Teachers Supporting Teachers in Urban Schools: What Iterative Research Designs Can Teach Us

Teachers Supporting Teachers in Urban Schools: WhatIterative Research Designs Can Teach Us

Elisa S. Shernoff, Ane M. Marínez-Lora, Stacy L. Frazier, Lara J. Jakobsons,and Marc S. Atkins

University of Illinois at Chicago

Deborah BonnerChicago Public Schools

Abstract. Despite alarming rates and negative consequences associated with urbanteacher attrition, mentoring programs often fail to target the strongest predictorsof attrition: effectiveness around classroom management and engaging learners;and connectedness to colleagues. Using a mixed-method iterative developmentframework, we highlight the process of developing and evaluating the feasibilityof a multi-component professional development model for urban early careerteachers. The model includes linking novices with peer-nominated key opinionleader teachers and an external coach who work together to (1) provide intensivesupport in evidence-based practices for classroom management and engaginglearners, and (2) connect new teachers with their larger network of colleagues.Fidelity measures and focus group data illustrated varying attendance ratesthroughout the school year and that although seminars and professional learningcommunities were delivered as intended, adaptations to enhance the relevance,authenticity, level, and type of instrumental support were needed. Implications forscience and practice are discussed.

Chronic turnover among new teachers isparticularly alarming, with up to 23% of pub-lic school teachers leaving within their firstfive years of teaching—14% migrating toother schools and 9% leaving the professionaltogether (Keigher, 2010). This “revolving

door” (Ingersoll, 2001) is exacerbated by newteachers being placed in the hardest-to-staffschools situated in high-poverty, urban com-munities where one-third to one-half of teach-ers leave within their first five years (Barnes,Crowe, & Schaefer, 2007; Borman & Dowl-

Stacy L. Frazier is now at the Center for Children and Families, College of Arts and Sciences, FloridaInternational University. The research reported here was supported by the Institute of Education Sciences,U.S. Department of Education, through Grant R305A090085 to University of Illinois at Chicago. Theopinions expressed are those of the authors and do not represent views of the Institute or the U.S.Department of Education. Portions of this paper were presented at the Society for Research on EducationalEffectiveness 2011 Conference.

Correspondence regarding this article should be addressed to Elisa S. Shernoff, Institute for JuvenileResearch, Department of Psychiatry, University of Illinois at Chicago, Chicago, IL 60608; e-mail:[email protected]

Copyright 2011 by the National Association of School Psychologists, ISSN 0279-6015

School Psychology Review,2011, Volume 40, No. 4, pp. 465–485

465

ing, 2008). Stressors related to teaching inurban schools are well documented, includingovercrowding, large class sizes, deterioratingconditions, and large numbers of students withunmet learning and mental health needs (Boyd& Shouse, 1997; Cappella, Frazier, Atkins,Schoenwald, & Glisson, 2008; Kataoka,Zhang, & Wells, 2002; Shernoff, Mehta, At-kins, Torf, & Spencer, 2011).

Chronic attrition creates financial hard-ships for urban districts, where scarce re-sources must be diverted to recruiting, hiring,and training (Barnes et al., 2007). Costs asso-ciated with teacher turnover in Chicago PublicSchools, for example, are estimated be-tween 76 and 128 million annually (Barnes etal., 2007). Whether turnover results in migra-tion or attrition, chronic discontinuity in staff-ing at the school level can create organiza-tional instability, in that teachers who remainbegin to view their workplace as undesirableand can experience weak organizational com-mitment (Kardos, Johnson, Peske, & Kauff-man, 2001). Regrettably, chronic turnoverfalls hardest on students, who are dependenton high-quality educators yet commonlytaught by inexperienced teachers over-whelmed by the stressors of urban schools andlikely to leave within those first few years(Smith & Ingersoll, 2004). Value-added stu-dent achievement data, for example, docu-ments that achievement gains are influencedmore by assigned teacher and teacher effec-tiveness than class size or composition andthat teachers with the strongest academicbackgrounds, a characteristic most often asso-ciated with student achievement, are the mostlikely to leave (Borman & Maritza Dowling,2008; Rivkin, Hanushek, & Kain, 2005).

Empirical Predictors of TeacherTurnover

Despite the alarming rates and negativeconsequences associated with teacher attrition,current induction programs and professionaldevelopment for new teachers often lack con-centrated, sustained support targeting the mostrobust empirical predictors of attrition. Givenmounting evidence suggests that students ex-

perience schools through teachers via the con-nection between effective instruction and ac-ademic success, targeted support for teachersworking in high poverty, urban schools is crit-ical to mental health promotion for at-riskchildren (see Atkins, Hoagwood, Kutash, &Seidman, 2010; Crone & Teddlie, 1995; Sher-noff et al., 2011). In this article, we highlightthe development and refinement of a multi-component service model in which early ca-reer teachers are linked with peer-nominatedmentors [key opinion leader (KOL) mentors]and a coach who provide intensive classroomsupport in evidence-based classroom manage-ment and motivation practices to promote stu-dents’ behavior and academic learning as wellas with the larger network of teachers to mod-erate the sense of uncertainty and turbulenceassociated with being a new teacher in anurban school.

Effectiveness in Classroom Management

Teachers’ skills in classroom manage-ment (e.g., order and discipline, environmentand climate, transitions) are related to studentachievement through their impact on effectiveinstruction and opportunities to learn (Creem-ers, 1994; Wang, Haertel, & Walber, 1997).However, beginning teachers consistentlyrank classroom management and student mis-behavior as the most stressful, complex, andpressing issue they face and a top reason forleaving their jobs (Ingersoll, 2001). Severalstudies have documented, for example, thatnovices attend more to disruptive behavior,rely more on punitive discipline strategies, andstruggle to monitor simultaneous classroomevents when compared to more experiencedteachers (Sabers, Cushing, & Berliner, 1991;Yoon, 2002). Building new teachers’ skillsand confidence in classroom management isparticularly crucial for urban educators, whereprevalence rates for disruptive behaviors arealmost three times national estimates (Tolan &Henry, 1996).

Effectiveness in Motivating Students

Perceptions that students are unmoti-vated and disengaged is also a key predictor of

School Psychology Review, 2011, Volume 40, No. 4

466

attrition among new teachers (Ingersoll,2001). Unmotivated students who exert littleeffort and give up quickly when challengedare a major concern in urban, low-incomeschools (Jones & Sandidge, 1997) and presentsignificant challenges to novices (Good &Brophy, 2003). Evidence suggests that whennew teachers experience early setbacks, theyinvest less effort in teaching, are more criticalof students, and resist trying new strategies(Allinder, 1994; Stein & Wang, 1988). Studiesalso document that teachers’ self-efficacy be-liefs related to student engagement and learn-ing are related to student achievement andmotivation (Stipek, 1996), teacher commitment(Coladarci, 1992), and longer term retention(Burley, Hall, Villeme, & Brockmeier, 1991).

Connectedness to Colleagues

Feeling ineffective in the classroom isexacerbated by the alienation and isolationthat often characterizes teaching. Althoughnew teachers are arguably the most at risk forexperiencing social isolation and the most de-pendent on meaningful collaboration withtheir colleagues, the cellular structure ofschools often leads to teaching in isolation andstruggling privately with problems (Johnson &The Project on the Next Generation of Teach-ers, 2004; Lortie, 2002). Studies of beginningteachers highlight that many feel “lost at sea”with few resources to help them survive(Berry, 2004; Kauffman, Johnson, Kardos,Liu, & Peske, 2002). New teachers often de-scribe their work as solitary, with few oppor-tunities to reflect on instruction, coteach, orplan lessons with colleagues (Johnson &Birkeland, 2003; Kardos & Johnson, 2007).The Project on the Next Generation of Teach-ers (2004), dedicated to developing new meth-ods for supporting and retaining high-qualityeducators for the past two decades, consis-tently identifies new teachers’ perceptions re-garding collegial connections, faculty cohe-siveness, and a supportive school communityas critical to deciding to remain in teaching(Kardos et al., 2001; Kardos & Johnson,2007). A longitudinal study of 50 first- andsecond-year teachers in the state of Massachu-

setts working in a wide variety of elementary,middle, and high schools documented thatnew teachers were more likely to remain inteaching and at their current school when theyexperienced frequent interactions among fac-ulty and perceived shared responsibility forstudent learning and school improvement(Kardos et al., 2001). In the current study, wepropose that collegial connections are a con-duit to building skills that lead to effectivenessand foster a sense of belongingness needed topromote long-term commitment to teaching.

Building on Indigenous Resources viathe KOL Mentor

In large urban districts, new teachersoften participate in induction programs fo-cused on logistics and designed to deal withlarge numbers of recruits at once. Reviews ofthe literature highlight that when mentoring ispart of induction, theoretical or empirical jus-tification for selecting and matching mentorswith mentees is lacking (e.g., Ingersoll &Kralik, 2004). Instead, mentors are selectedbased on seniority to fulfill policy require-ments, with limited regard for the technicalskills necessary to disseminate informationand support novices. One alternative involvesselecting mentors who are KOLs, designatedby peers as influential and socially connected(Atkins, Frazier et al., 2008; Rogers, 2005).KOL models extend current mentoring prac-tices by activating peer-nominated leader teach-ers or “near peers” to disseminate evidence-based practices to colleagues (Atkins, Graczyk,Frazier, & Adil, 2003; Atkins, Frazier et al.,2008; Neal et al., 2008; Neal, Neal, Atkins,Henry, & Frazier, 2011). KOLs are well posi-tioned to impact classroom practices and facili-tate social connections for new teachers, and inthe resource-limited environment of urbanschools, capitalize on indigenous resources tobuild school capacity and support the ongoingdissemination of effective practices (Atkins etal., 2003, 2010).

Classroom-Based Support via the Coach

Although KOL mentors are well-posi-tioned to support new teachers because of their

Teachers Supporting Teachers in Urban Schools

467

influential status and social connections, sev-eral practice constraints (e.g., extensive dutiesgiven their influential role; typical teachingassignments; district policies prohibiting dis-trict employees from participating in grant-related activities during contract hours) neces-sitated embedding external coaches hired withgrant funds into the original service model.The coaching model was developed based ontheoretical and empirical support for ongoing,classroom-based, modeling, practice, andguided feedback to promote the generalizationof knowledge and skills gained from profes-sional development into the classroom (An-drzejewski, Kirby, Morral, & Iguchi, 2001;Joyce & Showers, 2002; Poglinco et al.,2003). A core feature of coaching was ensur-ing adequate opportunities for intensive, sus-tained support to enhance new teachers’ effec-tiveness (Poglinco et al., 2003).

Current Study

Teachers Supporting Teachers in UrbanSchools is a 3-year study funded by the Insti-

tute of Education Sciences. The goals of thestudy include developing, adapting, and test-ing the feasibility and impact of a servicemodel for early career teachers designed toincrease their effectiveness in classroom man-agement and motivating learners and connect-edness to colleagues given the central role thatcohesion, trust, and belongingness plays inteaching (Bryk & Schneider, 2002). Figure 1(left side) illustrates the hypothesized relation-ships between the core service componentsand proximal teacher outcomes of interest dur-ing this initial development phase, while theright side depicts hypothesized distal studentand teacher outcomes relevant to a rigoroustest of the model in future studies. A funda-mental assumption of the model is that socialrelationships are crucial for novices not onlybecause they are a conduit for building skillsthat lead to effectiveness, but because connec-tions with colleagues fosters a sense of be-longingness within the workplace necessary toengender longer-term commitment to teach-ing. Social cognitive theory guides the model

Figure 1. Conceptual model representing the hypothesized relationship be-tween the three service components (i.e., Group Seminars, Coaching, andPLCs) and proximal teacher outcomes related to effectiveness and connect-edness and distal student learning and behavior outcomes in addition to distalteacher outcomes related to retention.


468

(Bandura, 1997), which proposes that learningoccurs vicariously by observing others whoserve as models and enactively by learningfrom doing. In this study, we argue that en-hancing new teachers’ perceptions regardingtheir ability to be effective with students andto develop positive relationships with col-leagues is crucial to retention. Otherwise, newteachers working in highly distressed urbanschools are left with few incentives to workthrough the inevitable challenges that arise.The service model is theorized to directly im-pact new teachers’ effectiveness and connect-edness via three core components: group sem-inars, coaching, and professional learningcommunities (PLCs).

Group Seminars

Group seminars were designed to occurtwice per month for early career teachers, withthe goal of disseminating evidence-basedclassroom management and student motiva-tion strategies (a complete list of evidence-based strategies used during Year 1 is avail-able from the first author). KOL mentors weretheoretically in the best position within theschool to lead the group seminars and influ-ence new teachers’ practice, and the coachwas invited to attend group seminars to pro-vide additional support and to plan for more indepth coverage of strategies in the context ofcoaching.

Coaching

Regular classroom-based coaching forearly career teachers by veteran teachers hiredwith grant funds was designed to provide ad-ditional opportunities for ongoing, real-timemodeling, demonstration, and coteaching ofthe evidence-based strategies introduced dur-ing the group seminars and tailoring the cur-riculum to new teachers’ individual needs.Classroom visits by the coach were central tocoaching activities, supported by pre- andpostconferences for planning and reflecting onthose classroom visits (Andrzejewski et al.,2001; Joyce & Showers, 2002; Poglinco et al.,2003).

PLCs

PLC meetings were scheduled monthlyfor all faculty—not just new teachers—andembedded reflective dialogue, staff collabora-tion, and collective responsibility for schoolimprovement into existing site-based profes-sional development (e.g., Bryk, Camburn, &Seashore Louis, 1999). PLCs were led byKOL mentors and designed as a deliberateinduction mechanism targeting new teachers’effectiveness through focused discussions andexposure to evidence-based strategies relatedto classroom management and engaging learn-ers while simultaneously promoting connect-edness with colleagues. Collective responsi-bility for school improvement was infusedthrough participants’ developing PLC groundrules, articulating school-wide norms for be-havior and learning, and prioritizing PLC top-ics. An overarching PLC theme was that if onecolleague was struggling, teachers were col-lectively responsible for supporting that teach-er—in the same spirit that student learningwas the responsibility of all school personnel.

Complimentary Role of KOL Mentorsand Coach

Effectiveness and connectedness servedas the foci of concentrated efforts to supportnew teachers working in urban schools and theKOL mentors played a critical role in bridgingthe core service components and addressingpersistent concerns about how to sustain pro-grams over time ( Annie E. Casey Foundation,2002). The role of the KOL mentor and coachrelated directly to our theory of change viamentors’ central role in promoting interactionsbetween early career teachers and their col-leagues, and through their leadership of thePLC meetings and group seminars designed tofoster a sense of belongingness in the work-place while coaches promoted new teachers’effectiveness through real-time, classroom-based support. Mentors and coaches alsohelped promote new teachers’ effectivenessaround the predictors of attrition through vi-carious and enactive learning opportunities.


469

Year 1 Activities

Year 1 activities provided an opportu-nity to test the initial feasibility of the modelwhile also enhancing its contextual fit andensuring the model reflected the needs, re-sources, and capacity of our school partners(Hoagwood, Burns, & Weisz, 2002). Contex-tual fit was operationalized as the alignment ofthe model and practices subsumed within themodel with the values, skills, and capacity ofteacher participants (Albin, Lucyshyn, Horner,& Flannery, 1996; Benazzi, Horner, & Good,2006). Using a mixed-method iterative devel-opment framework (see Protheroe, Bower, &Chew-Graham, 2007; Rounsaville, Carroll, &Onken, 2001), the current article describesYear 1 development, piloting, and adaptationof the model in partnership with one urbanschool, focusing on examining (1) the extentto which the core components were delivered,(2) teacher satisfaction with the model, and (3)key adaptations to enhance the contextual fitof the model.

Iterative Program Development

Figure 2 highlights the iterative programdevelopment process followed during Year 1,which was comprised of (1) engagement, (2)curriculum development, (3) training, and (4)initial feedback and modifications. These ini-tial phases proceeded in a fluid, ongoing, anditerative manner. That is, the engagement ofour community partners, development of thecurriculum, training of service providers, and

informal feedback occurred simultaneouslyand continuously; we cycled through this pro-cess before piloting the curriculum and col-lecting formal data to evaluate fidelity andteacher satisfaction.

Engagement

Although the relationship between theresearch team and our school partnersthrough a prior grant likely facilitated ourinitial entry into the school, those relation-ships grew and evolved over time, both fos-tering and influencing subsequent phases ofthe study. In addition, as described earlier, akey goal of the model was enhancing andfacilitating connectedness within the school—among classroom teachers, mentors, early ca-reer teachers, and coaches. Diffusion of inno-vations theory suggests that through relation-ships, teachers are encouraged to tryinnovative practices; thus, support and guid-ance around implementation was expected tostrengthen teacher relationships and lead togreater motivation to experiment with newstrategies (Rogers, 2005).

Engagement with our school partnersbegan with meetings twice per month with theprincipal, assistant principal, mentors, andguidance counselor during Spring 2009, inpreparation for piloting the model the follow-ing fall. Because our investigative team hadpartnered with the school in a previous study,we were able to capitalize on that relationship,which likely accelerated the engagement

Figure 2. Year 1 iterative model of program development and adaptation


470

phase for the current work. Early meetingsinvolved highlighting strengths, needs, and ex-isting school and district resources targetingclassroom management and motivating learn-ers; identifying existing school-wide normsaround instruction and behavior; and coalesc-ing around school goals and priorities for pro-fessional development activities. In addition,our school partners provided extensive feed-back on multiple drafts of the curriculum,problem-solved anticipated barriers to teacherparticipation, and established procedures forthe leadership team to genuinely endorse thecurriculum.

Curriculum Development

Three curricula were developed, reflect-ing the most current literature on classroommanagement and motivating learners, and cor-responding to the three core service compo-nents (i.e., group seminars, coaching, andPLCs). All three curricula were developedwith ongoing informal feedback by the schoolleadership team related to school norms andgoals around instruction. Group seminars fo-cused on identifying and altering antecedentsto disruptive behaviors via clear rules, rou-tines, and expectations, and applying consis-tent consequences (Evertson & Weinstein,2006). Coaching focused on providing class-room-based modeling and feedback to teach-ers and embedded reflective practice towardthe goal of enhancing new teachers’ ability toanalyze and identify areas of strength and needto enhance effectiveness. The PLC curriculumwas modeled on the Classroom AssessmentScoring System (CLASS; Pianta, La Paro, &Hamre, 2007), a reliable and standardizedmeasure of classroom processes associatedwith social and academic development that isalso used extensively as a professional devel-opment tool (Pianta, Mashburn, Downer,Hamre, & Justice, 2008). The PLC curriculumwas developed in close collaboration withCLASS developers and the CLASS was usedas one lens for enhancing teachers’ knowledgeand skills related to classroom managementand student engagement.

KOL Mentor and Coach Training

Initial training included a combinationof training and informal feedback. Two inten-sive days of training in Fall 2009 with the leadand fourth author were designed to introducementors and coaches to the model and curric-ulum and to solicit their feedback regardingneeds and priorities for teachers. Day 1 fo-cused on the process and experience of facil-itating group seminars and coaching; Day 2focused on the evidence-based strategies de-signed to enhance early career teachers’ effec-tiveness in classroom management and engag-ing learners. In January 2010, mentors partic-ipated in two additional days of training: thefirst focused on the process of facilitating PLCmeetings and the second focused on using theCLASS as a professional development tool(Pianta et al., 2008). Mentors and coachescontributed significantly to the development ofthe curriculum, thereby enabling us to capital-ize on their expertise and increase the rele-vance of services for teachers. Mentors werecompensated $1000 and earned credit towardstate recertification for training and earnedstipends for facilitating group seminars andPLCs. The lead and fourth author also facili-tated weekly supervision with coaches andmonthly supervision with mentors. Supervi-sion was highly structured, agenda-driven, andfocused on reviewing teachers’ progress inmeeting professional development goals,problem-solving barriers to progress, andplanning for next steps.

Initial Feedback and Modifications toEnhance Contextual Fit

Being responsive to and overcomingearly barriers were crucial to the ongoing col-laboration and iterative development of thecurriculum. Initial feedback from our schoolpartners through weekly meetings, reviewingthe curriculum, and trainings with coaches andKOLs suggested several adaptations to en-hance the contextual fit of the model. The firstadaptation included broadening the role of thecoach to support more teachers in the schoolwho were not early career teachers by ourdefinition but who nevertheless were strug-


471

gling with classroom management and engag-ing learners. Towards this effort, the principalcreated opportunities for the coach to attendgrade level meetings as a mechanism for pro-viding broader school support. Second, to en-hance teacher motivation, we incorporatedmotivational interviewing strategies (Miller &Rollnick, 2002) into the curriculum and initi-ated professional development activities by in-viting teachers to prioritize topics and selectfrom a menu of strategies. The principal like-wise suggested that we would enhance PLCparticipation by creating a “party-like” atmo-sphere, with raffles and door prizes. Third,collaboration with families was identified as ahigh priority for new teachers, and thus weadded universal (Good News Notes; Ruben-stein, Patrikakou, Weissberg, & Armstrong,2000) and targeted (Daily Report Card; Kel-ley, 1990) family involvement strategies intothe curriculum. Finally, feedback from thementors and coach suggested that classroommanagement and student engagement was in-extricably linked to content and instruction.Thus, we integrated lesson plans and statelearning standards into coaching conferencesand group seminars.

Method

Setting and Sample

School. One kindergarten througheighth (K-8) grade elementary school locatedon the west side of Chicago participated. Uni-versity and district internal review board ap-proval were obtained prior to initiating thestudy. The school was identified by screeningK-8 grade schools within this urban district(n � 325) based on the following criteria: 85%or greater low income, average reading scoreson statewide testing below the 30th percentile(M � 28, SD � 3.8), and school populationwithin one standard deviation of the districtmean (M � 702, SD � 306). The school wascharacterized as 98% African American and99% free or reduced-price lunch status andwas one of three schools who had partneredwith our investigative team in a prior mentalhealth services reform project in the compar-

ison condition (Atkins et al., 2008; Cappella etal., 2008).

Mentors. The identification and selec-tion of mentors followed systematic proce-dures described in our prior work (Atkins,Frazier et al., 2008; Neal et al., 2008, 2011).First, staff in an instructional role (i.e., K-8teachers, reading specialists, and paraprofes-sionals) participated in 10-min sociometric in-terviews in which they provided demographicinformation and nominated an unlimited num-ber of colleagues with whom they consultedregarding (1) classroom management and (2)motivating learners. Teachers (n � 22 of 23eligible staff) indicated how many times permonth they consulted with and how emotion-ally close they felt to each nominated col-league (1 � Not at All; 5 � Very Close).Second, sociometric interview data were usedto identify KOLs based on their peer-nomi-nated ability to provide direct advice on class-room management and motivating students tothe greatest number of K-8 grade teacherswithin their social network (Neal et al., 2008,2011). Conducting sociometric interviewswith all school personnel rather than just class-room teachers increases the likelihood of iden-tifying individuals within the school commu-nity who make important contributions to in-struction and was essential to accuratelyidentifying KOLs in close contact with themajority of teachers and thus best positionedto encourage the spread of information aboutevidence-based practices (Atkins, Frazier etal., 2008a; Neal et al., 2008, 2011). After thetop five teachers with the highest percentageof nominations in both classroom managementand student motivation domains were identi-fied, we ensured those teachers also met ourdemographic criteria (i.e., minimum 5 yearsteaching and 2 years working at the school).The top five peer-nominated colleaguesreached an average of 31% of teachers withintheir social network (range � 19%–57%)when compared to non-nominated teachers,who reached an average of 6% of their col-leagues (range � 0%–14%). The two mostinfluential teachers identified via the afore-mentioned procedures voluntarily transferred


472

to other schools in June 2009; therefore, thethird and fourth most influential teachers (bothreaching 24% of teachers within the network)were successfully recruited as mentors. Onementor was the lead literacy specialist at theschool–she was European American andhad 27 years of teaching experience. The othermentor was an African American male with 7years of experience as an eighth-grade scienceteacher. Both mentors had attained master’sdegrees.

Coaches. Coaches were selected from acadre of retired teachers affiliated with a localcollege of education with experience workingin urban schools, conducting classroom obser-vations, and facilitating feedback conferences.The first coach was a European Americanmale with 36 years of experience as a specialeducator and administrator with a master’sdegree in special education. He had been re-tired for 5 years, during which time he wasmentoring undergraduate students completingtheir student teaching assignments. He deliv-ered coaching from September through No-vember 2009, but responsibilities outside ofthe project grew increasingly demanding andhe resigned after 3 months of coaching. Apostdoctoral fellow with extensive training inclassroom management and consultation withurban teachers filled in as coach during theinterim 2 months until we hired and trained asecond coach. The second coach was an Afri-can American female with 37 years of expe-rience as a classroom teacher, consultant, andprincipal with a master’s degree inadministration.

Early career teachers. The plannedsample included early career teachers with 3or fewer years of teaching experience based onevidence that teachers are most likely to leaveteaching or transfer to other schools within thefirst 3 years of teaching (e.g., Guarino, San-tibanez, & Daley, 2006). However, a signifi-cant number of early career teaching positionswere eliminated district-wide during Year 1 ofthe study due to the economic recession; there-fore, we extended eligibility criteria to five orfewer years. Six teachers were eligible to par-

ticipate, five consented, three had 3 or feweryears of experience (see Table 1 for demo-graphic details).

Additional school personnel. At therequest of the leadership team, all school per-sonnel (e.g., general and special educators,paraprofessionals, security guards, readingspecialists, guidance counselor) were invitedto participate in monthly PLC meetings. Sev-enty-four percent (i.e., 17 of 23) of eligibleschool personnel consented to participate inthe PLCs and contributed data, including 2KOL mentors who cofacilitated the PLCs, 5early career teachers (described earlier),and 10 school personnel (see Table 1).

Measures

Fidelity measures. Three fidelity mea-sures were developed to examine adherence tothe core components of the service modelbased on the Multisystemic therapy model offidelity measurement (Schoenwald, Letour-neau & Halliday-Boykins, 2005). Accord-ingly, service recipients (i.e., early careerteachers, additional school personnel) reportedon the activities of service providers (i.e.,mentors and coaches) for each service compo-nent (i.e., group seminars, coaching, PLCs).After each group seminar, early career teach-ers completed a 16-item yes/no checklist indi-cating whether specific instructional methodswere used (e.g., didactic instruction, modelingof a strategy, practice, performance feedback,and problem-solving implementation barriers)and whether each element was helpful. At theconclusion of each PLC, participants com-pleted a 17-item yes/no checklist that similarlyasked about the use of specific instructionalmethods in addition to helpfulness ratings. Asdescribed previously, coaching encompassedthree types of service contacts (i.e., preconfer-ence, classroom visits, and postconference).Following each coaching contact, early careerteachers completed a yes/no checklist (rangingfrom 15 to 20 items) indicating whether eachcoaching element (e.g., modeling, perfor-mance feedback, observation) was used andhow helpful each element was. Coaches alsocompleted weekly logs tracking frequency and


473

Table 1Early Career Teacher and Additional School Personnel Demographic

Characteristics

Early Career Teachers (N � 5)

M Yrs. Teaching (SD) 3.4 (2.01)Range � 0–5 years

Grade/Role Kindergarten (n � 1)4th/5th Grade (n � 1)7th/8th Grade (n � 1)

Special Education (6–8th) (n � 1)Art (n � 1)

Highest Degree Master’s Degree through Alt. Cert. (n � 3)Bachelor’s in Education (n � 2)

Gender N � 5 Female

Race/Ethnicity African American (n � 3)European American (n � 1)

Asian American (n � 1)

Year 1 Data Contributed Group Seminar FidelityCoaching Fidelity

PLC FidelityMid Year Focus Group

Additional School Personnel (N � 10)

M Yrs. Teaching (SD) 15.9 (10.87)Range � 1–32 years

Grade/Role Head Start (n � 3)1st/2nd (n � 1)3rd/4th (n � 1)

Special Education (n � 4)Guidance Counselor (n � 1)

Highest Degree 40% Master’s Degree30% Bachelor’s Degree10% Associates Degree

20% HS Diploma or Equivalent

Gender 91% Female

Race/Ethnicity 73% African American18% European American

9% Biracial

Year 1 Data Contributed PLC FidelityPLC Focus Group

Note. Data reflect demographic data reported at baseline.


474

length of coaching contacts and contentcovered.

Focus groups. Teacher satisfaction wasalso assessed with one midyear focus groupwith all early career teachers (n � 5) after 3months of participating in coaching and groupseminars. A second focus group was con-ducted at the end of the first year after 4months of PLCs. The PLC focus group partic-ipants included two head start teachers, oneguidance counselor, one para-educator, in ad-dition to one kindergarten, 3rd-, 4th-, and 7th/8th-grade teacher (n � 7). Both focus groupswere 60 min long, conducted at the school,and moderated by the first author. The mod-erator followed a focus group protocol thatbegan with general questions (“Tell me aboutyour experience with the model so far”), fol-lowed by an introduction to the topic (“Now, Iwould like to know more about your experi-ence with group seminars/PLCs/coaching”)and open-ended probes for each area of in-quiry (“What parts of the group seminar haveyou found more and less helpful”). This for-mat ensured teachers could answer in an un-structured manner and allowed the moderatorto gain a deeper understanding of how teach-ers experienced each component (Morgan,1997). Focus groups were audiotaped andtranscribed verbatim by two experienced tran-scribers and the moderator checked the tran-scripts against the audiotapes for accuracy.

Analyses

Descriptive analyses examined the ex-tent to which the core components of themodel were delivered. Thematic analyseswere conducted using guidelines outlined byBraun and Clarke (2006) with the aid of Atlas.ti, a qualitative analysis software program(Murh & Friese, 2004). Braun and Clarke’s(2006) guidelines provided a flexible yetclearly delineated process for identifying andanalyzing thematic patterns: (1) becoming fa-miliar with the data, (2) generating initialcodes, (3) identifying themes, (4) reviewingand revising themes, and (5) refining themesand subthemes. The constant comparative ap-proach was used to ensure internal coherence

and consistency within each theme (Boeije,2002). Thematic analyses were conductedacross both focus groups, rather than sepa-rately to enhance the rigor of the analysis andbecause both groups were asked the same gen-eral questions regarding their experiences withthe model. However, the principle of constantcomparison was used to detect thematic simi-larities and differences in reported experiencesacross groups at each stage of the analysis(Boeije, 2002). The first and second authorindependently reviewed and coded each focusgroup transcript. Through an initial round ofopen coding, an initial list of descriptive codeswas developed by consensus. Low inter-rateragreement on descriptive codes at this stagewas addressed by meeting, clarifying, discuss-ing, and re-establishing consensus on opera-tional definitions followed by another round ofindependent coding. Kappa coefficients for thefinal list of 19 master codes ranged from 0.67to 1.0 (� � .81) indicating adequate inter-rateragreement (Thompson, McCaughan, Cullum,Dheldon, & Raynor, 2004). The identificationand refinement of themes was reached by con-sensus using the constant comparative ap-proach (Boeije, 2002).

Results

Implementation of the Model

Attendance records, fidelity tools, andcoaching logs were used to examine the de-gree to which the three core components weredelivered as designed. Completion rates forfidelity checklists were as follows: 92% forgroup seminars, 88% for PLCs, and 59% forcoaching.

Group seminars and PLCs. Figure 3highlights the percentage of early career teach-ers attending group seminars and PLCs. Tenof twelve planned group seminars were of-fered between the end of September 2009 andearly May 2010, suggesting that the majorityof seminars were delivered as planned. Atten-dance rates for group seminars varied acrossthe school year, ranging from 40% to 100%(M � 64%). Attendance rates for group sem-inars were strongest early in the school year,


475

weakest midyear (coinciding with the initia-tion of PLCs), and leveled off to 60% by theend of the school year. More than 90% ofgroup seminar checklists reflected that teach-ing methods such as didactic instruction, dis-cussion, troubleshooting, and modeling wereused versus 45% reflecting that active learningmethods were used.

Four PLCs were offered between Janu-ary and May 2010 suggesting the PLCs weredelivered monthly as originally intended withboth KOL mentors cofacilitating all PLCs.Attendance rates at the PLCs for the 5 earlycareer teachers (see Figure 3) ranged from60% to 100% (M � 80%) and 40% to 70%(M � 55%) for the ten additional school per-sonnel consenting to the research. Three ofthe 23 original eligible teachers participated inat least 50% of the PLC meetings withoutconsenting to the research. More than 90% ofPLC checklists reflected that the key instruc-tional methods were used by the KOL mentorswhen facilitating the PLCs (i.e., active learn-ing, video, troubleshooting, didactic, discus-sions, and handouts).

Coaching. Early career teachers partic-ipated in coaching activities between Septem-ber 2009 and May 2010. Coaching fidelity

measures and logs were ready for use in De-cember 2009; therefore, the following datarepresent a conservative estimate of servicereceipt (approximately 21 weeks excludingbreaks and statewide testing). Coaching logsreflected that new teachers participated in anaverage of 960 min or 16 hrs of coaching(SD � 2.45 hr). On average, early careerteachers spent 36% of their time in preconfer-ences (M � 20 min; SD � 20; range �1–105), 38% of their time in classroom visits(M � 52 min; SD � 18; range � 15–95),and 19% of their time in postconferences(M � 14 min.; SD � 12; range � 5– 60).Coaching checklists revealed that coachesrelied on discussion, observation, and didac-tic methods to a much greater degree thanperformance feedback, modeling, or prac-tice. Coaching checklists also reflected thatconferences covered behavioral manage-ment strategies (e.g., activity and classroomrules, preventing and responding to misbe-havior) to a much greater degree than rela-tionship building strategies with students orfamilies. Coaching logs corroborated thosedata and indicated that only 8% of coaching-related activities fell in the “other” category,including early engagement and relationship

Figure 3. Percentage of early career teachers attending the ten group seminarsand four PLCs during the 2009–2010 school year.


476

building, scheduling, and troubleshootinglogistics.

Response to the Model

Teacher satisfaction was examinedthrough a combination of fidelity checklistsand focus group data. More than 90% of groupseminar checklists reflected that early careerteachers experienced didactic instruction,handouts, discussion, problem-solving, mod-eling, and performance feedback as morehelpful relative to videos (58%), active learn-ing (75%), and practice using a strategy(64%). Ninety percent or more of PLC check-lists reflected that teachers experienced didac-tic instruction, handouts, and problem-solvingas more helpful than videos (80%) and activelearning (60%). More than 90% of the pre-and postconference checklists reflected thatearly career teachers experienced didacticmethods, discussion, performance feedback,and reflective practice as helpful. In contrast,only three-quarters of the checklists reflectedthat modeling by the coach, coteaching a les-son with the coach, and coach observation washelpful. Focus groups elicited more compre-hensive feedback regarding the content andformat of group seminars, coaching, andPLCs. Through thematic analyses of focusgroup data, three broad themes (i.e., satisfac-tion with the model, professional isolation,and enhancing feasibility) provide insight intoteachers’ initial response to the model andhow these experiences compared and con-trasted with their current work context. Thesethemes are described in greater detail below,with quotations highlighting key features ofthe exchange.

Satisfaction with the model. Onetheme that emerged across both focus groupswas a positive reaction to the service model.Three subthemes emerged from the analysis ofteachers’ responses to the two componentsrequiring teachers to meet in a group format—group seminars and PLCs. Those subthemesincluded treatment as professionals with ex-pertise to share (rather than passive recipientsof information), opportunities to share ideaswith colleagues, and respectful and profes-

sional treatment of teachers’ time. These threeteacher perceptions of the model appear to beinterdependent. That is, group seminars andPLCs created consistent, predictable opportu-nities to meet and provided a format that en-couraged the exchange of ideas, experiences,and expertise vertically from mentors and hor-izontally from colleagues. Early career teach-ers and PLC participants described welcomingthe opportunity for collegial interactions andbenefiting from hearing their colleagues sharegeneral classroom experiences in addition toexperiences with specific strategies. The fol-lowing are two quotes illustrating how thePLCs were perceived:

You guys had an agenda but…you let us talkand you let us discuss…. We bounced ideasoff of each other and even if we didn’t getthrough all of it, it was still to the pointwhere what we found was something that weneeded to focus on and that was not just aventing session but something that reallyhelped…. And I think why those worked sowell is because it was during the school day.You know, you ask somebody to come be-fore school or after school it’s so difficult.

I don’t really deal with classroom manage-ment on a regular basis but from an obser-vation standpoint, I saw it as a nice time forteachers to reflect and, like you said, thereare two pieces to it, just getting some con-crete things they can use in the classroom,and I think part of it is . . . teachers heredon’t like sit down for lunch and like talkand share ideas.

Professional isolation. Satisfactionwith the model appeared to be associated withteachers’ experience of professional isolationin their current work context. Several sub-themes describing professional isolation wereidentified: limited communication betweenteachers around practice, desire to communi-cate without the time to do so, lack of personalconnections with colleagues, and insularity.Data from both focus groups converged in thedescription of insular classrooms and verylimited opportunities for collegial exchange ofinformation, experiences, and expertise. De-spite significant overlap in reported experi-ences across the two focus groups, there wereseveral differences in what each focus groupemphasized when describing experiences of


477

isolation and insularity. The isolation de-scribed in the early career teacher focus grouphighlighted a lack of personal connection tocolleagues and not feeling fully indoctrinatedinto school routines (e.g., school-wide behav-ior management system), whereas insularity inthe context of the PLC focus group high-lighted a lack of knowledge regarding whatcolleagues were doing in their classrooms andthe type of instructional practices their col-leagues used. Across both focus groups, teach-ers highlighted significant time constraints(e.g., 20-min lunches spent preparing for thenext lesson) and infrequent and inconsistentopportunities to meet and talk with colleaguesat the root of the isolation and insularity theyexperienced. The PLCs and group seminarswere described as providing teachers with in-creased opportunities to share and collaboratewith colleagues, thereby reducing their senseof isolation. The following quote describes anearly career teacher’s experience of isolation:

No one knows what’s going on between thedifferent rooms and it’s really frustrating, Ithink because I would love to tell peoplewhat we’re doing in my room and I wouldlove to be able to use some of that in otherrooms and…. I think that’s the part that’sfrustrating. We’re so isolated from eachother.

Another early career teacher shared thatthrough the group seminars she was first ex-posed to the school-wide behavior manage-ment system using toothpicks that she nowimplements in her classroom:

The part that I’ve gotten the most is just thefact that I know what you guys are doing inyour classrooms. Like, I know your rulesbecause of these seminars. Like, I know what[Teacher X] does in her room and I don’tusually know that stuff because I’m sure youfeel the same way. Like, I didn’t know thetoothpick thing until this part of the year andI feel really bad about that because…. I don’tknow why I didn’t know it. So that kind ofstuff is nice because now I use it in my ownclassroom and I feel like a little bit moreinvolved.

Enhancing feasibility. Although com-ponents of the model (i.e., PLCs and groupseminars) created new opportunities forteacher collaboration and in turn helped miti-

gate some of the isolation and alienationteachers experienced at their school, teachersparticipating in both focus groups identifiedthe need to adapt the model and the strategiessubsumed within the model to better reflect therealities of their work environment and theirneed for concrete support. Three interrelatedthematic patterns related to teachers’ needsemerged: relevance (i.e., the extent to whichthe strategies were perceived as applicable andappropriate to teachers’ context), instrumentalsupport (i.e., the need for tangible materialsand concrete assistance around implementa-tion of new strategies), and perceptions ofeffectiveness (i.e., beliefs regarding the extentto which the strategies introduced would beeffective in reducing student behavior prob-lems and enhancing motivation). Feasibility—teachers’ beliefs regarding successful integra-tion of the strategies into their instructionalday and students’ ability to benefit from thosestrategies—was conceptualized as the byprod-uct of a service model that was highly relevantto teachers’ context, that included instrumen-tal support around implementation, and thatenhanced perceptions regarding effectiveness.

With regard to relevance, analyses re-vealed that teachers wanted access to trainingmaterials that reflected urban schools and thestudents they served and more discussion re-garding how to adapt the strategies accord-ingly. Teachers, for example, described want-ing access to videos with greater face validity,depicting overcrowded classrooms with heter-ogeneous learners and students with intensebehavioral and academic needs.

Teacher A: Yeah, I’m like, “Wow! I wish Ihad those kids!”[Laughter]Teacher B: You know, kids that sit there.Teacher A: Kids that sit there quietly andwait.

One early career teacher offered to vid-eotape her classroom to augment the authen-ticity of the training materials. Both new andmore experienced teachers also requestedmore time and consideration for how to mod-ify specific strategies to accommodate the de-velopmental and instructional level of individ-ual students.


478

Instrumental support was operational-ized as providing teachers with a training cur-riculum that included tangible and reproduc-ible materials (e.g., tokens, tickets) and settingaside time during group seminars and PLCs toprepare strategy-related materials. Instrumen-tal support also included specific instructionsand guidelines from mentors around when toimplement strategies (e.g., “I really just needyou to definitely do it for the week”). Teachersexplained that significant time constraints in-terfered with follow through and that theaforementioned instrumental support wouldenhance their implementation efforts. Finally,instrumental support also included teachers’need for concrete assistance around imple-mentation. Early career teachers, on one hand,described the need for more concrete supportfrom the coach, including feedback that wasaffirmative, constructive, and immediate, inaddition to coaches taking a more active roleduring classroom visits. PLC participants, onthe other hand, described concrete supportthey could provide to each other given theirdiverse skills and expertise. Media and tech-nology (i.e., e-mail, blogging, creating a ‘vir-tual school community’) emerged as a methodfor providing instrumental assistance to thoseteachers requiring more or different support.

Finally, teachers’ perception that a strat-egy would or could be effective appeared to beinfluenced by either personal success with astrategy or collegial endorsement. Below twoteachers discuss the use of tootles, a classwidestrategy designed to promote positive peer in-teractions and reduce tattling (Embry, Straate-meier, Richardson, Laugher, & Mitch, 2003).

Teacher A: If somebody else you know hasused it and you know those children and youknow that it would work personally in yourclassroom.

Teacher B: Right. . . Why were stuck ontootles because [early career teacher] didtootles and it was a lively, engaged conver-sation and she gave us that information . . .and I think because it was implemented inher classroom . . . It was very authentic whenshe did it.

Perceptions of effectiveness and will-ingness to experiment with a strategy was alsorelated to the depth of coverage of each strat-

egy. Specifically, teachers shared that despiteappreciating choice around different strategiesto use, they would have preferred more thor-ough coverage of fewer strategies—lessbreadth, more depth.

And just narrowing it down, you know? Nar-rowing it down to make it where, like yousaid, it’s just feasible in the classroom andyou know that you can hit on at least onething…. And whether you take away fivestrategies from the whole year or one strat-egy, at least…it was something that workedfor you.

Discussion

The current study launches a program ofresearch to support urban early career teachersby targeting common obstacles to teachingeffectively and by building in protective fac-tors to help novices thrive in a demandingwork context. The Teachers SupportingTeachers in Urban Schools (TST) model mar-shals resources through an influential mentorand external coach with the goal of enhancingnew teachers’ effectiveness and connectednessto colleagues. The mixed method iterative de-velopment framework allowed us to prelimi-narily examine during Year 1 whether themodel was delivered as intended and accept-able, and informed adaptations for subsequentfunding years with the addition of two newschools. The overarching goal of the 3-yearstudy is to create a service model (vs. a cannedcurriculum) that responds to the unique needsof each school context and can be adapted interms of structure and content to more closelymatch existing school norms, goals, and pri-orities around instruction and behavior.

Adaptations

The incorporation of early feedbackfrom our school partners allowed us to capi-talize on the continuous flow and exchange ofinformation between the research team, theleadership team, and the service providers(coaches and KOLs) early in the developmentprocess. A combination of implementationand teacher satisfaction data analyzed after ourfirst funding year informed a number of adap-


479

tations that guided future iterations of themodel, which are described below.

Less is more. Despite wanting to designa model that included a menu of options toenhance teacher motivation and engagement,in this time- and resource-limited environ-ment, teachers wanted more information aboutfewer strategies and specific instructions re-garding exactly when to implement them.Thus, during Year 2 of the study we integratedseveral seminar topics (e.g., rules, procedures,physical arrangement of the classroom, andpositive communication with families) into thefirst few coaching sessions to provide newteachers with access to critical strategies earlyin the school year. We also streamlined thegroup seminar and PLC curriculum duringYear 2 of the study to ensure mentors onlyintroduced one strategy per session and allo-cated more time to review, model, and discussadaptations. Rather than introducing new con-tent during subsequent meetings, more timewas allocated to discussing implementationbarriers, troubleshooting next steps, and adapt-ing the strategies to better fit teachers’ idio-graphic needs. In addition, we used schoolannouncements, reminder memos, and“themes of the week” (e.g., “Tootles Week”)to remind teachers of the school-wide strate-gies being implemented.

It is unclear whether the decline in groupseminar attendance midyear for early careerteachers was a function of time demands,identifiable stress cycles for teachers, or per-haps the perception by new teachers that PLCscould supplant group seminars (Hembling &Gilliland, 1981). Interestingly, PLC atten-dance rates followed a similar trend, withstrongest participation rates during initialPLCs, weakest during the middle sessions, andthen rebounding during the final meeting. Al-though we speculate that PLC attendance rateswere a function of scheduling, with poorestattendance on PLCs scheduled after schoolversus district-wide professional developmentdays, significant time constraints raised byfocus groups participants coupled withlower attendance rates for early career teach-ers when PLCs and group seminars were

scheduled concurrently (see Figure 3)prompted our team to incorporate regularfeedback sessions with early career teachersduring subsequent funding years to ensure thatprofessional development activities were notbecoming burdensome.

Lower completion rates for coaching fi-delity checklists (59%) completed by earlycareer teachers after each coaching contactrelative to PLC and group seminar checklists(88% and 92% completion rates, respectively)further suggested that completing fidelity mea-sures after every coaching contact (i.e., pre-conference, classroom visit, and postconfer-ence) was too burdensome for early careerteachers. Therefore, in an effort to create anefficient and effective suite of fidelity mea-sures (see Schoenwald et al., 2011), duringsubsequent funding years we reduced admin-istration of coaching fidelity from weekly tomonthly whereby teachers were asked to con-sider their most recent preconference, class-room visit, and postconference, and added arating of how representative each coachingsession was relative to other sessions takingplace that month. During Year 2 we also ad-ministered a parallel version of the coachingfidelity checklist to coaches in order to assessthe extent to which we find convergence be-tween early career teacher and coach ratingswhen the date of service is yoked. That is, ifwe learn that coaches can reliably and validlytell us what happened during coaching, wemay only need coaches to complete this mea-sure in a larger efficacy trial.

Multiple pathways to effectivenessand connectedness. The service model re-mains grounded in promoting early careerteachers’ effectiveness and connectedness toimprove teacher retention and support studentlearning. Teacher satisfaction data, for exam-ple, highlighted that the group format (i.e.,group seminars and PLCs) mitigated some ofthe isolation that teachers, particularly newerteachers, experienced in their school. It is im-portant to note that some early career teachershad up to 4 years of teaching experience attheir school and still experienced their work asinsular, which suggests that altering the social


480

context of a school takes time. Focus groupdata suggested that PLCs and seminars helpednovices become more socially integrated intothe school milieu and informed about existingschool-wide practices. These findings supportwhat is known about the critical role that so-cial connections, peer collaboration, and asupportive school community plays in helpingteachers cope with stressors, manage teachingdemands, and grow professionally (Bryk &Schneider, 2002; Shernoff et al., 2011). Insubsequent funding years, we also incorpo-rated media and the Internet (e.g., blogs, chatrooms, e-mail) to create a more flexible infra-structure to support teacher collaboration.

Findings also highlighted that teachers’perceived effectiveness in experimenting withinnovative practices appeared to be related inpart to the level and type of instrumental sup-port provided. This is consistent with sociallearning theory, which guided our conceptualmodel and highlights that learning and behav-ior are grounded in a social context and thatlearning occurs vicariously by observing oth-ers (Bandura, 1997). Coaching fidelity dataand focus group data, however, revealed thatearly career teachers spent a limited amount oftime in postconferences and needed moremodeling, demonstration, and timely feedbackfrom the coach, particularly following class-room visits.

Based on these early findings, severaladaptations to the coaching model were madeimmediately following the midyear focusgroup with early career teachers, which coin-cided with hiring a new coach. These modifi-cations included written feedback from thecoach after classroom visits when postconfer-ences were not possible to enhance communi-cation with and feedback to early career teach-ers. In addition, with permission from ourinternal review board, we encouraged coachesand early career teachers to use technology(e.g., e-mail, phone calls) to enhance commu-nication and to create greater fluidity aroundmodes for planning and feedback. Becauseearly career teachers also reported needingmore active modeling and demonstration ofstrategies during classroom visits and moreconstructive feedback during postconferences,

we incorporated more field-based training andcocoaching with the newly hired coach, withuniversity consultants playing a more activerole. This was done in an effort to maximizethe coaches’ understanding of, comfort with,and fidelity to an active coaching model.Coaching logs revealed that the mean numberof minutes spent in postconferences increasedfrom 12.5 to 17 after implementing changes tothe coaching model, suggesting that, at a min-imum, early career teachers and coaches werehaving more opportunities to communicatewith one another and plan together aftercoaching modifications were made. Fidelitychecklists, however, did not reveal changes inthe degree to which early career teachers per-ceived modeling, coteaching, or coach obser-vations to be more helpful after modifying thecoaching component. Taken together, thesefindings suggested during subsequent fundingyears the need to create more systematic op-portunities for the coach, the research team,and the early career teachers to strategize to-gether around how to enhance the coachingcomponent and more frequent opportunitiesfor informal feedback sessions and check-insthroughout the year.

We also learned that teachers’ percep-tions of effectiveness and feasibility were re-lated to their colleagues’ (not just KOLs) ex-perience of success. Hence, during the secondyear of the intervention we increased oppor-tunities for early career teachers to share theirimplementation successes given their early ex-posure to strategies through coaching. This isparticularly critical in light of not recruitingthe most influential mentors during Year 1;thus, part of creating a responsive, contextu-ally aligned service model included drawingon additional indigenous resources (e.g., dis-ciplinarians) to support implementation andsustainability of the model. Indeed, diffusionof innovations literature highlights that KOLsare often influential early on to encourageuptake, but that collegial relations may bemore important for maintenance—this is theway in which the tipping point is operational-ized—the point at which an innovation isadopted by the social network (Rogers, 2005).Thematic analyses also revealed that percep-


481

tions of effectiveness could be enhancedthrough tangible, relevant, and differentiatedtraining materials. Thus, during the followingyear we embedded “Make it Take It” activitiesinto group seminars and PLC meetings andallotted more time for teachers to prepare allstrategy-related materials (e.g., Good Behav-ior Game rules created on poster board, tootlescards developed). We also integrated addi-tional visual media into the curriculum thatmore closely reflected the reality of teachers’classrooms, including using commercial videoclips.

Limitations

Although it is unclear whether our ex-perience in one school generalizes to thebroader population of urban schools given thesmall and restricted sample of early careerteachers, Year 1 data provide a window intothe feasibility of the service model from theperspective of our community partners andservice recipients, which is an important firststep in development (Hoagwood et al. 2002).Given limitations related to the small samplesize, several steps were taken to conduct arigorous and transparent qualitative inquiry:inter-rater reliability for the descriptive mastercodes was obtained, a systematic process foridentifying thematic patterns was used (Braun& Clarke, 2006), and a step-by-step constantcomparative approach was used to identifythemes (Boeije, 2002). Although focus groupswere a logical methodology for the currentstudy given they facilitate the collection andanalysis of extensive data in a brief period oftime, several limitations to this methodologywarrant mention. First, because of the inherentcollective nature of focus groups, they oftenlimit the breadth and depth of informationobtained from each participant when com-pared to individual interviews. Second, focusgroups are contrived social settings created bythe researcher among individuals with a pre-sumed shared set of experiences in common(Kidd & Parshall, 2000; Morgan, 1997).Hence, data can be influenced by group dy-namics, including domineering group mem-bers, varying comfort in revealing personal

information in a group format, and individualpressure to conform and/or censor opinions(Kidd & Parshall, 2000). The group moderatoraddressed these potential problematic groupdynamics by encouraging different groupmembers, in particular more silent members,to share their perspectives and experiences.

Implications for Science and Practice

Given the urgent, intensive needs facingteachers working in inner-city, high-povertyschools, this program of research sheds initiallight on how to help new teachers overcomesome of the barriers to teaching effectively,remaining in the profession, and promotingpositive outcomes for at-risk learners. Becausementoring and induction programs to datehave scant evidence of a direct connectionwith student achievement, our service modelhas the capacity to further knowledge by fo-cusing explicitly on improving new teachers’practices around observable characteristics ofteaching related to positive outcomes forlearners.

This work reflects recent calls to de-velop contextually relevant interventions andservice models that better reflect the exigen-cies of real-world practice settings (Hoag-wood, Burns, Kiser, Ringeisen, & Schoen-wald, 2001; Hoagwood et al., 2002). Based onwhat we have learned thus far regarding im-plementation and satisfaction, we had an op-portunity to make key adaptations to themodel to meet the specific needs identified bythis small sample of urban teachers. Duringsubsequent funding years, we will assess theextent to which the model is producing prom-ising teacher outcomes related to perceivedand objective effectiveness and connected-ness. Although small scale, this study doesillustrate the extent and nature of support re-quired for early career teachers in addition tohow and under what conditions they are likelyto perceive a service model and specific strat-egies as feasible and effective.

References

Albin, R. W., Lucyshyn, J. M., Horner, R. H., & Flannery,K. B. (1996). A model for goodness of fit. In L. K.


482

Koegel, R. L. Koegel, & G. Dunlap (Eds.), Positivebehavior support: Including people with difficult be-havior in the community (pp. 81–98). Baltimore:Brookes.

Allinder, R. M. (1994). The relationship between efficacyand the instructional practices of special educationteachers and consultants. Teacher Education and Spe-cial Education, 17, 86–95. doi: 10.1177/088840649401700203

Andrzejewski, M. E., Kirby, K. C., Morral, A. R., &Iguchi, M. Y. (2001). Technology transfer throughperformance management: The effects of graphicalfeedback and positive reinforcement on drug treatmentcounselors’ behavior. Drug and Alcohol Dependence,63, 179–186.

Annie E. Casey Foundation. (2002). End games: Thechallenge of sustainability. New York: Author.

Atkins, M., Graczyk, P., Frazier, S., & Adil, J. (2003).Toward a new model for school-based mental health:Accessible, effective, and sustainable services in urbancommunities. School Psychology Review, 32, 503–514.

Atkins, M. S., Frazier, S. L., Leathers, S. J., Graczyk,P. A., Talbott, E., Jakobsons, L., et al. (2008). Teacherkey opinion leaders and mental health consultation inurban low-income schools. Journal of Consulting andClinical Psychology, 76, 905–908.

Atkins, M. S., Mehta, T., Frazier, S. L., Cappella, E.,Marínez-Lora, A., Shernoff, E., et al. (2008). Links toLearning: School and community mental health re-source alignment in support of children’s learning inhigh-poverty urban schools. Paper symposium at theannual meeting of the American Psychological Asso-ciation. Boston, MA.

Atkins, M. S., Hoagwood, K. E., Kutash, K., & Seidman,E. (2010). Toward the integration of education andmental health in schools. Administration and Policy inMental Health and Mental Health Services Research,37, 40–47. doi: 10.1007/s10488-010-0299-7

Bandura, A. (1997). Self efficacy: The exercise of control.New York: W. H. Freeman and Company.

Barnes, G., Crowe, E., & Schaefer, B. (2007). The cost ofteacher turnover in five school districts: A pilot study.National Commission on Teaching and America’s Fu-ture.

Benazzi, L., Horner, R. H., & Good, R. H. (2006). Effectsof behavior support team composition on the technicaladequacy and contextual fit of behavior support plans.The Journal of Special Education, 40, 160–170.

Berry, B. (2004). Recruiting and retaining ‘highly quali-fied teachers’ for hard-to-staff schools. NASSP Bulle-tin, 88, 5–27. doi: 10.1177/019263650408863802

Boeije, H. (2002). A purposeful approach to the constantcomparative method in the analysis of qualitative in-terviews. Quality and Quantity, 36, 391–409.

Borman, G. D., & Maritza Dowling, N. (2008). Teacherattrition and retention: A meta-analytic and narrativereview of the research. Review of Educational Re-search, 78, 367–409. doi: 10. 3102/0034654308321455

Boyd, W. L., & Shouse, R. C. (1997). The problems andpromise of urban schools. In H. J. Walberg, O. Reyes,& R. P. Weissberg (Eds.), Children and youth: Inter-disciplinary perspectives (pp. 141–165). ThousandOaks, CA: Sage Publications.

Braun, V., & Clarke, V. (2006). Using thematic analysisin psychology. Qualitative Research in Psychology, 3,77–101. doi: 10.1191/1478088706qp063oa

Bryk, A. S., Camburn, E., & Seashore Louis, K. (1999).Professional community in Chicago elementaryschools: Facilitating factors and organizational conse-quences. Educational Administration Quarterly, 35,751–781.

Bryk, A. S., & Schneider, B. (2002). Trust in schools: Acore resource for improvement. New York: RussellSage Foundation.

Burley, W. W., Hall, B. W., Villeme, M. G., & Brock-meier, L. L. (1991). A path analysis of the mediatingrole of efficacy in first-year teachers’ experiences, re-actions, and plans. Paper presented at the annual meet-ing of the American Educational Research Associa-tion, Chicago.

Cappella, E., Frazier, S. L., Atkins, M. S., Schoenwald,S. K., & Glisson, C. (2008). Enhancing schools’capacity to support children in poverty: An ecolog-ical model of school-based mental health services.Administration and Policy in Mental Health, 35,395– 409.

Coladarci, T. (1992). Teachers’ sense of efficacy andcommitment to teaching. The Journal of ExperimentalEducation, 60, 323–337.

Creemers, B. P. M. (1994). Effective instruction as a basisfor effective education in schools. Tijdschrift voorOnderwijsresearch, 19, 3–16.

Crone, L. J., & Teddlie, C. (1995). Further examination ofteacher behavior in differentially effective schools: Se-lection and socialization process. Journal of Class-room Interaction, 30, 1–9.

Embry, D. D., Straatemeier, G., Richardson, C., Laugher,K., & Mitch, J. (2003). The PAX Good BehaviorGame. Center City, MN: PAXIS Institute.

Evertson, C. M., & Weinstein, C. S. (Eds.). (2006). Hand-book of classroom management: Research, practice,and contemporary issues. Hillsdale, NJ: Lawrence Erl-baum.

Good, T. L., & Brophy, J. E. (2003). Looking in class-rooms. New York: Allyn & Bacon.

Guarino, C. M., Santibanez, L., & Daley, G. A. (2006).Teacher recruitment and retention: A review of therecent empirical literature. Review of Educational Re-search, 76, 173–208. doi: 10.3102/00346543076002173

Hembling, D. W., & Gilliland, B. (1981). Is there anidentifiable stress cycle in the school year? The AlbertaJournal of Educational Research, 27, 324–330.

Hoagwood, K., Burns, B. J., Kiser, L., Ringeisen, H., &Schoenwald, S. K. (2001). Evidence-based practice inchild and adolescent mental health services. Psychiat-ric Services, 52, 1179–1189. doi: 10.1176/appi.ps.52.9.1179

Hoagwood, K., Burns, B. J., & Weisz, J. R. (2002). Aprofitable conjunction: From science to service in chil-dren’s mental health. In B. J. Burns & K. Hoagwood(Eds.), Community treatment for youth: Evidence-based interventions for severe emotional and behav-ioral disorders (pp. 327–338). New York: OxfordPress.

Ingersoll, R. M. (2001). Teacher turnover and teachershortages: An organizational analysis. American Edu-cational Research Journal, 38, 499–534.


483

Ingersoll, R. M., & Kralik, J. M. (2004). The impact ofmentoring on teacher retention: What the researchsays [Research Review]. Denver, CO: Education Com-mission of the States.

Johnson, S, M., & Birkeland, S. E. (2003). Pursuing a“sense of success”: New teachers explain their careerdecisions. American Educational Research Journal,40, 581–617. doi: 10.3102/00028312040003581

Johnson, S. M., & The Project on the Next Generation ofTeachers (2004). Finders and keepers: Helping newteachers survive and thrive in our schools. San Fran-cisco: Jossey-Bass.

Jones, D. L., & Sandidge, R. F. (1997). Recruiting andretaining teachers in urban schools: Implications forpolicy and the law. Education and Urban Society, 29,192–203. doi: 10.1177/0013124597029002006

Joyce, B., & Showers, B. (2002). Student achievementthrough staff development. In B. Joyce, & B. Showers(Eds.), Designing training and peer coaching: Ourneeds for learning. Crozet, VA: Virginia Associationfor Supervision and Curriculum Development.

Kardos, S. M., Johnson, S. M., Peske, H. G., & Kauff-man, D. (2001). Counting on colleagues: New teach-ers encounter the professional cultures of theirschools. Educational Administration Quarterly, 37,250 –290.

Kardos, S. M., & Johnson, S. M. (2007). On their own andpresumed expert: New teachers’ experience with theircolleagues. Teachers College Record, 109, 2083–2106.Retrieved from http://www.gse.harvard.edu/�ngt/Kardos%20Johnson%20article%20Kardos_and_Johnson_TCR_2006.pdf

Kataoka, S. H., Zhang L., & Wells, K. B. (2002). Unmetneed for mental health care among U.S. children: Vari-ation by ethnicity and insurance status. American Jour-nal of Psychiatry, 159, 1548–1555. doi: 10.1176/appi.ajp.159.9.1548

Kauffman, D., Johnson, S. M., Kardos, S. M., Liu, S. M.,& Peske, H. G. (2002). “Lost at sea:” New teachers’experience with curriculum and assessment. TeachersCollege Record, 104, 273–300.

Keigher, A. (2010). Teacher attrition and mobility: Re-sults from the 2008–09 Teacher Follow-up Survey.(NCES 2010–353. U.S. Department of Education:Washington, DC: National Center for Education Sta-tistics. Retrieved April 20, 2010, from http://nces.ed.gov/pubsearch

Kelley, M. L. (1990). School-home notes: Promoting chil-dren’s classroom success. New York: Guilford Press.

Kidd, P. S., & Parshall, M. B. (2000). Getting the focusand the group: Enhancing analytical rigor in focusgroup research. Qualitative Health Research, 10, 293–308.

Lortie, D. C. (2002). Schoolteacher: A sociological study.Chicago, IL: University of Chicago Press.

Miller, W. R., & Rollnick, S. (2002). Motivational inter-viewing: Preparing people for change, Vol. 2. NewYork: Guilford Press.

Morgan, D. (1997). Focus groups as qualitative research(2nd ed.). Thousand Oaks, CA: Sage.

Murh, T., & Friese, S. (2004). User’s manual for ATLAS.ti 5.0 (2nd ed.). Berlin, Germany: Scientific SoftwareDevelopment.

Neal, J. W., Neal, Z. P., Atkins, M. S., Henry, D. H., &Frazier, S. L. (2011). Channels of change: Contrastingnetwork mechanisms in the use of interventions. Amer-

ican Journal of Community Psychology, 47, 277–286.doi 10.1007/s10464-010-9403-0

Neal, J., Shernoff, E. S., Frazier, S. L., Stachowicz, E.,Frangos, R., & Atkins, M. (2008). Change from within:Engaging teacher key opinion leaders in the diffusionof interventions in urban schools. The Community Psy-chologist, 41, 53–57.

Pianta, R. C., La Paro, K. M., & Hamre, B. K. (2007).Classroom Assessment Scoring System—CLASS. Bal-timore, MD: Brookes.

Pianta, R. C., Mashburn, A. J., Downer, J. T., Hamre,B. K., & Justice, L. (2008). Effects of web-mediatedprofessional development resources on teacher-childinteractions in pre-kindergarten classrooms. EarlyChildhood Research Quarterly, 23, 431–451.

Poglinco, S. M., Bach, A. J., Hovde, K., Rosenblum, S.,Saunders, M., & Spuovitz, J. A. (2003). The heart ofthe matter: The coaching model in America’s ChoiceSchools. Philadelphia, PA: Consortium for Policy Re-search in Education.

Protheroe, J. Bower, P., & Chew-Graham, C. (2007). Theuse of mixed methodology in evaluating complex in-terventions: Identifying patient factors that moderatethe effects of a decision aid. Family Practice, 24,594–600. doi: 10.1093/fampra/cmm066

Rivkin, S. G., Hanushek, E. A., & Kain, J. F. (2005).Teachers, schools, and academic achievement. Econo-metrica, 73, 417–458.

Rogers, E. M. (2005). Diffusion of innovations (5th ed.).New York: Free Press.

Rounsaville, B. J., Carroll, K. M., & Onken, L. S. (2001).NIDA’s Stage Model of behavioral therapies research:Getting started and moving on from stage I. ClinicalPsychology: Science and Practice, 8, 133–142.

Rubenstein, M., Patrikakou, E., Weissberg, R., & Arm-strong, M. (2000). Enhancing school-family partner-ships: A teacher’s guide. Chicago: University of Illi-nois at Chicago, Department of Psychology.

Sabers, D. S., Cushing, K. S., & Berliner, D. C. (1991).Differences among teachers in a task characterized bysimultaneity, multidimensionality, and immediacy. Amer-ican Educational Research Journal, 28, 63–88.

Schoenwald, S. K., Garland, A. F., Chapman, J. E., Fra-zier, S. L., Sheidow, A. J., & Southam-Gerow, M. A.(2011). Toward the effective and efficient measure-ment of implementation fidelity. Administration andPolicy in Mental Health and Mental Health ServicesResearch, 38, 32–43. doi: 10.1007/s10488-010-0321-0

Schoenwald, S. K., Letourneau, E. J., & Halliday-Boy-kins, C. (2005). Practicing therapist adherence to atransported family-based treatment for youth. Journalof Clinical Child and Adolescent Psychology, 34, 658–670. doi: 10.1207/s15374424jccp3404_8

Shernoff, E. S., Mehta, T., Atkins, M. S., Torf, R. &Spencer, J. (2011). A qualitative study of the sources andimpact of stress among urban teachers. School MentalHealth, 3, 59–69. doi: 10.1007/s12310-011-9051-z

Smith, T. M., & Ingersoll, R. M. (2004). What are theeffects of induction and mentoring on beginningteacher turnover? American Educational ResearchJournal, 41, 681–714. Retrieved from http://www.jstor.org/stable/3699442

Stein, M. K., & Wang, M. C. (1988). Teacher develop-ment and school improvement: The process of teacherchange. Teaching and Teacher Education, 4, 171–187.doi: 10.1016/0742-051X(88)90016-9


484

Stipek, D. J. (1996). Motivation and instruction. In D. C.Berliner & R. C. Calfee (Eds.), Handbook of educa-tional psychology (pp. 85–113). New York: Simon &Schuster Macmillan.

Thompson, C. T., McCaughan, D., Cullum, N., Sheldon,T. A., & Raynor, P. (2004). Increasing the visibility ofcoding decisions in team-based qualitative research innursing. International Journal of Nursing Studies, 41,15–20.

Tolan, P. H., & Henry, D. (1996). Patterns of psychopa-thology among urban poor children: Comorbidity andaggression effects. Journal of Consulting and ClinicalPsychology, 64, 1094–1099. doi: 10.1037//0022-006X.64.5.1094

Wang, M. C., Haertel, G. D., & Walber, H. J. (1997).Fostering resilience: What do we know? Principal, 77,18–20.

Yoon, J. S. (2002). Teacher characteristics as predictors ofteacher-student relationships: Stress, negative affect,and self-efficacy. Social Behavior and Personality, 30,485–493. doi: 10.2224/sbp.2002.30.5.485

Date Received: February 10, 2011Date Accepted: September 15, 2011

Action Editors: Elise Cappella, Wendy M.Reinke, and Kimberly E. Hoagwood �

Elisa S. Shernoff, PhD, is a Research Assistant Professor of Psychology in Psychiatry atthe Institute for Juvenile Research at the University of Illinois at Chicago. The overar-ching goal of her work is to expand the focus of mental health practice in urban schoolsto include supporting teacher effectiveness as a mechanism for promoting positiveacademic and behavioral outcomes for students living in urban poverty.

Ane M. Marínez-Lora, PhD, is a Research Assistant Professor of Psychology in Psychi-atry at the Institute for Juvenile Research at the University of Illinois at Chicago. Herresearch focuses on the provision of mental health services, cultural adaptation ofevidence-based interventions, the use of evidence-based practices by clinicians andteachers, and parental school involvement.

Stacy L. Frazier, PhD, is an Associate Professor in the Department of Psychology atFlorida International University. She directs a program of research that examines mentalhealth promotion for families living in urban poverty through capacity building in thenatural, neighborhood settings that support them, including schools, parks, and mentalhealth agencies.

Lara J. Jakobsons, PhD, is a Postdoctoral Fellow at the Institute for Juvenile Research atthe University of Illinois at Chicago. Her research focuses on examining factors relatedto improving children’s academic and behavioral outcomes in high-poverty schools.

Marc S. Atkins, PhD, is a Professor of Psychology and Psychiatry at the University ofIllinois at Chicago and Director of Research at the Institute for Juvenile Research. He hasa long-standing interest in the development of effective mental health services for childrenand families living in high poverty urban communities and has been the recipient ofnumerous grants from the National Institute of Mental Health and private foundations,including a developing center grant from the NIMH to develop and study new models formental health services in schools and communities.

Deborah Bonner, MA, has been a Principal within Chicago Public Schools for the past tenyears in addition to being an Assistant Principal for seven years. Prior to that she was aclassroom teacher for 15 years. She has been an advocate for children and families facingoverwhelming stressors related to urban poverty for over 30 years and is currentlydeveloping a parent literacy program through book clubs designed to enhance parentreading confidence and student reading achievement.


485

Copyright of School Psychology Review is the property of National Association of School Psychologists and its

content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's

express written permission. However, users may print, download, or email articles for individual use.

Documents

Teachers Supporting Teachers in Urban Schools: What Iterative Research Designs Can Teach Us