teacher learning and abstract instructional change… · teacher learning and instructional change:...

Teacher Learning andInstructional Change:How Formal andOn-the-Job LearningOpportunities PredictChange in ElementarySchool Teachers’Practice

Leigh Mesler PariseJames P. SpillaneNorthwestern University

Abstract

Recent education reform has emphasized theimportance of teacher learning in improvingclassroom instruction and raising studentachievement. This article focuses on teachers’learning opportunities, including formal profes-sional development and on-the-job learning thatoccurs through interactions with colleagues. Us-ing data from 30 elementary schools in a mid-sized urban school district, the authors con-currently explore the relationships betweenteachers’ formal professional development andon-the-job learning opportunities and instruc-tional change. Results suggest that formal pro-fessional development and on-the-job opportu-nities to learn are both significantly associatedwith changes in teachers’ instructional practicein mathematics and English language arts.

Recent education reform in the UnitedStates has increasingly defined acceptablelevels of mastery for students and centeredon holding schools accountable for studentoutcomes. As one strategy for raising stu-dent achievement, policymakers have fo-cused on improving the quality of publicschool teachers (Borko, 2004; Corcoran,1995b; Desimone, Porter, Garet, Yoon, &Birman, 2002; Garet, Porter, Desimone, Bir-man, & Yoon, 2001; Lieberman, 1995). Somepolicy initiatives focus on improving thequality of teachers entering the professionthrough state certification tests, more strin-gent degree requirements, and recruitmentefforts. At the same time, increased ac-countability pressure on schools requireslearning and change for the thousands ofteachers already in service, as they arepressed to implement new instructional ap-proaches in order to raise student achieve-ment (Cohen & Barnes, 1993; Cohen & Hill,

The Elementary School JournalVolume 110, Number 3© 2010 by The University of Chicago. All rights reserved.0013-5984/2010/11003-0004$10.00

2002; Drake, Spillane, & Hufford-Ackles,2001; Garet et al., 2001; Wilson & Berne,1999). There is great faith among schoolreformers and education researchers thataugmenting the learning opportunities ofpracticing teachers will enhance teacherperformance and lead to improved studentoutcomes (Borko, 2004). However, the extentto which teachers’ learning opportunities fa-cilitate change in their classroom practiceremains unclear. This study explores theempirical relationship between teachers’learning opportunities and changes in theirinstructional practice.

Seeking to understand the kinds oflearning opportunities to which teachershave access, educational researchers havefollowed two somewhat distinct lines ofresearch. The first has focused on teachers’formal learning opportunities, includingstructured professional development activ-ities and graduate education (Borko, 2004;Desimone, Porter, Garet et al., 2002; Garetet al., 2001; Guskey, 2002); the second lineof research has centered on teachers’ on-the-job learning and explored aspects ofschools’ organizational conditions that mayaffect teacher learning and change (Bryk,Camburn, & Louis, 1999; Louis, Marks, &Kruse, 1996; McLaughlin & Talbert, 2001;Scribner, Cockrell, Cockrell, & Valentine,1999; Wilson & Berne, 1999). While much ofthis work has been descriptive, researchersin both areas have begun to determinewhich learning opportunities are most ef-fective at facilitating change. However,these lines of research remain separate inthe empirical literature, which is problem-atic for both policy and practice because itis unclear whether time and money shouldbe spent on expanding teachers’ formalprofessional development, on working tobetter enable teachers to learn from theircolleagues on the job, or on some combina-tion of the two approaches.

This study seeks to marry these twolines of research by concurrently exploringthe empirical links between both formaland on-the-job learning opportunities and

teacher change. We begin with a review ofthe literature on teachers’ formal and on-the-job learning opportunities, as well asthe organizational conditions that may af-fect teacher learning. In our article, we usethe term formal learning opportunities to referto subject-specific professional develop-ment sessions, out-of-school teacher net-works, and coursework in math and En-glish. On-the-job learning opportunities referto interactions with colleagues aroundteaching and learning, including conversa-tions about instruction, peer observationand feedback, and advice seeking about in-struction. After reviewing the literature, wedescribe the methodology used to analyzethe relationship between teacher learningopportunities and teacher change in math-ematics and English language arts (ELA)teaching practice for elementary schoolteachers in a mid-sized urban school dis-trict. Finally, we report our main findings,which suggest that both formal profes-sional development and teachers’ on-the-job learning opportunities are statisticallysignificant predictors of teacher change inmath and ELA instruction. We concludewith a discussion of the implications of ourfindings for policy and practice.

Empirical and Theoretical AnchorsOur work is anchored in literature address-ing teachers’ opportunities to learn, includ-ing their formal professional developmentand the learning opportunities afforded bytheir interactions with colleagues on thejob. A major challenge in the existing liter-ature is that empirical studies of profes-sional development remain separate fromstudies of teachers’ on-the-job learning,though a number of theoretical pieces havejointly discussed them (e.g., Corcoran,1995a; Putnam & Borko, 2000). As a result,empirical research on teachers’ opportuni-ties to learn lacks a cohesive and compre-hensive framework for understanding andintegrating the various learning opportuni-ties that may affect teacher practice. In ad-

324 THE ELEMENTARY SCHOOL JOURNAL

MARCH 2010

dition, although some studies have takensubject matter into consideration by nar-rowing their scope to a single curriculardomain, work on formal and on-the-joblearning opportunities fails to make clear ifor how teachers’ opportunities to learn mayhave differential effects by school subject.In this section, we explore the extant liter-ature, arguing that these two lines of re-search must be bridged in order to furtherour understanding of teacher learning andchange.

Formal Learning OpportunitiesOver the last 20 years, amid calls for

changes in teaching practice and mountingefforts to increase the professionalization ofteaching, reformers and educators haveworked to expand professional develop-ment opportunities for teachers (Wilson &Berne, 1999). One example of policymakers’faith that increasing teacher participation informal learning opportunities will helpproduce desirable teacher and student out-comes is the requirement in the No ChildLeft Behind Act of 2001 that states ensurethe availability of high-quality professionaldevelopment for teachers in order to im-prove classroom instruction (Borko, 2004).In addition, state and district policies re-quire teachers to participate in formallearning opportunities, with the modalstate requirement being 120 hours over 5years (Hill, 2007). States also often requirethat teachers receive master’s degrees, al-low graduate coursework to count towardrecertification, and provide strong financialincentives for acquiring advanced degrees(Goldhaber & Brewer, 1998; Hill, 2007).While it is difficult to get an exact estimateof expenditures on teachers’ formal pro-fessional development (Corcoran, 1995a),schools, districts, states, and the federalgovernment spend at least millions, andlikely billions, of dollars on professionaldevelopment for teachers (Borko, 2004).This is a huge investment for which weknow little about returns (Rice, 2001).

Given the increased policy and fiscalemphasis on using formal learning oppor-tunities to facilitate teacher change, this sec-tion addresses the types of formal learningopportunities teachers experience and therelationships between formal learning op-portunities and changes in teacher practiceand student achievement. The literaturesuggests that the majority of teachers’ for-mal learning opportunities are in the formof workshops, special courses, graduatecoursework, and in-service days or confer-ences devoted to training teachers in a spe-cific set of ideas, techniques, or materials(Desimone, Porter, Garet et al., 2002; Garetet al., 2001; Hill, 2007; Little, 1993; NCES,2005). These opportunities traditionally fol-low what Little (1993) termed the “trainingparadigm,” in that they occur outside ofteachers’ classrooms at scheduled timesand are led by an expert seeking to train, orcommunicate new information to, groupsof teachers (Corcoran, 1995a; Feiman-Nemser, 2001). However, as discussed be-low, this type of professional developmentis not likely to facilitate change in teacherpractice.

Although most of teachers’ formallearning opportunities follow the trainingparadigm, teachers are increasingly partic-ipating in other types of formal profes-sional development that offer markedlydifferent opportunities. Referred to as “re-form professional development” by Garet,Desimone, and colleagues (Desimone, Por-ter, Garet et al., 2002; Garet et al., 2001;Porter, Garet, Desimone, Yoon, & Birman,2000), such opportunities frequently relatemore closely to teachers’ classroom con-texts than traditional activities, often in-volve active participation and collaborationbetween teachers, and may take place dur-ing the regular school day in teachers’classrooms or schools (Desimone, Porter,Garet et al., 2002; Garet et al., 2001). Reformprofessional development may involve or-ganized teacher study groups or networks,committees, mentoring, internships, andresource centers (Garet et al., 2001). Al-

TEACHER LEARNING AND CHANGE 325

though these types of learning opportuni-ties have become more widely available toteachers in recent years, the majority ofteachers do not participate in any reform-type formal learning opportunities (Garetet al., 2001; NCES, 2005; Porter et al., 2000).

A significant amount of work has fo-cused on describing the types of formallearning opportunities to which teachershave access, but considerably less researchhas linked these opportunities to teacherchange and student achievement. Whilethis work suggests that most of the profes-sional development that teachers receive,particularly graduate coursework and sin-gle workshops that follow the training par-adigm, is not consistently linked to changesin classroom practice (Cohen & Hill, 2002;Garet et al., 2001), it has also identified cer-tain characteristics of formal learning op-portunities that make them more likely tofacilitate learning and change. Specifically,activities are most effective at fosteringteacher change when they involve collec-tive participation of teachers from the sameschool, grade, or subject, are linked to teach-ers’ on-going daily activities, provide activelearning opportunities, and are content-specific (Desimone, Porter, Birman, Garet,& Yoon, 2002; Garet et al., 2001, 2008). Ingeneral, reform activities are more likelythan traditional activities to include theseattributes and they tend to be more success-ful at fostering change in teaching practice(Birman, Desimone, Porter, & Garet, 2000;Desimone, Porter, Garet et al., 2002; Garetet al., 2001).

With the exception of Garet and col-leagues’ 2008 study of reading professionaldevelopment, most of the recent empiricalwork regarding the effects of formal pro-fessional development on teacher changehas focused on mathematics. By exploringthe effects of subject-specific professionaldevelopment on teacher change, thesestudies have taken seriously the work ofStodolsky, Siskin, and others (e.g., Siskin,1991; Stodolsky, 1988; Stodolsky & Gross-man, 1995) who have suggested that the

subject matters when it comes to teacherpractice. However, it remains unclearwhether the aforementioned activities helpfoster change similarly across all subjectareas, or if, for example, certain opportuni-ties better facilitate learning in math than inELA.

While fewer studies have focused onthe relationship between teachers’ formallearning opportunities and student achieve-ment, there is some evidence linking higherstudent math achievement with teacherparticipation in sustained formal profes-sional development that is grounded incontent-specific pedagogy (Carpenter, Fen-nema, Peterson, Chiang, & Loef, 1989;Darling-Hammond, 2000; McCutchen et al.,2002; Saxe, Gearhart, & Nasir, 2001) andwell aligned with policy changes (Cohen &Hill, 2002; Hill, 2007). While empiricalwork is beginning to reveal which types ofactivities may be most effective at influenc-ing teacher and student outcomes, the factremains that the vast majority of the formallearning opportunities in which teachersparticipate are the kind of one-shot trainingsessions that research suggests are notlikely to facilitate teacher learning andchange (Desimone, Porter, Garet et al.,2002; Hill, 2007; NCES, 2005).

On-the-Job Learning Opportunitiesand Organizational ConditionsWhile formal learning opportunities

have taken center stage in the policy arena,some researchers have also focused on howteachers learn from their colleagues on thejob, outside of formal professional develop-ment activities. Work addressing on-the-job learning opportunities suggests thatlearning is fostered when professionalswork alongside others (Eraut & Hirsh,2007) asking questions and gathering infor-mation (Eraut & Hirsh, 2007; Frank, Zhao,& Borman, 2004), observing colleagues (Er-aut, 2004), and giving and receiving feed-back (Eraut & Hirsh, 2007). In this sectionwe review recent scholarship on teachers’


MARCH 2010

on-the-job interactions with coworkers andthe organizational conditions that may af-fect teacher learning.

On-the-job learning opportunities. Teach-ers’ on-the-job learning opportunities mayoccur throughout the school day in a widevariety of activities, including conversa-tions between teachers in the hallway, in-teractions with students, planning sessionswith colleagues, and meetings with par-ents. While we acknowledge that all ofthese instances may be important to indi-vidual teachers’ development, the primaryfocus of this piece is on those learning op-portunities that involve interaction be-tween colleagues, as studies have indicatedthat such activities constitute important po-tential learning opportunities (e.g., Little,2002; Smylie, 1995).

The extant literature on teacher collab-oration indicates that learning is fosteredwhen teachers engage in conversationsabout new material (Davis, 2003), discussstrategies for effective teaching (Brownell,Yeagar, Rennels, & Riley, 1997; Little, 2003),push one another to experiment aroundnew initiatives (Davis, 2003), work collab-oratively to share expertise (Little, 2003;Smylie, 1995), and interpret policy mes-sages (Coburn, 2001). Recent work byGoddard, Goddard, and Tschannen-Moran(2007) suggested that “the more teacherscollaborate, the more they are able to con-verse knowledgably about theories, meth-ods, and processes of teaching and learn-ing, and thus improve their instruction” (p.879). Further, this work explored the em-pirical links between teacher collaborationand student achievement and, while theauthors’ operationalization of collaborationalso included teachers’ participation inschool decision making, they found thathigher levels of teacher collaboration wereassociated with higher student achieve-ment on high-stakes tests in both math andreading, after controlling for school and in-dividual factors (Goddard et al., 2007). Inaddition, Bryk et al. (1999) found that whenteachers engaged in peer observation and

feedback, opening their practice up to scru-tiny by a colleague, they learned abouttheir colleagues’ teaching practices andwere encouraged to “ask questions abouttheir practice and to view it in a more an-alytic fashion” (p. 754).

Studies have also suggested that thestrength of interpersonal relationships maybe important when it comes to learningfrom interactions with colleagues. Social in-teractions, and specifically advice seeking,are associated with the transfer of informa-tion, which is essential for learning andknowledge development (Frank et al., 2004;Reagans & McEvily, 2003; Uzzi, 1997). Inschools, strong ties support teachers’ jointsense-making about instructional policyand reform, which can enable high-fidelityimplementation (Coburn, 2001; Spillane,1999). Additionally, social interactions thatspan an organization’s boundaries mayalso be important for learning becausethey can provide access to new informa-tion and potentially minimize conformityand group think among organizationalmembers (Hansen, 1999; Leana & Pil, 2006;Reagans & McEvily, 2003; Wenger, 1998).

Organizational conditions. While someresearch has centered on the specific behav-iors or activities involved in teachers’ on-the-job learning opportunities, other work hasaddressed the conditions under which theseopportunities typically take place—those or-ganizational arrangements and norms thatmay support learning. This work often incor-porates teacher behaviors (e.g., looking atstudent work and discussing instruction), butit also extends the discussion of on-the-joblearning by defining the quality of teacherrelationships and characteristics of the schoolorganization that foster teacher learning andchange (Bryk et al., 1999; Bryk & Schneider,2002; Coburn, 2001; Lee & Smith, 1996; Little,1982; Spillane, 1999). When the school is char-acterized by norms of trust among teachersand between teachers and administrators(Bryk & Schneider, 2002; Tschannen-Moran,2001), beliefs regarding collective responsibil-ity for student learning (Lee & Smith, 1996),


and shared norms supporting opennessand innovation (Louis et al., 1996; Rosen-holtz, 1985; Scribner, Hager, & Warne,2002; Smylie, 1988), staff engagement in on-the-job learning activities is both morelikely to occur and more likely to producechange. Studies have also shown that schoolsscoring high on these measures are better atraising student achievement (Louis & Marks,1998), fostering specific instructional changes(Bryk et al., 1999), and implementing newpolicies (Coburn, 2001; Spillane, 1999).

In addition, school leaders play an im-portant role in establishing a school’s orga-nizational context. In doing so, they toomay affect teacher learning in the work-place. School leaders who endorse knowl-edge sharing among teachers and createinternal structures that promote collabora-tion are most effective at fostering changewithin their schools (Bryk et al., 1999; Bryk& Schneider, 2002; Fullan, 2002; Youngs &King, 2002). Furthermore, school leaderswho communicate clear expectations toteachers and concrete goals for studentachievement can encourage teachers to im-prove their practice (Leithwood, 1992). As awhole, this work suggests that teachers’ on-the-job learning opportunities and theirschools’ organizational conditions play animportant role in fostering teacher learningand change.

While the last 2 decades of researchhave greatly expanded the knowledge baseon teacher learning, the field remains some-what segmented into two broad areas. Al-though some theoretical and descriptivework on teacher learning couples teachers’formal and on-the-job learning opportu-nities (e.g., Corcoran, 1995a; Putnam &Borko, 2000), empirical studies of teachers’learning opportunities have yet to do so.While we acknowledge that there maysometimes be overlap between formal andon-the-job learning opportunities, we feelthat the data give us traction for looking atdifferences between the types of learningopportunities in which the teachers in oursample engaged in their schools. Our work

seeks to understand the relative links be-tween elementary school teachers’ formaland on-the-job learning opportunities andtheir changes in their classroom practice inmathematics and ELA. Specifically, we ex-plore the following questions: When teach-ers’ formal and on-the-job opportunities tolearn are taken into account, what are therelative impacts of these different activitieson teacher change in instructional practice?Are these relationships mediated by teach-ers’ perceptions of their schools’ organiza-tional conditions? Finally, are the asso-ciations between formal and on-the-joblearning opportunities different for teach-ers’ changes in math and ELA instruction?

MethodData: Sample and Data-CollectionProceduresThe data for this study come from an

evaluation of a leadership professional de-velopment program in a mid-sized urbanschool district in the southeastern UnitedStates. Data were collected from all of thedistrict’s 30 elementary schools. The aver-age school had approximately 600 students,65% black students, 28% white students,and 64% of students who qualified for freeor reduced-price lunch. As part of a mixed-method evaluation, school staff members inthe 30 elementary schools were asked tocomplete an 18-page questionnaire. Ques-tionnaire items were primarily closed-ended and asked about the school staffmembers’ work in and out of the classroomand their involvement in school improve-ment efforts. The subject sections askedquestions about the school as a workplace,school leadership, professional develop-ment and school change, and the respon-dent’s background. In two open-endedquestions, respondents were also asked todescribe their in-school social networks bylisting the names of people from whomthey seek advice about mathematics andreading/language arts or English instruc-tion. Surveys were administered to in- and


MARCH 2010

out-of-classroom personnel, except theprincipals, who completed a different in-strument, in each of the district’s 30 ele-mentary schools at two time points: springof 2005 and spring of 2007. In total, 1,210elementary school staff members re-sponded to the survey in 2005 and 1,194responded in 2007; respective responserates were 89% and 83%. The sample forthe present study was limited to self-contained kindergarten through fifth-gradeclassroom teachers responsible for bothmath and ELA instruction. After selectingthese respondents, 714 respondents from2005 and 704 respondents from 2007 wereincluded in the analyses.

Scale DevelopmentRelevant to this study, the school staff

questionnaire included questions regard-ing teachers’ formal professional develop-ment, on-the-job learning opportunities,perceptions of school organizational condi-tions, and individual characteristics. Themeasures included in subsequent analysesare detailed below. In addition, specificitems and alphas for each scale are in-cluded in the Appendix. All scales werereliable with alphas above or equal to .70,and most were highly reliable with alphasabove .90 (Cronbach, 1951). Scales used inthis study were developed using a combi-nation of previous empirical work as wellas relevant literature on each dimensionof on-the-job learning opportunities andschool organizational conditions. Previousscale development using the same teachersurvey identified reliable constructs relatedto teachers’ learning opportunities and theschool organization (Goldring, Huff, Stitz-iel Pareja, & Spillane, 2008; Goldring, Spil-lane, Huff, Barnes, & Supovitz, 2006; Supo-vitz, Sirinides, & May, in press). Supovitz etal. (in press) conducted confirmatory factoranalysis to validate the use of similar scalesfrom the teacher survey by examining sur-vey items related to teacher change in in-struction, collaborative discussion, peer

observation and feedback, teachers’ per-ceptions of the school environment, andprincipal leadership. The authors providedinitial group membership, permitting itemsto then migrate iteratively to dimensionsthat better explain item variance, but noitem migrated from its hypothesized di-mension.

MeasuresThis section details the measures used

in subsequent analyses. The dependentvariables are Change in Math TeachingPractice and Change in ELA Teaching Prac-tice. The measures of formal learning op-portunities are Math and ELA ProfessionalDevelopment, Math and English Courses,and Outside Network Participation. Themeasures of on-the-job learning opportuni-ties are Collaborative Discussion, Peer Ob-servation and Feedback, and Math andELA Advice Seeking. Finally, the measuresof organizational conditions are Profes-sional Learning Community and PrincipalDevelops Goals. The analyses also includea number of controls, including teacher ef-ficacy, race, and gender, as described be-low.

Change in math teaching practice andchange in ELA teaching practice. On a7-point scale ranging from not at all to agreat deal, participants were asked to indi-cate how much they changed their teachingthis year for the following items: studentassessment, student grouping, materialsused, topics covered, teaching methodsused, kinds of work students do, kinds ofquestions asked, and understanding of theneeds of individual students in their class.While the link between changes in teacherpractice and student achievement is oftenimplicit in education policies, recent workusing these data has found significant linksbetween teachers’ changes in practice andstudent achievement on state assessments(Supovitz et al., in press). Respondents an-swered separately for math and ELA, andthe items for each subject were then aver-


aged to create the math change and ELAchange variables.

Math professional development andELA professional development. Math pro-fessional development and ELA profes-sional development are two measures ofteachers’ formal learning opportunities. Ona 4-point scale ranging from none to 8�sessions, participants were asked to indicatethe number of professional developmentsessions they participated in this year in (1)mathematics teaching, and (2) reading/lan-guage arts or English teaching.

Math and English courses. Additionalmeasures of formal learning opportunitiesincluded in subsequent analyses are teach-ers’ coursework in math and English. On a6-point scale ranging from none to 16�classes, participants were asked to indicatethe number of undergraduate or graduatelevel courses they had taken in mathemat-ics and in English or a related language artsfield.

Outside network participation. On a6-point scale ranging from never to 10 ormore times, respondents were asked to indi-cate how often they participated in a net-work with other teachers outside of theirschool this year. Characterized by Garet etal. (2001) as a type of reform professionaldevelopment, teacher networks outside ofthe school may provide teachers with op-portunities to learn and may bring new in-formation into teachers’ own schools (Re-agans & McEvily, 2003; Wenger, 1998). Weseparate this measure from the math andELA professional development variablesbecause we asked separate questions abouttheir attendance at math and English pro-fessional development and about their par-ticipation in a teacher network outside ofthe school.

Collaborative discussion. The collabo-rative discussion measure captures teach-ers’ behaviors around conversation withcolleagues regarding teaching and learn-ing. On a 5-point scale ranging from neverto more than 2 days/week, participants wereasked questions regarding their conversa-

tions with colleagues around issues ofteaching and learning. On the same 5-pointscale, respondents were also asked to indi-cate how often they had in-depth discus-sions about their teaching with anotherclassroom teacher. Finally, on a 7-pointscale ranging from never to more than once aweek, respondents were asked to indicatehow often they had scheduled meetingswith other teachers in the school to dis-cuss and plan curriculum or teaching ap-proaches. Standard scores were calculatedfor each individual item, and items werethen averaged to create the collaborativediscussion variable.

Peer observation and feedback. On a5-point scale ranging from never to morethan 2 days/week, participants were asked toindicate how often they participated in fourdifferent observation and feedback activi-ties around instruction and student work.Researchers have found that when teachersengage in observation and feedback, theymay learn from colleagues and they askmore questions to improve their own teach-ing practice (Bryk et al., 1999; Little, 1990).Items were averaged to create the peer ob-servation and feedback variable.

Math advice seeking and ELA adviceseeking. The term out degree is a measureused in network analysis designed to captureadvice-seeking interactions and opportuni-ties for learning around specific subject mat-ter. Respondents were asked, “To whom doyou turn for advice or information aboutmathematics instruction?” and “To whom doyou turn for advice or information aboutreading/language arts or English instruc-tion?” Respondents could list up to sevendifferent sources of information on each sub-ject. As an indicator of tie strength, respon-dents were also asked to indicate how oftenthey turned to each source for advice, rang-ing from yearly to daily. We call this variableadvice seeking here in order to ease interpre-tation for the reader. The math advice seek-ing and ELA advice seeking measures werecreated by totaling the frequency with which


MARCH 2010

subject-specific advice was sought from allsources listed.1

Professional learning community. As ameasure of teachers’ perceptions regardingtheir school’s organizational conditions, re-spondents were asked about the school’sprofessional learning community. On a4-point Likert scale ranging from stronglydisagree to strongly agree, teachers re-sponded to items about teacher trust, open-ness between teachers, and support forteacher innovation. On a 5-point scale rang-ing from none to nearly all, teachers werealso asked questions about how manyteachers in the school took collective re-sponsibility for school improvement andstudent learning. Standardized items wereaveraged to create the professional learningcommunity variable. We recognize that thismeasure groups together a number of con-structs that are often described as distinctfrom one another (e.g., collective responsi-bility and relational trust); these constructsfactored together into a reliable single mea-sure in these data.

Principal develops goals. On a 4-pointLikert-type scale ranging from strongly dis-agree to strongly agree, respondents were alsoasked about how well the principal com-municates and develops clear instructionalgoals. Sample items included whether theprincipal “clearly communicates expectedstandards for math instruction in this school”and whether the principal “communicates aclear vision for our school.” We note that thisvariable and the professional learning com-munity variable are measures of teachers’perceptions of the school organization, as thesurvey elicited respondents’ feelings abouttheir colleagues and the principal; the aboveon-the-job learning measures asked directlyabout their behaviors regarding discussionand interaction with colleagues around in-struction.

Teacher efficacy. On a 4-point Likertscale ranging from strongly disagree tostrongly agree, respondents were asked torespond to seven statements regardingtheir ability to raise student achievement

and help the school improve. Items wereaveraged to create the teacher efficacy vari-able. By including this measure in our anal-yses, we acknowledge previous work indi-cating that the extent to which teachersimplement changes in their classroom prac-tice is significantly affected by their feelingsof efficacy (e.g., Guskey, 1988; Tschannen-Moran, Hoy, & Hoy, 1998).

Teacher characteristics. In addition toteacher efficacy, individual teacher charac-teristics included as control variables insubsequent analyses include number ofyears as a teacher, gender, race, and theteacher’s class size.

Data Analysis PlanIn order to test whether the data from

2005 and 2007 were structurally different,we created a set of interactions between a2007 dummy variable and all of the explan-atory variables included in the full model.We then conducted an F-test to determinewhether the interaction terms were jointlysignificant when added to the full model.The inclusion of the interaction terms didnot significantly improve the goodness offit for either ELA teaching practice (F(14,1001) � 1.10, p � ns) or math teaching prac-tice (F(14, 995) � 1.11, p � ns). These resultssuggest that the structure of the data from2005 and 2007 are not sufficiently differentto justify using two separate models, as therelationships between the predictors anddependent variables did not change overtime. Therefore, the data were pooled and1,418 observations were included in thesample.

To explore the relationship betweenteacher learning opportunities and teacherchange in classroom practice, five ordinary-least-squares multiple regression modelswere computed for changes in math andELA. For each subject (math and ELA), Mod-els 1 and 3 regressed change in teaching prac-tice on formal and on-the-job learning oppor-tunities, respectively. Models 2 and 4 addedschool fixed effects and individual teacher


characteristics, including age, race, years ofteaching experience, teacher efficacy, andclass size, to examine the effect of these con-trols on the respective relationships betweenformal and on-the-job learning opportunitiesand teacher change. Including school fixedeffects helped address omitted variable biasat the school level by accounting for any ob-served or unobserved school-level factorsthat may affect teacher change, as onlywithin-school variation is used to producemodel estimates. In addition, robust standarderrors were calculated in Models 2 and 4 toadjust for clustering by participant ID to ac-count for the nonindependence caused byteachers who responded in both 2005 and2007. Finally, the full model (Model 5) in-cluded all previous variables and controlsand added organizational conditions to themodel.

Although using hierarchical linear mod-eling was considered, as teachers arenested within schools, nearly all of the vari-ation in the dependent variables waswithin schools. Specifically, 96% of the vari-ation in change in both math and ELAteaching practice was within schools.

Therefore, we concluded that using schoolfixed-effects models would be sufficient foraddressing our research questions and moreparsimonious than computing hierarchicallinear models.

ResultsDescriptive ResultsDescriptive statistics for the measures

included in this study are presented in Ta-ble 1. Sample means and standard devia-tions are listed for the pooled sample—used in subsequent analyses—and separatelyfor 2005 and 2007. Although there are sta-tistically significant differences in the meanlevels of change in ELA instruction, collab-orative discussion, and ELA professionaldevelopment in 2005 and 2007, the previ-ously discussed F-tests indicated that therelationships between these variables andthe dependent variables did not changeover time. The key variables of interest inthis study—change in math and ELAteaching—indicate that, on average, teach-ers implemented moderate changes in theirclassroom practice and that there was sub-

TABLE 1. Pooled Sample Descriptive Statistics

VariablePooled Sample

Mean (SD) Range2005

Mean (SD)2007

Mean (SD) p

Change in math instructiona 3.73 (1.51) 1.00–7.00 3.72 (1.55) 3.73 (1.47) nsChange in ELA instruction 3.97 (1.42) 1.00–7.00 4.06 (1.44) 3.87 (1.39) p � .05Math professional development 2.24 (2.27) .00–8.00 2.18 (2.33) 2.31 (2.21) nsELA professional development 2.89 (2.59) .00–8.00 3.11 (2.81) 2.67 (2.32) p � .01Math undergrad/grad courses 4.21 (3.34) .00–16.00 4.11 (3.37) 4.30 (3.34) nsEnglish undergrad/grad courses 5.54 (4.08) .00–16.00 5.60 (4.17) 5.48 (3.99) nsOutside network participation 2.93 (1.69) 1.00–6.00 2.93 (1.71) 2.91 (1.66) nsCollaborative discussion 3.23 (.82) 1.10–5.33 3.16 (.84) 3.29 (.80) p � .01Peer observation and feedback 2.10 (.98) 1.00–5.00 2.10 (1.02) 2.09 (.93) nsMath advice seeking 4.53 (5.14) .00–34.00 4.60 (5.39) 4.47 (4.88) nsELA advice seeking 4.66 (5.08) .00–34.00 4.59 (5.17) 4.73 (4.98) nsProfessional learning community 3.21 (.56) 1.23–4.23 3.21 (.55) 3.20 (.57) nsPrincipal develops goals 3.36 (.60) 1.00–4.00 2.88 (.66) 2.91 (.66) nsTeacher efficacy 2.95 (.42) 1.14–4.00 2.96 (.42) 2.94 (.41) nsYears experience 13.42 (9.65) 1.00–47.00 13.62 (9.32) 13.21 (10.01) nsClass size 18.42 (4.34) 1.00–45.00 18.39 (4.50) 18.44 (4.18) nsGender 94.1% female 94.6% female 93.6% female nsRace 71.4% white,

25.7% black69.8% white,26.4% black

73.0% white,23.2% black

ns

NOTE.—N � 1,418 (2005: 714; 2007: 704).aFor all variables, standard scores were used in all analyses.


MARCH 2010

stantial teacher-level variation in theamount of change implemented. Generaltrends in these data also indicate that teach-ers had positive attitudes toward theirschools’ organizational conditions, partici-pated in collaborative discussions with col-leagues more than a few times per month,engaged in peer observation and feedbacka few times per year, participated in a net-work with teachers outside of their schooltwo times per year, and attended more thantwo professional development sessions peryear in both math and ELA. Additionally,teachers in this sample were relatively wellexperienced, with a mean of approximately13 years of experience and a median (notlisted in Table 1) of 12 years of experience.

Table 2 presents the zero-order correla-tions of the primary variables included insubsequent analyses. The correlations be-tween the dependent variables—changesin math and ELA teaching practice—andteachers’ formal and on-the-job learningopportunities were low to moderate, rang-ing from .08 to .25. Collaborative discussionhad the strongest positive association withchanges in both ELA and math, at .23 and.25, respectively. Finally, nearly all of thecorrelations between variables were signif-icant at the p � .01 level.

Regression ResultsFormal learning opportunities. Table 3

presents five multiple-regression modelspredicting teacher change in math and ELAteaching practice. The first four modelswere run in order to test the impact ofadding controls for individual characteris-tics and school effects on the relationshipsbetween teacher change and formal and on-the-job learning opportunities. The finalmodel was designed to test the relative im-pacts of all of these learning opportunitieson teacher change in practice. For each sub-ject, Models 1 and 3 are essentially uncon-ditional models, including only the inde-pendent variables of interest—formal andon-the-job learning opportunities, respec-

tively. Model 1 suggests that formal profes-sional development has a fairly small buthighly significant association with changein both math and ELA practice. This asso-ciation held up well in Model 2, whichadded individual teacher characteristicsand school fixed effects, indicating that thisrelationship is not strongly mediated bythese controls. Specifically, a single stan-dard deviation (SD) increase in profes-sional development is associated with ap-proximately a .18 SD and .15 SD increase inthe change in math and ELA teaching prac-tice (respectively). In addition, a single SDincrease in outside network participation isassociated with approximately a .06 SD in-crease in the change in both math and ELApractice. The number of math courses ateacher has taken is not predictive of theirchange in practice for either math or ELA.Model 2 also indicates that teacher efficacyis also a statistically significant predictor ofteacher change in both math and ELAteaching practice. For both subjects, Model2 explained 12% of the variation in teacherchange in practice.

On-the-job learning opportunities. Mod-els 3 and 4 in Table 3 explore the relationshipbetween teachers’ on-the-job learning op-portunities and teacher change in math andELA classroom practice. Model 3 indicatesthat for both subjects, on-the-job learningopportunities are significantly associatedwith teachers’ changes in practice. Control-ling for school effects and teacher character-istics in Model 4, the relationships betweenchange in teaching practice and collabora-tive discussion are statistically significant forboth math and ELA, indicating that this as-sociation is not dependent upon teachers’school or individual characteristics. Collabo-rative discussion is the strongest predictor ofteacher change, as a single SD increase incollaborative discussion is associated with a.19 SD change in math and .17 SD change inELA teaching practice. In addition, adviceseeking is significantly, though relativelyweakly, associated with change in practice,with a single SD increase in advice seeking


TA

BL

E2.

Cor

rela

tion

Tab

le(P

oole

dSa

mpl

e)

Var

iabl

eM

ath

Cha

nge

EL

AC

hang

eM

ath

PDE

LA

PDM

ath

Cou

rses

Eng

lish

Cou

rses

Out

sid

eN

etC

olla

bD

isc

Obs

/Fd

bck

Mat

hO

utD

gE

LA

Out

Dg

PLC

Prin

cipa

lD

evG

oals

Effi

cacy

Exp

Cha

nge

inE

LA

inst

ruct

ion

.734

**M

ath

prof

essi

onal

dev

elop

men

t.2

07**

.128

**E

LA

prof

essi

onal

dev

elop

men

t.0

72*

.193

**.3

47**

Und

ergr

ad/

grad

mat

hco

urse

s.0

74**

.027

.137

**.0

55*

Und

ergr

ad/

grad

Eng

lish

cour

ses

.049

.018

.108

*.1

35*

.650

**O

utsi

de

netw

ork

part

icip

atio

n.1

22**

.126

**.1

48**

.136

**.0

34.0

44C

olla

bora

tive

dis

cuss

ion

.253

**.2

33**

.212

**.1

64**

.120

**.0

78**

.192

**Pe

erob

serv

atio

nan

dfe

edba

ck.1

34**

.114

**.1

43**

.106

**.0

72.0

71*

.173

**.3

02**

Mat

had

vice

seek

ing

.132

**.1

24**

.076

**.0

16.0

86*

.038

.100

**.2

26**

.088

**E

LA

advi

cese

ekin

g.0

76**

.116

**.0

48.0

86**

.056

.027

.078

**.1

97**

.065

*.8

30**

Prof

essi

onal

lear

ning

com

mun

ity

.088

**.0

83**

.163

**.0

20.0

83*

.016

.138

**.2

74**

.142

**.1

73**

.147

**Pr

inci

pal

dev

elop

sgo

als

.120

**.1

18**

.184

**.1

31**

.055

*.0

48.1

36**

.272

**.1

17**

.142

**.1

22**

.562

**T

each

eref

ficac

y.1

73**

.189

**.1

48**

.068

**.0

89*

.063

.156

**.2

12**

.109

**.1

16**

.089

**.4

41**

.428

**Y

ears

expe

rien

ce�

.006

�.0

08.0

64*

.050

.113

**.1

05**

�.0

36.0

53*

.109

**�

.052

*�

.079

**.1

59**

.161

**.1

29**

Cla

sssi

ze.0

52.0

09.1

15**

�.0

30.1

17*

.074

.010

.001

.009

.053

.010

.000

�.0

03�

.027

�.0

20

*Cor

rela

tion

issi

gnifi

cant

atth

e.0

5le

vel

(tw

o-ta

iled

).**

Cor

rela

tion

issi

gnifi

cant

atth

e.0

1le

vel

(tw

o-ta

iled

).

334

TA

BL

E3.

Cha

nge

inT

each

ing

Prac

tice

Reg

ress

ion

Mod

els

Ind

epen

den

tV

aria

ble

Cha

nge

inM

ath

Tea

chin

gPr

acti

ceC

hang

ein

EL

AT

each

ing

Prac

tice

(1)

(2)

(3)

(4)

(5)

(1)

(2)

(3)

(4)

(5)

Form

alPD

Form

alPD

w/

Con

trol

s

On-

the-

Job

Opp

ortu

niti

esO

nly

On-

the-

Job

Opp

ortu

niti

esw

/C

ontr

ols

Full

Mod

elFo

rmal

PDFo

rmal

PDw

/C

ontr

ols

On-

the-

Job

Opp

ortu

niti

esO

nly

On-

the-

Job

Opp

ortu

niti

esw

/C

ontr

ols

Full

Mod

el

Form

alle

arni

ngop

port

unit

ies:

Mat

hpr

ofes

sion

ald

evel

opm

ent

.191

**.1

75**

...

...

.139

**..

...

...

...

...

.(.0

28)

(.030

)(.0

31)

Mat

hco

urse

s.0

42.0

24..

...

..0

05..

...

...

...

...

.(.0

27)

(.029

)(.0

28)

EL

Apr

ofes

sion

ald

evel

opm

ent

...

...

...

...

...

.180

**.1

40**

...

...

.106

**(.0

27)

(.029

)(.0

29)

Eng

lish

cour

ses

...

...

...

...

...

�.0

12�

.021

...

...

�.0

33(.0

27)

(.027

)(.0

27)

Out

sid

ene

twor

kpa

rtic

ipat

ion

.096

*.0

64*

...

...

.030

.106

**.0

69*

...

...

.035

(.027

)(.0

28)

(.028

)(.0

26)

(.028

)(.0

27)

On-

the-

job

lear

ning

oppo

rtun

itie

s:C

olla

bora

tive

dis

cuss

ion

...

...

.217

**.1

89**

.180

**..

...

..2

04**

.175

**.1

62**

(.027

)(.0

31)

(.033

)(.0

28)

(.030

)(.0

31)

Peer

obse

rvat

ion

and

feed

back

...

...

.062

*.0

56.0

46..

...

..0

47.0

48.0

42(.0

27)

(.029

)(.0

30)

(.027

)(.0

30)

(.030

)M

ath

advi

cese

ekin

g..

...

..0

78**

.067

*.0

63*

...

...

...

...

...

(.026

)(.0

28)

(.029

)E

LA

advi

cese

ekin

g..

...

...

...

...

...

...

..0

72**

.062

*.0

59*

(.026

)(.0

28)

(.029

)O

rgan

izat

iona

lco

ndit

ions

:Pr

ofes

sion

alle

arni

ngco

mm

unit

y..

...

...

...

.�

.062

...

...

...

...

.002

(.041

)(.0

32)

Prin

cipa

ld

evel

ops

goal

s..

...

...

...

..0

52..

...

...

...

..0

49(.0

36)

(.036

)T

each

eref

ficac

y..

..1

60**

...

.144

**.1

25**

...

.194

**..

..1

73**

.143

**(.0

31)

(.030

)(.0

33)

(.030

)(.0

29)

(.032

)Sc

hool

fixed

effe

cts

No

Yes

No

Yes

Yes

No

Yes

No

Yes

Yes

Tea

cher

cont

rols

No

Yes

No

Yes

Yes

No

Yes

No

Yes

Yes

R2

.054

.128

.074

.145

.169

.049

.130

.061

.148

.168

Obs

erva

tion

s1,

328

1,30

91,

418

1,37

71,

309

1,33

11,

312

1,41

81,

377

1,31

2

NO

TE.—

Mod

els

1an

d3,

stan

dar

der

rors

inpa

rent

hese

s.M

odel

s2,

4,an

d5,

robu

stst

and

ard

erro

rsin

pare

nthe

ses.

Tea

cher

cont

rols

(not

show

n)in

clud

era

ce,

gend

er,y

ears

expe

rien

ce,a

ndcl

ass

size

.*p

�.0

5.**

p�

.01.

335

corresponding to a .06 SD increase in changefor both subjects. Finally, as in Model 2 forformal learning opportunities, Model 4 sug-gests that teachers who feel more efficaciousabout their work implement greater changesin their math and ELA teaching practice.

Full model. The full model includedteachers’ formal professional development,on-the-job learning opportunities, organi-zational conditions for learning, teachercontrols, and school fixed effects. Simulta-neously examining the relationships be-tween teacher change, opportunities forlearning, and organizational conditions,Model 5 shows that formal professional de-velopment, collaborative discussion, andadvice seeking remain statistically signifi-cant predictors of teacher change in mathand ELA teaching practice. Results arecomparable across subject areas, as a singleSD increase in professional development isassociated with a .14 SD change in mathand .11 SD change in ELA teaching prac-tice, a single SD increase in advice seekingis associated with a .06 SD increase inchange for both subjects, and a single SDincrease in collaborative discussion is asso-ciated with respective increases in changein teaching practice of .18 SD and .16 SD formath and ELA. Furthermore, the relation-ship between teacher efficacy and changepersists after all variables and controlswere added to the full model. Post hoc sig-nificance tests indicated that there wereneither significant differences by curriculardomain nor between the collaborative dis-cussion, formal professional development,and efficacy coefficients. In sum, aftercontrolling for individual teacher charac-teristics and school fixed effects, both for-mal and on-the-job learning opportunitieswere statistically significant predictors ofchanges in teachers’ math and ELA class-room practice, though we acknowledgethat the coefficients are fairly small.

Robustness checks. In order to test thefindings from the full school fixed-effectsmodels, a series of robustness checks wereconducted. Table 4 presents the results

from the first robustness check, whichaimed to verify the relationships betweenformal learning opportunities and teacherchange. To do this, ELA formal profes-sional development was added to themodel predicting changes in math teachingpractice and math professional develop-ment was added to the ELA change model.The math and ELA professional develop-ment measures were moderately correlated(.35, Table 2), but the results presented in

TABLE 4. Robustness Checks for Pooled Sample

Independent VariableMath

ChangeELA

Change

Formal learningopportunities:

Math professionaldevelopment .158** .028

(.032) (.032)Math courses .005 . . .

(.028)ELA professional

development �.051 .096**(.029) (.031)

English courses . . . �.035(.027)

Outside networkparticipation .031 .032

(.028) (.027)On-the-job learning

opportunities:Collaborative discussion .183** .159**

(.033) (.032)Peer observation and

feedback .046 .041(.029) (.030)

ELA advice seeking . . . .060*(.029)

Math advice seeking .062* . . .(.029)

Organizational conditionsfor learning:

Professional learningcommunity �.063 �.001

(.042) (.039)Principal develops goals .057 .048

(.036) (.036)Teacher efficacy .126** .143**

(.033) (.032)R2 .171 .168Observations 1,309 1,312

NOTE.—Robust standard errors in parentheses.Teacher controls (not shown) include race, gender,years experience, and class size.

*p � .05.**p � .01.


MARCH 2010

Table 4 suggest that they measure differentconstructs, as ELA professional develop-ment failed to predict changes in mathteaching and vice-versa. Rather than beingsome general construct captured by bothtypes of formal development, this robust-ness check suggests that teachers receivedspecific content knowledge from theseopportunities because the subject-specifictraining remains predictive of change in theteaching practice even when training in theother subject is entered into the model.

Table 5 presents robustness checks

aimed at addressing possible simultaneityproblems in the above analyses resultingfrom the independent and dependent vari-ables being collected at the same timepoint. For example, it is possible thatchanges in the independent variables re-flect changes in the dependent variables(e.g., Did teachers collaborate more becausethey had to or planned to implementchanges in their practice?). In order to ad-dress this issue, the sample for the robust-ness checks in Table 5 was restricted toapproximately 400 teachers who completed

TABLE 5. Robustness Checks for Teachers in Both Years

Independent Variable

(1) (2) (1) (2)2007 Math

Change2005 Math

Change2007 ELAChange

2005 ELAChange

2005 Measures:Formal learning opportunities:

Math professional development .048 (.056) .172** (.062) . . . . . .Math courses .029 (.067) .076 (.054) . . . . . .ELA professional development . . . . . . �.052 (.076) .053 (.060)English courses . . . . . . .056 (.062) .025 (.058)Outside network participation �.004 (.046) �.062 (.073) .010 (.041) �.040 (.056)

On-the-job learning opportunities:Collaborative discussion .138* (.058) .115� (.065) .183* (.067) .103 (.075)Peer observation and feedback �.032 (.049) .044 (.063) �.038 (.052) .018 (.061)ELA advice seeking . . . . . . .046 (.031) .054 (.057)Math advice seeking .049 (.036) .079 (.054) . . . . . .

Organizational conditions for learning:Professional learning community .089 (.073) .009 (.103) .142* (.055) .000 (.092)Principal develops goals .027 (.074) .052 (.087) .044 (.068) .024 (.083)

Teacher efficacy .077 (.070) .147* (.062) .026 (.070) .214** (.066)2007 Measures:

Formal learning opportunities:Math professional development .095� (.054) . . .Math courses �.088 (.066) . . .ELA professional development . . . �.003 (.072)English courses . . . �.075 (.055)Outside network participation �.005 (.072) �.049 (.075)

On-the-job learning opportunities:Collaborative discussion .055 (.056) .078 (.069)Peer observation and feedback .079 (.055) .022 (.074)ELA advice seeking . . . �.031 (.061)Math advice seeking .008 (.036) . . .

Organizational conditions:Professional learning community �.004 (.072) .022 (.073)Principal develops goals .116 (.064) .093 (.070)

Teacher efficacy �.028 (.048) .032 (.075)R2 .199 .304 .202 .313Observations 412 385 416 394

NOTE.—Robust standard errors in parentheses. Teacher controls (not shown) include race, gender, yearsexperience, and class size.

�p � .10.*p � .05.**p � .01.


the school staff questionnaire in both 2005and 2007.

In Table 5, Model 1 for both math andELA change, lagged predictors from 2005were used to predict 2007 change in teach-ing practice. The relationship between col-laborative discussion and change in prac-tice held up well in both the math and ELAmodels. This suggests that in the cross-sectional model, the collaborative discus-sion variable was not reflecting changes inthe dependent variable—change in class-room practice. While the relationship be-tween formal professional developmentand change in practice did not hold up wellin Model 1, this may not indicate simulta-neity problems, but rather that teachers’professional development is not designedto have lagged effects on change in instruc-tion over time. For example, if a teacherwas trained in 2005, she would likely beexpected to implement changes in herteaching practice in 2005 and not necessar-ily continue to make changes in 2007 basedon training received 2 years earlier.

Model 2 in Table 5 regressed 2005changes in math and ELA teaching practiceon full sets of predictors from both 2005and 2007. Including the 2007 measures ofpredictor variables helps control for omit-ted variables that might influence both the2007 measures and the 2005 predictors andoutcomes. In addition, if the 2007 variablessignificantly predicted changes in 2005teaching practice, this would signal a prob-lem with the data. This model suggests thatthe relationships between formal profes-sional development, collaborative discus-sion, advice seeking, and teacher efficacyidentified in the full model for the pooledsample (Table 3, Model 5 math) are robustfor changes in math teaching practice, asthe 2005 versions of these variables re-mained predictive over and above the fu-ture measures of these variables. However,this robustness check indicates that the re-lationship between organizational condi-tions for learning and teacher change inmath may be questionable, as the 2007 ver-

sion of the principal develops goals vari-able was more predictive of 2005 changesin teaching practice than the 2005 version ofthat variable for both math and ELA. Forchanges in ELA teaching practice, the rela-tionships identified in the full model for thepooled sample (Table 3, Model 5 ELA) alsoheld up, as the 2005 coefficients on profes-sional development, collaborative discus-sion, and teacher efficacy were larger thanthe 2007 coefficients. In addition to theserobustness checks, change models, whichwere designed to assess whether changes inlearning variables predicted changes inteaching change (e.g., If a teacher attendedmore professional development in 2007than in 2005, did her teaching practicechange more in 2007?) were also analyzed.The results from these analyses, however,are not included here, as they neitherstrongly supported nor negated the find-ings from the pooled school fixed-effectsmodels.

DiscussionUsing a sample of elementary school teachersresponsible for both math and ELA instruc-tion, this study examined the empirical linksbetween teachers’ formal professional devel-opment and on-the-job learning opportuni-ties and teacher change. For a large sample ofelementary school teachers, school fixed-effects models provide empirical confirma-tion that self-reports of both formal and on-the-job learning opportunities are statisticallysignificant predictors of teachers’ reportedchanges in math and ELA classroom practice.We believe this study adds an important andpractical contribution to the literature, as ourresults suggest that concurrently exploringdifferent types of teacher learning opportuni-ties is worthwhile.

First, our results show that consideringboth formal professional development andon-the-job learning opportunities and theirrelations with teacher change is meaning-ful. This finding has implications for con-ceptions about what constitutes a produc-


MARCH 2010

tive opportunity for teacher learning andchallenges policymakers’ reliance on for-mal professional development as the pri-mary mechanism for improving the qualityof in-service teachers. In addition, Hill(2004) suggested that a major gap in theliterature on teacher professional develop-ment is that few studies examine the devel-opment opportunities of typical teachers,and we feel that we have helped addressthat gap by exploring the learning oppor-tunities of a sample of self-contained ele-mentary school teachers in one district.

Extant literature has jointly discussedteachers’ formal professional developmentand on-the-job learning opportunities inreviews and theoretical pieces. For exam-ple, Putnam and Borko (2000) posited thata combination of approaches to teacherlearning and change involving formal pro-fessional development and engagement ininstructional discourse with colleagues maybest foster changes in teachers’ practices. Themajor contribution of this study is that it sub-stantiates the notion that both types of oppor-tunities support teacher learning and helpfacilitate teachers’ changes in practice. Al-though this study did not include a follow-upassessing the quality of teachers’ changes inpractice, we believe these changes to be ben-eficial, as other work with these data hasfound teachers’ reported changes in practiceto be linked to improvements in studentachievement (Supovitz et al., in press).

Among the on-the-job learning opportu-nities explored in this study, collaborativediscussion between teachers was the stron-gest predictor of teacher change in math andELA classroom practice. While other studieshave not examined the relative effects of thedifferent on-the-job learning opportunitiesexplored here, the notion that collaborativediscussion between teachers plays an impor-tant role in their development is well sup-ported by previous work suggesting that col-laboration improves teachers’ knowledgebase for teaching (Brownell et al., 1997; God-dard et al., 2007). One additional conceivableexplanation for the link between collabora-

tive discussion and teacher change in practiceis that when teachers choose to engage withcolleagues, they do so around subject matterthat is directly relevant to their current teach-ing practice. As previously discussed, workon formal professional development has sug-gested that activities closely tied to teachers’daily work and focused on specific contentare more likely to facilitate learning andchange (Desimone, Porter, Garet et al., 2002;Garet et al., 2001; Porter et al., 2000; Smylie,1995), so although the present data did notinclude this information, it is plausible thatthe same is true for teachers’ on-the-job col-laborative discussion.

In addition to informing our under-standing of the potential value of on-the-job learning opportunities for teacher learn-ing and change, this study has implicationsregarding spending on teachers’ formalprofessional development. Across the UnitedStates, district estimates of spending onteachers’ formal professional developmentrange from 2% to 5% of the yearly budgetbeing spent on the activities alone, and thisamount nearly doubles when salary in-creases awarded for educational attainmentsare also considered (CPRE, 1996). Financialsupport varies widely at the state level, buteven conservative estimates suggest thatmost states contribute millions of dollars toteacher professional development (CPRE,1996; Ward, St. John, & Laine, 1999). Onefundamental problem is that very little isknown about the tangible returns to this in-vestment. The evidence we do have suggeststhat the majority of these funds are spent onformal learning opportunities for teachersthat have little impact on their classroompractice (Corcoran, 1995a; Desimone, Porter,Garet et al., 2002; Garet et al., 2001). The re-sults from this study provide evidence thatteachers do engage in valuable formal learn-ing opportunities, as content-specific profes-sional development is linked to teachers’changes in classroom practice.

The results of this study also suggest,however, that there may be additionalstrategies for augmenting teacher learning


and improving the quality of public schoolteachers than are currently promoted byeducation policy. If the learning opportuni-ties in which teachers engage in theirschools are at least as effective at fosteringchange as costly formal learning opportu-nities, then it may be worthwhile to incor-porate more on-the-job learning opportuni-ties into teachers’ work lives. The notionthat teachers should be provided withmore opportunities to learn from andwith their colleagues through collabora-tion is not a new concept within the fieldof education. In fact, researchers andsome reformers have been advocating set-ting aside time for teacher collaborationfor many years (e.g., Darling-Hammond,1996; Darling-Hammond & McLaughlin,1995; Hiebert & Stigler, 2000; Little, 1982;Rosenholtz, 1989b; Smylie, 1994), andsome local school systems have already in-stituted this practice (Rosenholtz, 1989a;Spillane, Mesler, Croegaert, & Sherer Zolt-ners, 2009). These strategies for teacherlearning and change rarely find their wayinto education policy, especially at the stateand federal levels. There are, however, ahandful of states that have recognized thepotential value of on-the-job learning op-portunities by including them in teacherpolicies. For instance, rather than requiringthat teachers accumulate professional de-velopment hours for recertification solelythrough formal learning opportunities, Vir-ginia, Vermont, and a few other states alsoallow peer observation and participation incollaborative curriculum development tocount toward recertification (CPRE, 1996).

A key issue here, however, concerns howpolicymakers at the federal, state, and districtlevels might work to influence on-the-joblearning opportunities for teachers, andwhich level of the educational system mightbe best positioned to do so. We acknowledgethat it may be easier for policymakers tomandate that teachers spend a specifiednumber of hours in formal learning activitiesthan to require them to engage in on-the-joblearning opportunities. The results of this

work may therefore be especially relevant tothe work of school leaders, who may inten-tionally structure teachers’ work in order toallow them opportunities to engage in dis-cussion and joint work with colleagues (Spil-lane et al., 2009). This is not to say that schoolleaders can easily compel teachers to engagein productive collaboration, as some collabo-ration may increase conflict (Achinstein,2002), but they may use different strategies toshape teachers’ schedules and promote activ-ities that have been shown to cultivate pro-ductive teacher collaboration, such as team-ing and appointing teacher leaders (Drago-Severson, 2007). In addition, while they werenot a substantial factor in teachers’ learningopportunities in the district we studied,2coaches may play an important role in facil-itating teachers’ on-the-job learning opportu-nities (Showers & Joyce, 1996).

In addition to providing empirical evi-dence that both formal and on-the-job learn-ing opportunities are significantly associatedwith teachers’ changes in instructional prac-tice, we found two relatively unexpected re-sults. First, the results suggest teachers’ per-ceptions of their schools’ organizationalconditions have markedly different relation-ships to change than their on-the-job behav-iors. In addition, while we expected that therelationships between teacher learning op-portunities might differ between math andELA, we did not find significant differencesbetween the two subjects.

Previous work on professional commu-nity has often combined measures of teach-ers’ perceptions about their school’s organi-zational climate, including teachers’ feelingsabout trust between teachers and staff mem-bers’ collective responsibility for studentlearning, and measures of teachers’ collegialbehaviors, including the frequency withwhich teachers engage in collaboration andpeer observation (e.g., Bryk et al., 1999; Louis& Marks, 1998; Louis et al., 1996). While priorstudies have found that a meaningful scale isformed when these types of items are com-bined, the related items in this data set didnot form a reliable scale. In separating teach-


MARCH 2010

ers’ perceptions about the organizational cli-mate from their collegial interactions, thisstudy found that, in a sense, actions speaklouder than words. One contribution of thisstudy is the finding that teacher perceptionsand actions may operate in distinct wayswith respect to their relationships to teachers’changes in instructional practice. Specifically,teachers’ on-the-job behavior was consis-tently related to their changes in practice, andthis relationship was not affected by the in-troduction of teachers’ perceptions of organi-zational conditions into the model.

With regard to curricular domain, thisstudy suggests that associations betweenlearning opportunities and changes in prac-tice are comparable in math and ELA, whichis in itself a notable finding given the litera-ture outlining variation in teacher practice bysubject (e.g., Drake et al., 2001; Siskin, 1991;Stodolsky, 1988; Stodolsky & Grossman,1995). Prior work indicated that advice seek-ing may be more influential in teachers’ mathpractice than in other subject areas, as Stodol-sky and Grossman (1995) found that second-ary teachers’ lessons were more coordinatedwith their colleagues in math than in othersubjects. Stodolsky and Grossman’s work,however, focused specifically on high schoolteachers who were specialized in their teach-ing assignments. Our findings suggest thatthis may not be true for self-contained ele-mentary school teachers who are responsiblefor teaching both subjects, but further workwith a larger sample might better detect dif-ferences by curricular domain. It is also plau-sible that the current policy environment,which places a dual emphasis on math andELA, has led to teachers treating the two sub-jects more similarly than they once did.

LimitationsThe data for this study include detailedmeasures of many aspects of teachers’work practice. This study does, however,have limitations. First, the data set includesrich information regarding teachers’ on-the-job learning opportunities, but its mea-

sures of formal learning opportunities areless detailed. While one of the question-naire’s strengths is that it asks about formaldevelopment sessions addressing specificsubject matter, one limitation is that it doesnot differentiate between varying formalprofessional development formats (e.g., re-form vs. traditional) or include detailedinformation on the timing of graduatecoursework. Another potential limitation isthat the data are self-reports of teacherchange and participation in learning oppor-tunities. Mayer (1999) reported that whileteacher self-reports of classroom practicemay be quite reliable, survey instrumentsare unable to precisely measure the qualityof teachers’ practice. Nonetheless, as men-tioned previously, other work with thesedata has found statistically significant rela-tionships between teacher change and stu-dent achievement (Supovitz et al., in press).Finally, the relationships discussed hereinare correlational, and although a number ofrobustness checks have been conducted toassess the strength of the relationships inthe school fixed-effects models, causal ar-guments are beyond the scope of the data.

Conclusion and Future ResearchBy concurrently analyzing the empirical re-lationships between elementary schoolteachers’ formal professional developmentand on-the-job learning opportunities andchange in math and ELA instruction, thisarticle extends the existing research onteacher learning and change. Its findingsindicate that the opportunities in whichteachers engage within their school build-ings are at least as predictive of teacherchange as are the subject-specific formalprofessional development sessions they at-tend. Furthermore, this study suggests thatit may be worthwhile for both school lead-ers and policymakers to pay more attentionthan they traditionally have to on-the-joblearning opportunities by allocating timefor teachers to collaborate or adjusting pol-icy levers regarding teacher recertification.


Future research in this area should ex-plore the relative relationships between dif-ferent types of formal learning opportunities(e.g., workshops vs. ongoing mentoring, con-ferences vs. teacher networks, etc.), on-the-job opportunities, and teacher change. Con-ducting such an analysis may determine thatthe relationship between formal professionaldevelopment and teacher change identifiedby this study underestimated the effects ofcertain activities, but overestimated the ef-fects of others. Additional next steps includeexploring the links between teacher learningopportunities, teacher change, and studentachievement. Gaining a better understandingof the relative impacts of these opportuni-ties on raising student achievement—theprimary goal of many education policies—will provide useful information to teachers,school leaders, and policymakers. Finally,because this study focused only on elemen-tary school teachers, middle and highschool teachers should be included in sub-sequent analyses, as the relationships be-tween the primary variables of interest maydiffer significantly by school level. In addi-tion to the knowledge contributed to thefield by the present study, this future re-search will help answer important ques-tions regarding the most effective strategiesfor improving the knowledge and skills ofin-service teachers so that they are betterequipped to improve student outcomes.

Appendix

School Staff Questionnaire ItemsMath Change (� � .95)Please indicate how much you changed the fol-lowing aspects of your math teaching this year:

Student assessmentStudent groupingMaterials usedThe topics coveredThe teaching methods you useThe kinds of work you have students doThe kinds of questions you ask studentsYour understanding of the needs of in-

dividual students in your class

ELA Change (� � .93)Please indicate how much you changed the fol-lowing aspects of your reading/language arts orEnglish teaching this year:

Student assessmentStudent groupingMaterials usedThe topics coveredThe teaching methods you useThe kinds of work you have students doThe kinds of questions you ask studentsYour understanding of the needs of in-

dividual students in your class

Collaborative Discussion (� � .90)This school year, how often have you had con-versations with colleagues about the followingtopics:

What helps students learn the bestDevelopment of new curriculumThe goals of this schoolManaging classroom behaviorYour reading/language arts of English

instructionYour math instructionContent or performance standards in

reading/language arts or EnglishContent or performance standards in

math

This school year, how often did you have sched-uled meetings with other teachers in this schoolto discuss and plan curriculum or teaching ap-proaches?

This school year, how often did you have in-depth discussions about your teaching with thefollowing people:

Another classroom teacher

Peer Observation and Feedback (� � .84)This school year, how often did you observe anyof the following people teach?


This school year, how often did the followingpeople observe you teach?


This school year, how often did the followingpeople give you feedback after observing youteach?


MARCH 2010


This school year, how often did the followingpeople review your students’ work?


Professional Learning Community (� � .94)Indicate the extent to which you agree or dis-agree with each of the following statementsabout the school in which you work:

Teachers in this school respect colleagueswho are expert in their craft

Teachers in this school trust each otherTeachers in this school really care about

each otherTeachers respect other teachers who take

the lead in school improvement ef-forts

Many teachers openly express their pro-fessional views at faculty meetings

Teachers in this school are willing toquestion one another’s views on is-sues of teaching and learning

We do a good job of talking throughviews, opinions, and values

Teachers are expected to continuallylearn and seek out new ideas in thisschool

Teachers are encouraged to experimentin their classrooms in this school

Teachers are encouraged to take risks inorder to improve their teaching

How many teachers in this school do the follow-ing:

Take responsibility for helping one an-other do well

Help maintain positive student behaviorin the entire school

Take responsibility for improving theoverall quality of teaching in the school

Principal Develops Goals (� � .93)Mark the extent to which you disagree or agreewith each following statements about the prin-cipal in this school:

Clearly communicates expected stan-dards for reading/language arts orEnglish instruction in this school

Clearly communicates expected stan-dards for math instruction in thisschool

Encourages teachers to raise test scores

Makes clear to the staff his or her expec-tations for meeting instructional goals

Communicates a clear vision for ourschool

Communicates clear standards for stu-dent learning

Teacher Efficacy (� � .70)

I am capable of making the kinds ofchanges expected in this school

The kinds of changes expected in thisschool are helping my students reachhigher levels of achievement

I strongly value the kinds of changesexpected in this school

If I try really hard, I can get through toeven the most difficult and unmoti-vated students

I am uncertain how to teach some of mystudents [Scale reversed]

My students’ peers influence their moti-vation more than I do [Scale reversed]

Most of a student’s performance de-pends on the home environment, so Ihave limited influence [Scale re-versed]

Notes

Work on this article was supported by theDistributed Leadership Studies with fundingfrom the Institute for Education Sciences (grantR305E040085) and the National Science Founda-tion (grant 0412510). We are grateful to our col-leagues on the study for their help with datacollection and data analysis: Carol Barnes, EricM. Camburn, Lisa Dorner, Ellen Goldring, Jona-thon Supovitz, Jason Huff, Henry May, BethSanders, James Sebastian, James Pustejovsky,and Amber Stitziel Pareja. We are grateful toGreg Duncan for his comments on an earlierdraft of this article. All opinions and conclusionsor recommendations expressed in this paper arethose of the authors and do not necessarily re-flect the views of any of the funding agencies.Address all correspondence to Leigh MeslerParise at [email protected].

1. For example, if a respondent sought mathteaching advice from two colleagues daily(weight of 5), one colleague monthly (weight of3), and one colleague yearly (weight of 1), therespondent’s weighted math advice seekingmeasure would be 14 (5 � 5 � 3 � 1).

2. Half of the teachers in the sample did notinteract with a coach during the school year, andthose teachers who did work with a coach did so


infrequently (only 20% interacted with a coachmore than a few times per year).

References

Achinstein, B. (2002). Conflict amid community:The micropolitics of teacher collaboration.Teachers College Record, 104(3), 421–455.

Birman, B. F., Desimone, L., Porter, A. C., &Garet, M. S. (2000). Designing professionaldevelopment that works. Educational Leader-ship, 57(8), 28–33.

Borko, H. (2004). Professional development andteacher learning: Mapping the terrain. Edu-cational Researcher, 33(8), 3–15.

Brownell, M. T., Yeagar, E., Rennels, M. S., &Riley, T. (1997). Teachers working together:What learning educators and researchersshould know. Teacher Education and SpecialEducation, 20, 340–359.

Bryk, A. S., Camburn, E., & Louis, K. S. (1999).Professional community in Chicago elemen-tary schools: Facilitating factors and organi-zational consequences. Educational Adminis-tration Quarterly, 35(5), 751–781.

Bryk, A. S., & Schneider, B. L. (2002). Trust inschools: A core resource for improvement. NewYork: Russell Sage.

Carpenter, T. P., Fennema, E., Peterson, P. L.,Chiang, C. P., & Loef, M. (1989). Usingknowledge of children’s mathematics think-ing in classroom teaching: An experimentalstudy. American Educational Research Journal,26(4), 499–531.

Coburn, C. E. (2001). Collective sensemakingabout reading: How teachers mediate read-ing policy in their professional communities.Educational Evaluation and Policy Analysis,23(2), 145–170.

Cohen, D. K., & Barnes, C. A. (1993). Pedagogyand policy. In D. K. Cohen, M. W. McLaugh-lin, & J. E. Talbert (Eds.), Teaching for under-standing: Challenges for policy and practice (pp.207–239). San Francisco: Jossey-Bass.

Cohen, D. K., & Hill, H. C. (2002). Learning pol-icy: When state education reform works. NewHaven, CT: Yale University Press.

Corcoran, T. C. (1995a). Helping teachers teachwell: Transforming professional development.Philadelphia: University of Pennsylvania.

Corcoran, T. C. (1995b). Transforming professionaldevelopment for teachers: A guide for state poli-cymakers. Washington, DC: National Gover-nors’ Association.

CPRE (1996). Policies and programs for professionaldevelopment of teachers: Profiles of the states.

Philadelphia: Consortium for Policy Re-search in Education.

Cronbach, L. J. (1951). Coefficient alpha and theinternal structure of tests. Psychometrika,16(3), 297–334.

Darling-Hammond, L. (1996). The quiet revolu-tion: Rethinking teacher development. Edu-cational Leadership, 53(6), 4–10.

Darling-Hammond, L. (2000). Teacher qualityand student achievement: A review of statepolicy evidence. Education Policy Analysis Ar-chives, 8(1), 1–49.

Darling-Hammond, L., & McLaughlin, M. W.(1995). Policies that support professional de-velopment in an era of reform. Phi DeltaKappan, 76(8), 597–604.

Davis, K. S. (2003). “Change is hard”: What sci-ence teachers are telling us about reform andteacher learning of innovative practices. Sci-ence Education, 87(1), 3–30.

Desimone, L., Porter, A. C., Birman, B. F., Garet,M. S., & Yoon, K. S. (2002). How do districtmanagement and implementation strategiesrelate to the quality of the professional de-velopment that districts provide to teachers?Teachers College Record, 104(7), 1265–1312.

Desimone, L., Porter, A. C., Garet, M. S., Yoon,K. S., & Birman, B. F. (2002). Effects of pro-fessional development on teachers’ instruc-tion: Results from a three-year longitudinalstudy. Educational Evaluation and Policy Anal-ysis, 24(2), 81–112.

Drago-Severson, E. (2007). Helping teacherslearn: Principal leadership for adult growthand development. Teachers College Record,109(1), 70–125.

Drake, C., Spillane, J. P., & Hufford-Ackles, K.(2001). Storied identities: Teacher learningand subject-matter context. Journal of Curric-ulum Studies, 33(1), 1–23.

Eraut, M. (2004). Informal learning in the work-place. Studies in Continuing Education, 26(2),247–273.

Eraut, M., & Hirsh, W. (2007). The significance ofworkplace learning for individuals, groups, andorganisations. Oxford: University of Oxford,SKOPE.

Feiman-Nemser, S. (2001). From preparation topractice: Designing a continuum to strengthenand sustain teaching. Teachers College Record,103(6), 1013–1055.

Frank, K. A., Zhao, Y., & Borman, K. (2004).Social capital and diffusion of innovationswithin organizations: The case of computertechnology in schools. Sociology of Education,77, 148–171.

Fullan, M. (2002). The change leader. EducationalLeadership, 59(8), 16–20.


MARCH 2010

Garet, M. S., Cronen, S., Eaton, M., Kurki, A.,Ludwig, M., Jones, W., et al. (2008). The im-pact of two professional development interven-tions on early reading instruction and achieve-ment. Washington, DC: NCEE.

Garet, M. S., Porter, A. C., Desimone, L., Birman,B. F., & Yoon, K. S. (2001). What makes pro-fessional development effective? Resultsfrom a national sample of teachers. AmericanEducational Research Journal, 38(4), 915–945.

Goddard, Y., Goddard, R., & Tschannen-Moran,M. (2007). A theoretical and empirical inves-tigation of teacher collaboration for schoolimprovement and student achievement inpublic elementary schools. Teachers CollegeRecord, 109(4), 877–896.

Goldhaber, D. D., & Brewer, D. J. (1998). Whenshould we reward degrees for teachers? PhiDelta Kappan, 80(2), 134–138.

Goldring, E., Huff, J., Stitziel Pareja, A., & Spil-lane, J. P. (2008, March). Measuring principals’content knowledge of learning-centered leader-ship. Paper presented at the meeting of theAmerican Educational Research Associa-tion, New York.

Goldring, E., Spillane, J. P., Huff, J., Barnes,C. A., & Supovitz, J. (2006, April). Measuringthe instructional leadership competence of schoolprincipals. Paper presented at the meeting ofthe American Educational Research Associ-ation, San Francisco.

Guskey, T. R. (1988). Teacher efficacy, self-concept, and attitudes toward the imple-mentation of instructional innovation. Teach-ing and Teacher Education, 4(1), 63–69.

Guskey, T. R. (2002). Professional developmentand teacher change. Teachers and Teaching,8(3), 381–391.

Hansen, M. T. (1999). The search-transfer prob-lem: The role of weak ties in sharing knowl-edge across organization subunits. Adminis-trative Science Quarterly, 44(1), 82–85.

Hiebert, J., & Stigler, J. W. (2000). A proposal forimproving classroom teaching: Lessonsfrom the TIMSS video study. ElementarySchool Journal, 101(1), 3–20.

Hill, H. C. (2004). Professional developmentstandards and practices in elementaryschool mathematics. Elementary School Jour-nal, 104(3), 215–231.

Hill, H. C. (2007). Learning in the teaching work-force. The Future of Children, 17, 111–127.

Leana, C. R., & Pil, F. K. (2006). Social capitaland organizational performance: Evidencefrom urban public schools. Organization Sci-ence, 17(3), 353–366.

Lee, V. E., & Smith, J. B. (1996). Collective re-sponsibility for learning and its effects on

gains in achievement for early secondaryschool students. American Journal of Educa-tion, 104(2), 103–147.

Leithwood, K. (1992). The move toward trans-formational leadership. Educational Leader-ship, 49(5), 8–12.

Lieberman, A. (1995). Practices that supportteacher development. Phi Delta Kappan,76(8), 591–596.

Little, J. W. (1982). Norms of collegiality andexperimentation: Workplace conditions ofschool success. American Educational ResearchJournal, 19(3), 325–340.

Little, J. W. (1990). The persistence of privacy:Autonomy and initiative in teachers’ profes-sional relations. Teachers College Record,91(4), 509–536.

Little, J. W. (1993). Teachers’ professional devel-opment in a climate of educational reform.Educational Evaluation and Policy Analysis,15(2), 129–151.

Little, J. W. (2002). Locating learning in teachers’communities of practice: Opening up prob-lems of analysis in records of everydaywork. Teaching and Teacher Education, 18(8),917–946.

Little, J. W. (2003). Inside teacher community: Rep-resentations of classroom practice. TeachersCollege Record, 105(6), 913–945.

Louis, K. S., & Marks, H. M. (1998). Does pro-fessional community affect the classroom?Teachers’ work and student experiences inrestructuring schools. American Journal of Ed-ucation, 106(4), 532–575.

Louis, K. S., Marks, H. M., & Kruse, S. (1996).Teachers’ professional community in re-structuring schools. American Educational Re-search Journal, 33(4), 757–798.

Mayer, D. (1999). Measuring instructional prac-tice: Can policymakers trust survey data?Educational Evaluation and Policy Analysis,21(1), 29–45.

McCutchen, D., Abbott, R., Green, L., Beretvas,S., Cox, S., Potter, N., et al. (2002). Beginningliteracy: Links among teacher knowledge,teacher practice, and student learning. Jour-nal of Learning Disabilities, 35(1), 69–86.

McLaughlin, M. W., & Talbert, J. E. (2001). Pro-fessional communities and the work of highschool teaching. Chicago: University of Chi-cago Press.

NCES (2005). Characteristics of public school teach-ers’ professional development activities: 1999–2000. Washington, DC: U.S. Department ofEducation.

Porter, A. C., Garet, M. S., Desimone, L., Yoon,K. S., & Birman, B. F. (2000). Does professionaldevelopment change teaching practice? Results


from a three-year study. Washington, DC:American Institutes for Research and De-partment of Education.

Putnam, R. T., & Borko, H. (2000). What do newviews of knowledge and thinking have tosay about research on teacher learning? Ed-ucational Researcher, 29(1), 4–15.

Reagans, R., & McEvily, B. (2003). Networkstructure and knowledge transfer: The ef-fects of cohesion and range. AdministrativeScience Quarterly, 48(2), 240–267.

Rice, J. K. (2001). Cost framework for teacher prep-aration and professional development. Washing-ton, DC: Finance Project.

Rosenholtz, S. J. (1985). Effective schools: Inter-preting the evidence. American Journal of Ed-ucation, 93(3), 352–388.

Rosenholtz, S. J. (1989a). Teacher’s workplace: Thesocial organization of schools. New York:Longman.

Rosenholtz, S. J. (1989b). Workplace conditionsthat affect teacher quality and commitment:Implications for teacher induction programs.Elementary School Journal, 89(4), 421–439.

Saxe, G. B., Gearhart, M., & Nasir, N. S. (2001).Enhancing students’ understanding of mathe-matics: A study of three contrasting ap-proaches to professional support. Journal ofMathematics Teacher Education, 4, 55–79.

Scribner, J. P., Cockrell, K. S., Cockrell, D. H., &Valentine, J. W. (1999). Creating professionalcommunities in schools through organiza-tional learning: An evaluation of a schoolimprovement process. Educational Adminis-tration Quarterly, 35(1), 130–160.

Scribner, J. P., Hager, D. R., & Warne, T. R.(2002). The paradox of professional commu-nity: Tales from two high schools. Educa-tional Administration Quarterly, 38(1), 45–76.

Showers, B., & Joyce, B. (1996). The evolution ofpeer coaching. Educational Leadership, 53, 12–16.

Siskin, L. S. (1991). Departments as differentworlds: Subject subcultures in secondaryschools. Educational Administration Quarterly,27(2), 134–169.

Smylie, M. A. (1988). The enhancement functionof staff development: Organizational andpsychological antecedents to individualteacher change. American Educational Re-search Journal, 25(1), 1–30.

Smylie, M. A. (1994). Redesigning teachers’work: Connections to the classroom. Reviewof Research in Education, 20(2), 129–177.

Smylie, M. A. (1995). Teacher learning in theworkplace: Implications for school reform.

In T. R. Guskey & M. Huberman (Eds.), Pro-fessional development in education: New para-digms and practices. New York: Teachers Col-lege Press.

Spillane, J. P. (1999). External reform initiativesand teachers’ efforts to reconstruct theirpractice: The mediating role of teachers’zones of enactment. Journal of CurriculumStudies, 31(2), 143–175.

Spillane, J. P., Mesler, L., Croegaert, C., & ShererZoltners, J. (2009). Organizational routines andschool-level efforts to establish tigh coulping:Changing policy, changing work practice?Manuscript submitted for publication.

Stodolsky, S. S. (1988). The subject matters: Class-room activity in math and social studies. Chi-cago: University of Chicago Press.

Stodolsky, S. S., & Grossman, P. L. (1995). Theimpact of subject matter on curricular activ-ity: An analysis of five academic subjects.American Educational Research Journal, 32(2),227–249.

Supovitz, J., Sirinides, P., & May, H. (in press).How principals and peers influence teach-ing and learning. Educational AdministrationQuarterly.

Tschannen-Moran, M. (2001). Collaboration andthe need for trust. Journal of Educational Ad-ministration, 39(4), 308–331.

Tschannen-Moran, M., Hoy, A. W., & Hoy,W. K. (1998). Teacher efficacy: Its meaningand measure. Review of Educational Research,68(2), 202–248.

Uzzi, B. (1997). Social structure and competitionin interfirm networks: The paradox of em-beddedness. Administrative Science Quarterly,42(1), 35–67.

Ward, J. G., St. John, E. P., & Laine, S. W. M.(1999). State programs for funding teacher pro-fessional development. Oak Brook, IL: NorthCentral Regional Educational Lab.

Wenger, E. (1998). Communities of practice: Learn-ing as a social system. Retrieved November17, 2007, from www.infed.org/biblio/com-munities_of_practice.htm

Wilson, S. M., & Berne, J. (1999). Teacher learningand the acquisition of professional knowledge:An examination of research on contemporaryprofessional development. Review of Researchin Education, 24, 173–209.

Youngs, P., & King, M. B. (2002). Principal lead-ership for professional development to buildschool capacity. Educational AdministrationQuarterly, 38(5), 643–670.


MARCH 2010

teacher learning and abstract instructional change… · teacher learning and instructional change:...

Documents