonstration in whichthe presenter taught these struggling read-ers to use the hand cues, and participantswatched as these students successfully de-coded new words. It was noted that teach-ers acknowledged the success of the handgestures with students for whom othermethods had not been successful. Many ex-pressed their intention to use the strategiesin their own classrooms as a result of seeingthe demonstration.

Treatment 3 (information � model-ing � practice). Treatment 3 added a pro-tected mastery experience with the in-clusion of a one-and-a-half-hour practicesession. During the practice session, partic-

234 THE ELEMENTARY SCHOOL JOURNAL

DECEMBER 2009

This content downloaded from 195.78.108.157 on Wed, 21 May 2014 00:58:59 AMAll use subject to JSTOR Terms and Conditions

ipants worked in groups to make decisionsregarding how they would use the strate-gies, plan appropriate lessons for their stu-dents, and practice implementation of thestrategies (i.e., lessons that they could usewith their students the following day). To-tal time for this treatment was 4 hours and30 minutes.

Treatment 4 (information � model-ing � practice � coaching). Treatment 4added a stronger mastery experience withthe inclusion of follow-up coaching in theuse of the new skill. Coaching took place inthe weeks following the workshops and in-cluded three components: (1) A 30-minutesmall-group review of hand gestures, (2) a15-minute one-on-one coaching session indialogue with the presenter, and (3) a 30-minute coaching session with the presenterin the teacher’s classroom. Not only didthis coaching provide a mastery experienceas teachers were supported in their imple-mentation of a new strategy, it also pro-vided a more individualized and specificexperience of verbal persuasion. For thoseteachers who requested it, the coachingmay have also included a second vicariousexperience within their own classroom set-ting as they observed the presenter use theTucker Signing Strategies with their stu-dents. Total time for this treatment was 5hours and 45 minutes.

Physiological arousal as a source of self-efficacy was not directly tested in the fourprofessional development formats used inthis study, although teachers were nodoubt experiencing some emotional re-sponse to what they were learning.Whether these feelings were beneficial ordetrimental to their self-efficacy beliefs wasnot explored in this study.

ParticipantsThe participants were 93 primary (K–2)

and resource teachers from nine schoolswithin five different school districts. Throughcluster sampling, schools were assigned toworkshop treatment groups through strati-

fied random selection. In each of the par-ticipating schools, all kindergarten throughsecond-grade teachers participated in thestudy. The unit of analysis was the individ-ual teacher. The 93 teachers included in thestudy completed both the initial and finalsurveys.

InstrumentationThree measures were used in this study.

These included the Teachers’ Sense of Effi-cacy Scale (Tschannen-Moran & WoolfolkHoy, 2001), an adaptation of the TeacherSense of Efficacy for Literacy Instruction(Tschannen-Moran & Johnson, 2004), and ameasure of implementation of the Tuckermethod. These are described below.

Teacher self-efficacy. Teachers’ self-efficacy beliefs were measured using the12-item Teacher Sense of Efficacy Scale([TSES]; Tschannen-Moran & Woolfolk Hoy,2001). Instructions directed the respon-dents to “Please respond to each of thequestions by considering the combinationof your current ability, resources, and op-portunity to do each of the following inyour present position.” Teachers rated theitems on a 9-point scale with anchors at 1(none at all), 3 (very little), 5 (some degree),7 (quite a bit), and 9 (a great deal). The scaleincluded three four-item subscales: Efficacyfor Instructional Strategies, Efficacy forClassroom Management, and Efficacy forStudent Engagement. The following are ex-amples of items within each subscale:

Efficacy for Instructional Strategies:

• How well can you implement alterna-tive strategies in your classroom?

Efficacy for Classroom Management:

• How much can you do to calm a stu-dent who is disruptive or noisy?

Efficacy for Student Engagement:

• How much can you do to get studentsto believe they can do well in school-work?

SOURCES OF SELF-EFFICACY 235

This content downloaded from 195.78.108.157 on Wed, 21 May 2014 00:58:59 AMAll use subject to JSTOR Terms and Conditions

The alpha coefficient of internal consistencyof the TSES for the current study was .90 forboth administrations of the instrument.

Teacher self-efficacy for reading in-struction. Seven items adapted from theTschannen-Moran and Johnson (2004) mea-sure of teachers’ sense of efficacy for liter-acy instruction were included to determineteachers’ self-efficacy for the teaching ofreading (TSERI). The instructions and re-sponse anchors were the same as those forthe TSES. Examples of the reading itemsfollow:

• To what extent can you help your stu-dents monitor their own use of read-ing strategies?

• To what extent can you teach thesound/letter relationship to your stu-dents?

• How much can you do to meet theneeds of struggling readers?

The obtained alpha coefficients for theTSERI were .91 in the first administrationand .88 in the second.

Implementation. Implementation of thetarget teaching strategy was assessed attwo stages. A one-item preassessment wasdesigned to determine the level of imple-mentation of the Tucker Signing Strategiesfor Reading on the survey given prior totraining. This item was used to identifyparticipants who were already using theTucker strategy. Since the intent was tostudy teacher participation in training witha new strategy, teachers who respondedwith a 7 or higher were removed from thesample. The posttraining implementationmeasure included five additional items thatwere used to determine the relationship be-tween teacher self-efficacy and implemen-tation of the strategy presented in the train-ing. Examples of the implementation itemsfollow:

• To what extent do you use the TuckerReading Strategies? (Used on bothpre- and postsurveys)

• To what extent do you use the Tuckerhand cues to help your students figure

out unknown words when they arereading?

• To what extent do you use the Tuckerhand cues to meet the needs of strug-gling readers?

The alpha coefficient for the six-item imple-mentation scale was .99, indicating thatwhen a teacher indicated a particular levelof implementation, she tended to indicatethe same level across the various aspects ofreading instruction tapped in the six items.

Data CollectionSchool districts were selected for partic-

ipation on the basis of their willingness toinclude all primary teachers in the schoolbuilding rather than offering the workshopto volunteers after hours in a central loca-tion. This allowed for a more realistic pro-fessional development experience. Oncepermission was granted to conduct work-shops and collect data, principals were con-tacted to set dates for the workshops. At thebeginning of each workshop, the researcherexplained the purpose of the study, assuredconfidentiality, and asked teachers to com-plete the survey honestly. Follow-up sur-veys were administered to all participantsapproximately 1 month after the work-shops. Surveys were coded numerically sothat initial responses could be matchedwith final responses by individual.

Data AnalysisDescriptive statistics were computed

for each treatment group and for thefull sample for pre- and posttests of TSESand TSERI as well as implementation.Repeated-measures ANOVAs were con-ducted to compare the four treatment for-mats across time. A follow-up analysiscompared changes in teacher self-efficacyin general and teacher self-efficacy forreading instruction that were greater thana standard deviation. Finally, an ANOVAwas used to explore the effect of treat-ment format on implementation of theTucker Signing Strategies for Reading.

236 THE ELEMENTARY SCHOOL JOURNAL

DECEMBER 2009

This content downloaded from 195.78.108.157 on Wed, 21 May 2014 00:58:59 AMAll use subject to JSTOR Terms and Conditions

LimitationsA limitation of this study design was

that as new sources of self-efficacy wereadded in the third and fourth treatmentformats, there was also an increase intime, presenting an untested confoundingvariable. The amount of contact time withthe trainer ranged from 3 hours for Treat-ments 1 and 2 to 53⁄4 hours for those inTreatment 4. In addition, all data wereself-reported. This meant that the level ofimplementation was not corroboratedwith classroom observations to confirmwhether or not teachers were actually im-plementing the new strategies at the levelthey reported, except for those teachers inTreatment 4 during which the trainermade classroom visits and discussed withteachers the implementation challengesthey were facing. In addition, teachers’self-reports of self-efficacy are indicatorsof self-beliefs related to motivation andare not necessarily well calibrated to theiractual effectiveness in classroom instruc-tion. Finally, the survey responses mayhave suffered from response bias, al-though the generally moderate correla-tions between scales and across timewould suggest this was not a significantproblem.

ResultsThe findings of this study demonstratedsome surprising relationships betweentreatment formats that included varioussources of self-efficacy and changes in self-efficacy. The expectation that increasing

self-efficacy-relevant input in additive fash-ion would result in steady increases in self-efficacy was not confirmed. The results alsoraised questions regarding how treatmentconditions and self-efficacy beliefs were re-lated to implementation of a new teachingstrategy.

Descriptive StatisticsDescriptive statistics for each of the four

treatment groups for the initial and finalteacher self-efficacy scores, as well as theimplementation scores, are displayed in Ta-ble 1. Responses were on a 1–9 scale, rang-ing from “none at all” to “a great deal.” Themean scores for initial teacher self-efficacy(TSES) and initial teacher self-efficacy forreading instruction (TSERI) were quite sim-ilar, with TSES scores ranging from 6.91 to7.19, and TSERI scores ranging from 6.96 to7.20. These ranges are similar to thosefound by Tschannen-Moran and WoolfolkHoy (2001). The posttest for final teacherself-efficacy ranged from 7.24 to 7.69 andfinal teacher self-efficacy for reading in-struction ranged from 7.11 to 7.99. Allshowed sufficient variability, with stan-dard deviations ranging from .65 to 1.17.Mean implementation scores ranged from2.71 to 6.78 indicating substantial variancein implementation across treatment groupsand standard deviations ranging from 1.67to 2.43 within groups.

Changes in Self-Efficacy BeliefsA steady increase in self-efficacy was

expected from Treatment 1 through Treat-

TABLE 1. Mean Self-Efficacy and Implementation Scores

Treatment 1 Treatment 2 Treatment 3 Treatment 4 Overall(N � 28) (N � 21) (N � 20) (N � 24) (N � 93)

M SD M SD M SD M SD M SD

Initial TSES 7.19 (1.00) 6.91 (.68) 7.00 (.98) 6.98 (.86) 7.03 (.89)Final TSES 7.52 (.84) 7.24 (.65) 7.27 (1.04) 7.69 (.78) 7.45 (.84)Initial TSERI 7.17 (1.07) 7.20 (1.03) 7.08 (1.20) 6.96 (.99) 7.10 (1.06)Final TSERI 7.78 (.77) 7.22 (.90) 7.11 (1.17) 7.99 (.81) 7.56 (.96)Implementation 3.45 (1.82) 2.71 (1.99) 3.73 (2.43) 6.78 (1.67) 4.20 (2.49)

SOURCES OF SELF-EFFICACY 237

This content downloaded from 195.78.108.157 on Wed, 21 May 2014 00:58:59 AMAll use subject to JSTOR Terms and Conditions

ment 4 because of the additive approachof including additional sources of self-efficacy-relevant input for each subsequenttreatment. A repeated-measures ANOVAdid reveal significant increases in teachers’self-efficacy (TSES) across time [F(3, 89) �33.42, p � .01], however, in this case thetreatment effect was not significant. Thismeant that all teachers tended to expressincreased general self-efficacy at the time ofthe posttest (approximately 1 month later)regardless of which treatment group theywere in (see Table 2).

For teachers’ self-efficacy for reading in-struction (TSERI), there was no main effectfor treatment, but there was a significanteffect for time, again demonstrating anoverall increase in self-efficacy for readinginstruction at the time of the posttest [F(3,89) � 19.69, p � .01]. In this case, there wasalso a significant interaction effect betweentime and treatment [F(3, 89) � 6.49, p � .01](see Table 3). An examination of these datarevealed an unexpected pattern (see Fig. 1).While the TSERI scores were homogeneousto begin with, two groups—Treatment 1

and Treatment 4—showed significant gainsin TSERI, while the two middle treatmentgroups—Treatment 2 and Treatment 3—re-mained unchanged.

Further exploration led to the discoverythat these measures of central tendencywere masking a drop in self-efficacy amonga substantial number of teachers. To ex-plore this dynamic, we recoded the posttestscores as falling within a standard devia-tion of the initial scores, those with losses ofmore than a standard deviation, and thosewith a gain of greater than a standard de-viation (based on the SD of initial self-efficacy scores). Only 1% of the generalTSES scores dropped more than a standarddeviation, and while 24% saw increases ofmore than a standard deviation, the chi-square was not significant. For TSERI, how-ever, 14% and 20% of the teachers in Treat-ment 2 and Treatment 3, respectively, sawdeclines in their self-efficacy of greater thana standard deviation. On the other extreme,19% and 10%, respectively, of teachers inthese same two treatment groups saw in-creases in their TSERI of more than a stan-

TABLE 2. Repeated-Measures ANOVA for Teacher Self-Efficacy

Type IIISum of Squares df Mean Square F Sig.

Between-subjects effects:Treatment 2.782 3 .927 .730 .537Error 113.024 89 1.270

Within-subjects effects:Time 7.732 1 7.732 33.420 .000Time � treatment 1.438 3 .479 2.072 .110Error (time) 20.592 89 .231

TABLE 3. Repeated-Measures ANOVA for Teacher Self-Efficacy for Reading Instruction

Type IIISum of Squares df Mean Square F Sig.

Between-subjects effects:Treatment 4.971 3 1.657 1.067 .367Error 138.195 89 1.553

Within-subjects effects:Time 8.235 1 8.235 19.687 .000Time � treatment 8.144 3 2.715 6.490 .001Error (time) 37.228 89 .418

238 THE ELEMENTARY SCHOOL JOURNAL

DECEMBER 2009

This content downloaded from 195.78.108.157 on Wed, 21 May 2014 00:58:59 AMAll use subject to JSTOR Terms and Conditions

dard deviation. Furthermore, no teachers inTreatments 1 and 4 had TSERI declines ofmore than a standard deviation, while 25%of Treatment 1 teachers and 38% of Treat-ment 4 teachers had increases of over a fullstandard deviation (see Table 4).

Implementation of Reading Strategies,Teacher Self-Efficacy, and FormatThe final research question explored the

relationship of self-efficacy and treatmentformat in relation to the implementationof the teaching strategy introduced. AnANOVA was conducted to explore the dif-ferences in implementation of the Tucker

FIG. 1.—Changes in teacher self-efficacy for read-ing instruction.

TABLE 4. Changes in Self-Efficacy Larger Than 1 Standard Deviation

Changes in Self-EfficacyaLess Than 1 SDBelow the Mean

Within 1 SDof the Mean

Greater Than 1 SDAbove the Mean Total

Treatment 1:Count 0 23 5 28% within treatment .0 82.1 17.9 100.0

Treatment 2:Count 0 17 4 21% within treatment .0 81.0 19.0 100.0

Treatment 3:Count 1 16 3 20% within treatment 5.0 80.0 15.0 100.0

Treatment 4:Count 0 14 10 24% within treatment .0 58.3 41.7 100.0

TotalCount 1 70 22 93% within treatment 1.1 75.3 23.7 100.0

Changes in Self-Efficacyfor Reading Instructionb

Less Than 1 SDBelow the Mean

Within 1 SDof the Mean

Greater Than 1 SDAbove the Mean Total

Treatment 1:Count 0 21 7 28% within treatment .0 75.0 25.0 100.0

Treatment 2:Count 3 14 4 21% within treatment 14.3 66.7 19.0 100.0

Treatment 3:Count 4 14 2 20% within treatment 20.0 70.0 10.0 100.0

Treatment 4:Count 0 15 9 24% within treatment .0 62.5 37.5 100.0

Total:Count 7 64 22 93% within treatment 7.5 68.8 23.7 100.0

a�2(6, 93) � 9.40, ns.b�2(6, 93) � 13.37, p � .05.

SOURCES OF SELF-EFFICACY 239

This content downloaded from 195.78.108.157 on Wed, 21 May 2014 00:58:59 AMAll use subject to JSTOR Terms and Conditions

Signing Strategies for Reading based ontreatment format. The results revealed thatwhen controlling for initial self-efficacy,40% of the variance in implementationcould be explained by treatment format[F(3, 89) � 19.57, p � .01]. A Tukey post-hoctest identified Treatment 4 (information,demo, practice, and coaching) as the onlytraining format that varied significantly fromeach of the other three groups in terms ofimplementation of the instructional strategyintroduced. Follow-up coaching was the ele-ment that distinguished Treatment 4 fromthe other training models.

DiscussionThis study explored the relationship be-tween professional development formatsand changes in teacher self-efficacy beliefsand teachers’ self-reports of implementa-tion of a new teaching strategy. Four pro-fessional development formats for the pre-sentation of the Tucker Signing Strategiesfor Reading were designed to include, inadditive fashion, three of the four sourcesof self-efficacy identified by Bandura (1997).The first three formats included verbal per-suasion in the form of information aboutthe teaching strategy, vicarious experiencein the form of a demonstration of the strat-egy with local students, and a protectedmastery experience in the form of a plan-ning and practice session with colleagues.These all were related to modest gains inteacher self-efficacy, and the first was re-lated to gains in teacher self-efficacy forreading instruction. They were not, how-ever, related to increases in implementationof the target teaching strategy.

In keeping with Bandura’s (1997) asser-tions, the most powerful professional de-velopment format included an authenticmastery experience embedded in the teach-er’s regular teaching context. This masteryexperience included the use of the strategyin the teacher’s own classroom with thesupport of a coach providing specific ver-bal persuasion and feedback (and poten-

tially a second vicarious experience withthe teacher’s own students). Through theaddition of follow-up coaching sessions,teachers increased their self-efficacy forreading instruction and also their imple-mentation of the Tucker Signing Strategiesfor Reading.

As in previous studies of teacher self-efficacy using the TSES, the mean scoreswere above the midpoint on the responsescale, with means around 7 on a 9-pointscale (Tschannen-Moran & Woolfolk Hoy,2001), but with standard deviations be-tween .6 and 1.1 there should not be a con-cern of ceiling effects. These strong meansreveal that career teachers tend to have afairly strong sense of efficacy for teaching.Closer inspection of the results revealedthat few practicing teachers selected re-sponse options 1–3 (“none at all” to “verylittle”) concerning their capability as teach-ers. Studies that segregate novice teachershave found their means to be significantlylower than those of career teachers (al-though still above the midpoint), suggest-ing that teachers with low self-efficacy tendto either increase their self-efficacy or exitthe field within the first few years of teach-ing (Tschannen-Moran & Woolfolk Hoy,2007).

For the primary-grade teachers in thisstudy, general sense of efficacy for teachingand self-efficacy for reading instructionwere highly correlated, sharing 55% of thevariance in these constructs. This sharedvariance was not surprising because pri-mary teachers who consider themselvesskillful at teaching reading are also likely toconsider themselves capable teachers ingeneral because the teaching of reading isso central to the task of teaching in theprimary grades. Although self-efficacy is acontext- and task-specific construct (Ban-dura, 1997) that has been shown to varywithin teachers across subject and grade-level classes at the secondary level (Rau-denbush et al., 1992; Ross et al., 1996, 1999),for these primary teachers, the group ofstudents they were referencing in their as-

240 THE ELEMENTARY SCHOOL JOURNAL

DECEMBER 2009

This content downloaded from 195.78.108.157 on Wed, 21 May 2014 00:58:59 AMAll use subject to JSTOR Terms and Conditions

sessments of general self-efficacy and self-efficacy for reading instruction were thesame. Nonetheless, there was sufficient dis-crepancy to justify analyzing general self-efficacy for teaching and self-efficacy forreading instruction as separate constructs.

The Tucker Signing Strategies for Read-ing workshop was targeted specifically tothe skill of teaching beginning reading andhad a significant effect on teacher self-efficacy for reading instruction in the firstand fourth treatment formats. The meanof general teachers’ self-efficacy also in-creased during the course of the study,however, even though the instructionalstrategy taught only targeted reading in-struction for beginning readers. This find-ing bears further exploration in future stud-ies because it suggests that providingteachers with instructional strategies andsupport in one content area may transfer toa more general sense of self-efficacy.

With exposure to a new teaching strat-egy, a substantial proportion of teachers intwo of the four treatment groups saw sub-stantial declines in self-efficacy for readinginstruction. This drop indicates that the de-velopment of self-efficacy is more complexthan expected. It seems that the awarenessof a new instructional strategy that isshown to have an impact on strugglingreaders caused some teachers to reassesstheir definition of good teaching and toprovoke a recalibration of their own self-efficacy beliefs against this new standard.Without coaching to assist teachers in theimplementation of the new skill, a signifi-cant proportion of teachers were left feelingmore inadequate than they had before. InTreatment 4, the group that received coach-ing, virtually all of the teachers experiencedincreases in both general teacher self-efficacy and teacher self-efficacy for read-ing instruction, and nearly 4 in 10 experi-enced gains of more than a standarddeviation in both types of self-efficacy.

One might argue that the findings ofthis study have more to do with the contacttime involved in the various treatments

rather than the proposed sources of self-efficacy. The pattern of gains and dips inself-efficacy, however, offers evidence thatcontact time was not the only salient vari-able, although it may have played a role.The contact times for Treatment 1 andTreatment 2 were the same, and yet therewas a .61 increase in self-efficacy for read-ing instruction in Treatment 1 and almostno mean change in Treatment 2, in partbecause of declines in self-efficacy for read-ing instruction in Treatment 2. Treatment 3extended the workshop from 3 hours to 41⁄2hours, and yet again the mean change inself-efficacy for reading instruction wasnegligible because over half of teachers ev-idenced declines in their self-efficacy forteaching reading, with 20% evidencinglosses of more than a standard deviation.Treatment 4 extended the total contact timeto 53⁄4 hours, and over 90% of the teachersin this treatment experienced increases inboth general self-efficacy and self-efficacyfor reading instruction.

Because self-efficacy has been associ-ated with willingness to try new strategiesin previous research, the relationship be-tween initial self-efficacy and implementa-tion of the new strategy was tested. Thelack of a relationship between initial self-efficacy and implementation raises a chal-lenge to studies that have found that thelevel of implementation of a new methoddepends on the self-efficacy of teachers(Poole & Okeafor, 1989; Scribner, 1999). Inthe present study, variance in the level ofimplementation of the new instructionalstrategy was related to the addition of thefollow-up coaching components of train-ing, not to initial self-efficacy. There is,however, some alignment with earlier re-search in the finding of a relationship be-tween teachers’ final self-efficacy for read-ing instruction and implementation.

ImplicationsThe findings from this study suggest thatthe process of influencing the self-efficacy

SOURCES OF SELF-EFFICACY 241

This content downloaded from 195.78.108.157 on Wed, 21 May 2014 00:58:59 AMAll use subject to JSTOR Terms and Conditions

beliefs of practicing teachers is not astraightforward one. Because self-efficacyis a motivational construct associated withthe willingness to try new practices, persis-tence, and resilience in the face of setbacks,the expectation of the researchers was thatthe inclusion of additional sources of self-efficacy in the four formats would leadto incremental increases in teacher self-efficacy that would result in incrementalincreases in implementation of the new in-structional strategy. A more complex pat-tern emerged, however, of both increasesand decreases in self-efficacy beliefs.

The results of this study demonstratedthat verbal persuasion, vicarious experi-ences, and even a limited mastery experi-ence did not prove to be particularly pow-erful in creating the conditions to supportimplementation of a new instructionalstrategy when they took place in a largegroup setting. Furthermore, the resultslend support to the importance of an au-thentic task-specific mastery experienceand of individualized verbal persuasion inraising self-efficacy beliefs and supportingimplementation of a new teaching strategy(Bandura, 1997). Only in the real setting cana teacher experience a true test of his or hercapabilities. Thus, as teachers’ bolsteredself-efficacy supports their successful im-plementation of the new strategy, the im-plementation experience becomes a mas-tery experience that contributes to futureself-efficacy assessments. The findings ofthis study support professional develop-ment models that mediate the effect ofthe reassessment of participants’ analysisof the teaching task and their personalcompetency as they participate in skill-development programs (Tschannen-Moran,Woolfolk Hoy, & Hoy, 1998).

The relationship evidenced betweenprofessional development format and im-plementation of a new instructional strat-egy in this study aligns with previousfindings that the form that professional de-velopment takes is an important variable inthe likelihood that the instructional strate-

gies introduced will be implemented in theclassroom. Previous researchers have alsofound that experiences limited to telling(verbal persuasion) and watching others(vicarious experiences) were minimally ef-fective in leading to instructional changeand have emphasized the need for provid-ing teachers with feedback and support intheir implementation efforts (Guskey, 1989;Joyce & Showers, 1988; Stein & Wang,1988). In the present study, professional de-velopment training that included follow-upcoaching, in which participants receivedsupport as they developed the new teach-ing skill, was related to increased imple-mentation. The lack of relationship be-tween implementation and the first threetraining formats underscores the impor-tance of these follow-up coaching experi-ences in providing teachers needed assis-tance as they attempt to implement a newskill.

The primary-grade teachers in thisstudy tended to view themselves as doing agood job of teaching in general, and specif-ically of teaching reading, as evidenced bythe fairly high mean scores of both teacherself-efficacy in general and teacher self-efficacy for reading instruction in the initialdata collection. It is also true that in each ofthe nine schools, second- and third-gradestudents were found for the demonstra-tions that had not yet started reading. Inmost cases, those children had been thestudents of several of the teachers present.Prior to the Tucker workshop, it may havebeen that when these teachers had a stu-dent in their class who did not learn toread, it did not disrupt their self-efficacy asteachers because they consoled themselveswith the belief that a few students simplycan’t be taught to read during the primaryyears. Receiving as passive observers infor-mation about a teaching strategy that mostwere unfamiliar with in Treatment 1 mayhave challenged some of the teachers’ as-sumptions about the requirements of read-ing instruction, but if, as Gregoire (2003)has proposed, they did not feel personally

242 THE ELEMENTARY SCHOOL JOURNAL

DECEMBER 2009

This content downloaded from 195.78.108.157 on Wed, 21 May 2014 00:58:59 AMAll use subject to JSTOR Terms and Conditions

implicated, they may not have engaged inthe systematic processing of the informa-tion. When the teachers in Treatment 2 wit-nessed a demonstration in which strug-gling readers who had been students intheir classes and had not yet learned toread were taught to decode complex wordswithin a 20-minute demonstration, many ofthem may have felt personally implicated.Witnessing the trainer achieving successwhere they had failed may have causedsome teachers to adjust their evaluation oftheir own capability to teach reading bycreating new, higher standards for accept-able practice. The added component ofpractice in Treatment 3 allowed an oppor-tunity for limited skill development butalso revealed the difficulties inherent inputting it into practice, such that even moreteachers began to interpret the interventionas a threat as opposed to a challenge (Gre-goire, 2003), resulting in a drop in theirself-efficacy for teaching reading. Thefollow-up conversations and assistance re-ceived during coaching may have allowedfor deeper processing of the informationpresented and helped bring teachers’ skilllevels in line with their revised standards ofgood practice, resulting in strengthenedself-efficacy.

A substantial proportion of the partici-pants in this study experienced a decreasein their self-efficacy for reading instructionwith exposure to this new instructionalstrategy. This dip in self-efficacy beliefs iscomparable with previous studies thathave found that in the initial stages of im-plementation of a new skill, self-efficacybeliefs may suffer (Guskey, 1984; Ross,1994; Stein & Wang, 1988; Woolfolk Hoy &Burke-Spero, 2005). It is interesting that thegreatest “dips” in self-efficacy came amongtreatment groups with only a small level ofimplementation. These decreases in self-efficacy for reading instruction may pro-vide further evidence that teachers’ calibra-tion of the level of their content knowledgeis not especially accurate (Cunningham etal., 2004), such that it is vulnerable to a

reassessment in the face of evidence of ateaching strategy that is shown to be suc-cessful with students with whom they havefailed. Wheatley (2002) challenged initia-tives for promoting increases in teacher self-efficacy beliefs as too simplistic and has pro-posed that entertaining doubts about one’sefficacy might actually provoke greateropenness to new ideas. Thus if doubts,as reflected in decreases in teacher self-efficacy beliefs, are part of an expectedgrowth pattern as teachers learn to imple-ment a new skill, they should be viewed asnatural and desirable reflections that helpcreate conditions of readiness to benefitfrom professional learning. To overcomethese doubts, however, teachers may needassistance such as follow-up coaching inconfronting and resolving the anxiety anduncertainty inherent in changing profes-sional practice (Guskey, 1988; Ross, 1992).There is a need for more research to explorethe effects of teacher self-efficacy doubts onteacher development as they attempt to in-crease their skill and stabilize their imple-mentation of new strategies.

That implementation of the readingstrategy introduced was related to finalself-efficacy for reading instruction and notto general self-efficacy is of note because itdemonstrates the importance of assessingself-efficacy as a task-specific construct(Bandura, 1997). Although teachers whoparticipated in this professional develop-ment program generally came away with astronger sense of their capability as teach-ers, this was not necessarily related to theirwillingness or ability to implement the in-structional strategy introduced. Increasesin self-efficacy for reading instruction mayhave led teachers to take the risks involvedin attempting a new instructional strategy,or conversely it may have been the result ofsuccessfully having implemented the newstrategy and having witnessed improvedstudent learning as a result. Scholars haveproposed that teacher self-efficacy is a cy-clical process in which self-efficacy be-comes both a product and a constructor of

SOURCES OF SELF-EFFICACY 243

This content downloaded from 195.78.108.157 on Wed, 21 May 2014 00:58:59 AMAll use subject to JSTOR Terms and Conditions

experiences (Tschannen-Moran, WoolfolkHoy, & Hoy, 1998). Although evidence ofincreased student learning was not in-cluded in the study, Guskey (1988, 1989)has likewise noted that teacher self-efficacyand willingness to innovate are both acause and an outcome of witnessing im-provements in student achievement.

The findings of this study revealed thecomplexity of the development of self-efficacy and the kinds of support neededfor teachers to implement a new instruc-tional strategy. Future studies are neededto probe these complexities further. Howmuch self-efficacy is enough to sustain theself-confidence needed to take on the risksinherent in change, and how much is toomuch, leading to complacency and an im-munity to change? If decreases in self-efficacy are to be expected, and to someextent are even helpful to fostering a moti-vation for change, at what point do theybecome debilitating? Are lecture-basedworkshops that raise self-efficacy but arenot powerful enough to lead to implemen-tation to be abandoned, or do they have arole to play in setting the stage for futurechange? Is it counterproductive to presentteachers with compelling evidence of newinstructional strategies that raise their stan-dards of professional practice without pro-viding the kinds of supports needed tomeet those higher standards themselves?How much coaching is needed to sustaindeep conceptual change and lasting in-structional changes? Which components ofcoaching are the most powerful in support-ing self-efficacy beliefs and implementa-tion? These questions offer rich new ave-nues for future investigation into the role ofself-efficacy in the professional develop-ment of teachers.

References

Anderson, R., Greene, M., & Loewen, P. (1988).Relationships among teachers’ and students’thinking skills, sense of efficacy, and student

achievement. Alberta Journal of EducationalResearch, 34(2), 148–165.

Ashton, P., & Webb, R. (1986). Making a differ-ence: Teachers’ sense of efficacy and studentachievement. New York: Longman.

Bandura, A. (1977). Self-efficacy: Toward a uni-fying theory of behavioral change. Psycho-logical Review, 84, 191–215.

Bandura, A. (1997). Self-efficacy: The exercise ofcontrol. New York: W. H. Freeman.

Cole, C., & Majd, M. (2003). Tucker Signing Strat-egies for Reading national study. Blooming-ton: Indiana University. Retrieved June 25,2009, from http://www.ahaprocess.com/files/Tuckernationalstudy1.pdf

Cunningham, A. E., Perry, K. E., Stanovich,K. E., & Stanovich, P. J. (2004). Disciplinaryknowledge of K–3 teachers and their knowl-edge calibration in the domain of early liter-acy. Annals of Dyslexia, 54, 139–167.

Gregoire, M. (2003). Is it a challenge or a threat?A dual process model of teachers’ cognitionand appraisal processes during conceptualchange. Educational Psychology Review, 15,147–179.

Guskey, T. (1984). The influence of change ininstructional effectiveness upon the affectivecharacteristics of teachers. American Educa-tional Research Journal, 21, 245–259.

Guskey, T. (1986). Staff development and theprocess of teacher change. Educational Re-searcher, 15(5), 5–12.

Guskey, T. (1988). Teacher efficacy, self-concept,and attitudes toward the implementationof instructional innovation. Teaching andTeacher Education, 4(1), 63–69.

Guskey, T. (1989). Attitude and perceptualchange in teachers. International Journal ofEducational Research, 13(4), 439–454.

Guskey, T., & Passaro, P. (1994). Teacher effi-cacy: A study of construct dimensions.American Educational Research Journal, 31,627–643.

Haney, J. J., Wang, J., Keil, C., & Zoffel, J. (2007).Enhancing teachers’ beliefs and practicesthrough problem-based learning focused onpertinent issues of environmental health sci-ence. Journal of Environmental Education,38(4), 25–33.

Joyce, B., & Showers, B. (1988). Student achieve-ment through staff development. White Plains,NY: Longman.

McKinney, M., Sexton, T., & Meyerson, M.(1999). Validating the efficacy-based changemodel. Teaching and Teacher Education, 15,471–485.

244 THE ELEMENTARY SCHOOL JOURNAL

DECEMBER 2009

This content downloaded from 195.78.108.157 on Wed, 21 May 2014 00:58:59 AMAll use subject to JSTOR Terms and Conditions

McMaster, P. (2003). Learning through gestur-ing: The Tucker reading strategy. VirginiaEducational Leadership, 2(1), 39–43.

Midgley, D., Feldlaufer, H., & Eccles, J. (1989).Change in teacher efficacy and student self-and task-related beliefs in mathematics dur-ing the transition to junior high school. Jour-nal of Educational Psychology, 81(2), 247–258.

Poole, M., & Okeafor, K. (1989). The effects ofteacher efficacy and interactions among ed-ucators on curriculum implementation. Jour-nal of Curriculum and Supervision, 4(2), 146–161.

Raudenbush, S. W., Rowan, B., & Cheong, Y. F.(1992). Contextual effects on the self-perceived efficacy of high school teachers.Sociology of Education, 65, 150–167.

Ross, J. (1992). Teacher efficacy and the effects ofcoaching on student achievement. CanadianJournal of Education, 17(1), 51–65.

Ross, J. (1994). The impact of an inservice topromote cooperative learning on the stabil-ity of teacher efficacy. Teaching and TeacherEducation, 10(4), 381–394.

Ross, J., & Bruce, C. (2007). Professional devel-opment effects on teacher efficacy: Results ofrandomized field trial. Journal of Research inEducation, 101, 50–60.

Ross, J. A., Cousins, J. B., & Gadalla, T. (1996).Within-teacher predictors of teacher effi-cacy. Teaching and Teacher Education, 12, 385–400.

Ross, J. A., Cousins, J. B., Gadalla, T., & Hannay,L. (1999). Administrative assignment ofteachers in restructuring secondary schools:The effect of out-of-field course responsibil-ity on teacher efficacy. Educational Adminis-tration Quarterly, 35, 782–804.

Scribner, J. (1999). Teacher efficacy and teacherprofessional learning: Implications for schoolleaders. Journal of School Leadership, 9, 209–234.

Shahid, J., & Thompson, D. (2001). Teacher effi-cacy: A research synthesis. Paper presented atthe annual meeting of the American Educa-

tional Research Association, Seattle (ERICDocument Reproduction No. ED 453 170).

Smylie, M. (1988). The enhancement function ofstaff development: Organizational and psy-chological antecedents to individual teacherchange. American Education Research Journal,25, 1–30.

Stein, M., & Wang, M. (1988). Teacher develop-ment and school improvement: The processof teacher change. Teaching and Teacher Edu-cation, 4, 171–187.

Timperley, H., & Phillips, G. (2003). Chang-ing and sustaining teachers’ expectationsthrough professional development in liter-acy. Teaching and Teacher Education, 19,627– 641.

Tschannen-Moran, M., & Johnson, D. (2004,April). Teachers’ sense of efficacy for literacyinstruction. Paper presented at the annualmeeting of the American Educational Re-search Association, San Diego.

Tschannen-Moran, M., & Woolfolk Hoy, A.(2001). Teacher efficacy: Capturing an elu-sive construct. Teaching and Teacher Educa-tion, 17, 783–805.

Tschannen-Moran, M., & Woolfolk Hoy, A.(2007). The differential antecedents of self-efficacy beliefs of novice and experiencedteachers. Teaching and Teacher Education, 23,994–956.

Tschannen-Moran, M., Woolfolk Hoy, A., &Hoy, W. K. (1998). Teacher efficacy: Itsmeaning and measure. Review of EducationalResearch, 68, 202–248.

Tucker, B. (2001). Tucker Signing Strategies forReading. Highland, TX: Aha! Process, Inc.

Wheatley, K. (2002). The potential benefits ofteacher efficacy doubts for educational re-form. Teaching and Teacher Education, 18,5–22.

Woolfolk Hoy, A., & Burke-Spero, R. (2005).Changes in teacher efficacy during the earlyyears of teaching: A comparison of fourmeasures. Teaching and Teacher Education, 21,343–356.

SOURCES OF SELF-EFFICACY 245

This content downloaded from 195.78.108.157 on Wed, 21 May 2014 00:58:59 AMAll use subject to JSTOR Terms and Conditions