The effect of teacher coaching with performance feedback on behavior-specific praise in inclusion classrooms.

The effect of teacher coaching with performance feedback on behavior-specific praise in inclusion classrooms. Abstract This study extends teacher coaching with written performancefeedback as a means to increase behavior-specific praise statements(BSPS) in inclusion classrooms at the high school level. Three inclusionmath teachers increased their use of BSPS during mathematicsinstruction. On-task intervals and frequency of BSPS were measuredthroughout the study. The study provides support for teacher coachingwith performance feedback as an effective method for teacher training.Favorable teacher responses indicate it may be an acceptable source ofprofessional development. Future researchers may explore introducingother teaching strategies and feedback on other dependent variables intothe teacher coaching sessions. Limitations are inconclusive findings onthe effectiveness of BSPS to influence student time on-task. Futureresearch should include a more sensitive, systematic measurement ofon-task and disruptive behaviors. High school teachers provide instruction for a diverse and everchanging population (Putnam, Spiegel, & Bruininks, 1995; Villa,Thousand, Nevin, & Listen, 2005). Students with high-incidencedisabilities are enrolled in inclusion classrooms taught by generaleducation teachers (Conderman & Katsiyannis, 2002; Handler, 2006;Reschly & Christenson, 2006) who may be unprepared to manage theacademic and behavioral needs of students with disabilities that mayinterfere with instruction and student learning (Baloglu, 2009; Cwikla,2004; Van Reusen, Shoho, & Barker, 2000). Meanwhile, accountabilityfor annual yearly progress requires teachers to present fast-paced,content-rich instruction to prepare students for end of course exams andgraduation assessments (Mastropieri & Scruggs, 2001; No Child LeftBehind, 2001). Emphasis on teaching grade-level, standards-basedcurriculum to classes with multiple ability levels of students mayreduce teachers' time for planning activities that encourageparticipation from all students. However, instructional strategies toencourage participation of students with disabilities are essential toengage students in the learning process (Bost & Riccomini, 2006;Council for Exceptional Children, 1987; Murray & Pianta, 2007). Inaddition, disruptive behavior due to academic and/or social deficits maybe reduced with basic classroom management strategies incorporated intoteaching (Conroy, Sutherland, Snyder, Al-Hendawi, & Vo, 2009; Gunter& Jack, 1993; Simonsen, Fairbanks, Briesch, Myers, & Sugai,2008). Academic and Behavior Problems are Cyclic A lack of participation and misbehavior are often generated as away for students to avoid task demands, avert failure, and avoid peerembarrassment when failure occurs (Colvin, 2004; Miles & Stipek,2006; Scott, Nelson, & Liaupsin, 2001). To increase studentparticipation, teachers need to acknowledge small successes each dayimmediately and consistently reinforcing student effort (Bost &Riccomini, 2006; Brophy, 1979; Scott et al., 2001). Yet high schoolteachers often lecture, model, ask questions, give directions, andmonitor independent student practice increasing demands on compliantstudents while rarely incorporating motivational techniques orevidence-based instructional strategies into everyday teaching (Carr,Taylor, & Robinson, 1991; Schumaker et al., 2002). When disruptivebehaviors occur teachers often respond with warnings, threats, or officereferrals to gain compliance (Villa et al., 2005). A teacher's reaction to misbehavior may provide attentionunintentionally reinforcing the inappropriate behavior (Shores &Wehby, 1999) or prompting the beginning of a coercive cycle that isdifficult to end (Colvin, 2004). Although there are various reasonsstudents engage in inappropriate behavior during instruction, teachersare able to improve classroom behavior by adjusting when and how theyrespond (Baloglu, 2009; Rathel, Drasgow, & Christie, 2008; Shores& Jack, 1993; Thomas, Becker, & Armstrong, 1968; Van Acker &Grant, 1996). One strategy that may be incorporated into regular,on-going classroom instruction and increase appropriate, active studentparticipation is behavior-specific praise (Brophy 1980; Conroy,Sutherland, Snyder, & Marsh, 2008; Ferguson & Houghton, 1992;Simonsen et al., 2008; Sutherland, Alder, & Gunter, 2003). Behavior-specific Praise Encourages Participation Used consistently, behavior-specific praise statements (BSPS) arean effective teaching strategy that may positively affect classroombehavior by increasing student time on-task, responding, and correctanswers (Ferguson & Houghton, 1992; Sutherland et al., 2003;Sutherland, Wehby, & Yoder, 2002). Behavior-specific praise isapproval with an explanation of the appropriate behavior exhibited(Gunter & Jack, 1993; Sutherland, 2000; Sutherland, Lewis-Palmer,Stichter, & Morgan, 2008). Offered contingent on a student'ssocial or academic action, BSPS are effective when they occurfrequently, in a timely manner, and are related to the student'seffort (Conroy et al., 2009; Simonsen et al., 2008) whether offered toindividuals or class-wide. Kirby and Shields (1972) examined the effect of BSPS on the timespent engaged and the completion of math problems of a seventh gradestudent. Findings indicate that on-task behavior increased from abaseline mean of 47% to an intervention mean of 97% when praised.Likewise, the number of math problems completed increased from a mean of.47 problems per minute to a mean of 1.44 during intervention. Likeother new teaching skills, providing assistance and support to teachersmay increase the use of BSPS (Gunter, Jack, DePaepe, Reed, &Harrison, 1994; Rathel et al., 2008; Shores & Jack, 1993). Teacher Coaching Provides Support for BSPS One strategy to assist teachers with increasing the use of BSPS isteacher coaching with performance feedback. Teacher coaching withperformance feedback has been effective in assisting teachers withincreasing the use of BSPS (Hawkins & Heflin, in press; Reinke,Lewis-Palmer, & Merrel, 2008; Sutherland, Wehby, & Copeland,2000), increasing student opportunities to respond (Sutherland et al.,2003), improving teacher communication with students (Rathel et al.,2008), and improving strategies of effective instruction (Hasbrouck& Christen, 1997; Stichter, Lewis, Richter, Johnson, & Bradley,2006). Teacher coaching works much like coaching athletes (Joyce &Showers, 1982; Maeda, 2001) as it involves: (a) choosing a skill todevelop, (b) reviewing information regarding the skill, (c) discussingthe skill application, (d) practicing the skill, (e) collecting datathrough observations to analyze the effect, and (f) providing feedback(Hendrickson, Sorka, & Gable, 1988; Joyce & Showers, 1982;Maeda, 2001; Rathel et al, 2008). Teacher coaching has been implemented as part of preservice (Rathelet al, 2008), mentoring (Hasbrouck & Christen, 1997), andcollaboration (Stichter et al., 2006) programs using various termsincluding peer-coaching, observation-feedback, collaboration, collegialcoaching, and challenge coaching (Cwikla, 2004; Hendrickson et al.,1988; Maeda, 2001; Showers & Joyce, 1996; Stichter et al., 2006;Sutherland et al., 2000). The goal is to provide non-evaluative feedbackto improve teaching practices by increasing individual's awarenessof personal interactions while teaching (Garmston, 1987; Sutherland,2000). Discussion is based on data from the observation, providing anopportunity for reflection and interchange between colleagues (Joyce& Showers, 1982; Stichter et al., 2006). Increasing Teacher Use of BSPS Sutherland et al. (2000) examined the effect ofobservation-feedback on the rate of a teacher's BSPS in aself-contained elementary class for students with emotional andbehavioral disorders (E/BD). Observations took place three days a weekduring social-skills instruction. The intervention was verbal feedbackon the observed rate of BSPS for each 15-minute session. After theinitial baseline data were collected, the teacher was provided with thebenefits of BSPS and encouraged to meet a goal of six BSPS during each15-minute session. Prior to each observation the observer reminded theteacher of the goal and provided examples of BSPS. After eachobservation the observer met with the teacher to praise his use of BSPS,report the number of BSPS observed, and repeat examples of BSPS heardduring the lesson. The mean rate of BSPS increased from 1.3 to 6.7 per15-minutes. The mean rate then dropped to 1.7 per 15-minutes during thewithdrawal phase, increasing to a mean rate of 7.8 when the observationfeedback intervention was reinstated. Student on-task intervals wereobserved with mean percentages increasing from 48.7% to 85.6% during thefirst intervention phase, decreasing to 62.2% during withdrawal, andincreasing to 83.3% during the final intervention phase. The increasedrates of BSPS did not maintain once the intervention was withdrawn. Reinke et al. (2008) replicated and extended the research onobservation-feedback to increase BSPS with general education teachersacross elementary school grades (i.e., first, second, and fifth grades)during daily mathematics instruction. Additional components wereincorporated including a classroom check-up to evaluate environment andvisual feedback. Mean rates of BSPS increased with the classroomcheck-up and self-monitoring, but increased more with the classroomcheck-up and visual feedback. Visual feedback included a graphindicating rates of BSPS, nonbehavior-specific praise statements,reprimands, and disruptions. Disruptions decreased in two of theclassrooms, while the other two classes were inconsistent. Maintenanceobservations a month after the final intervention showed increased ratesof praise and decreased disruptions continued across all fourclassrooms. Using video self-modeling (VSM) and visual performance feedback,Hawkins and Heflin (in press) extended the research on the rate of BSPSto a self-contained program serving high school students with E/BD.Prior to each observation, the observer met with the teacher for 10minutes to provide feedback of the dependent variables presented on aline graph, to review an edited version of the video from the previoussession, to provide examples of BSPS, and to praise the teacher forspecific components of the BSPS viewed on the video that emphasized thedesirable student behavior. During the study, all three teachers'mean rates of BSPS increased steadily. Although one teacher was notobserved providing any BSPS during the maintenance observations theother two teachers provided their highest rates during maintenanceobservations. These studies provided teacher support for learning and practicingBSPS using teacher coaching prior to the observations and various modelsof feedback following observations. Further extension of teachercoaching with performance feedback research to increase BSPS iswarranted to investigate the effectiveness for supporting inclusionteachers at the high school level (Colvin, 2004; Mastropieri &Scruggs, 2001). To formulate a teacher coaching intervention more usefulfor colleagues to implement within a school as opposed to having anoutside observer, it will be useful to examine less time-consumingmethods for teacher coaching (e.g., a fixed but not daily schedule). The purpose of this study was to extend the research on teachercoaching and BSPS to the high school level, specifically in generaleducation collaborative inclusion math classrooms with students with andwithout high incidence disabilities including those identified withE/BD, learning disabilities (LD), and other health impairments (OHI).The study examined (1) the effect teacher coaching with writtenperformance feedback had on the frequency of teachers' BSPS withhigh school students, and (2) the effect of BSPS on student on-taskbehavior. Teachers were observed to determine the level of maintenanceafter teacher coaching sessions ended, and social validity was assessedusing a survey at the conclusion of the intervention. Method Participants and Setting This study occurred in three inclusion suburban high schoolclassrooms in a metropolitan area of a southeastern city (see Table 1for school demographics). The teacher participants included two femaleand one male teacher; two general education teachers and one specialeducation teacher (see Table 2 for teacher demographics). The schooloperated by block scheduling with classes offered on alternate days(i.e., A/B days). These three teachers co-taught three different ninthgrade Math I inclusion classes for students repeating ninth grade due tofailing math the previous school year (see Table 1 for classdemographics). Although the appropriateness of the math curriculum formeeting the needs of these students is unknown, the students wereassigned to these classes as a second attempt to earn math creditrequired for graduation. In an effort to provide additional support tothe students, rather than following the standard block scheduling, mathclasses were taught alternating the core class with an additional classto provide students 90 minutes of instruction daily using the state mathcurriculum. Although students had math everyday, they alternated betweenteachers therefore sessions were conducted on alternating days.Table 1School and Classroom DemographicsSuburban High School with 2055 Students: 233 Students with Disabilities(109 ninth grade)Grade Number Age Gender Ethnicity Disabilities (1)9 786 14-17 51%F 51% African 61 LD 49% M American 19 OHI 19% Caucasian 13 ED 27% Hispanic 2 AU 3% Asian 1 DHH 2 VI 4 MID 3 MOID 4 SID 9th Grade Students Retained to Repeat MathematicsClass 1 20 15-16 45% Female 2 LDKelly 55% Male American 2 OHI 25% Caucasian 15% HispanicClass 2 24 15-16 50% Female 33% African 3 LDJaime 50% Male American 1 ED 17% Caucasian 1 OHI 50% HispanicClass 3 18 15-17 39% Female 44% African 4 LDChris 61% Male American 3 ED 22% Caucasian 33% HispanicNote: 1: LD = learning disabilities; OHI = other health impaired; ED =emotional disorder; AU = autism; DHH = deaf and hard of hearing; VI =visually impaired; MID = mild intellectual disability; MOID = moderateintellectual disability; SID = severe intellectual disabilityTable 2 Teacher DemographicsParticipant Ethnicity Highest Degree Certification Years of ExperienceKelly Caucasian Masters Secondary 8 MathJaime Hispanic Bachelors Secondary 2 MathChris African Masters Special 2 American Education Mathematical units taught over the 14 weeks of observation includedgeometry, algebraic equations, probability, and graphs. Observationsoccurred during 15 minutes of active instruction of the core classincluding a variety of activities such as review of the previousday's lesson, introduction of new material, whole group instructionwith guided practice, content review for upcoming tests, group work(i.e., stations), independent practice, and checking answers oncompleted work. Dependent Variables and Data Collection The two dependent variables measured in this study werebehavior-specific praise statements (BSPS) and on-task behavior. Thesevariables were measured during 15-minute observations that occurred atthe same time each day the class met, provided the participants(teachers) were present in class. Behavior-specific praise statements. Behavior-specific praisestatements were defined as statements of approval provided to a studentor students by a teacher that included a description of the behaviorbeing reinforced (Sutherland et al., 2000). For example, when reviewingmath problems with the student the teacher may have said, "Thestrategies you used for problems 3 and 4 are excellent. You haveincluded all of the steps in a neat and easy to understand process. Goodwork!" This statement told the student what was done well. Anotherexample is when a teacher told the class, "Thank you for cominginto class quietly and having your homework on your desk ready toreview." The students knew exactly which behaviors pleased theteacher. Behavior-specific praise statements did not include nonspecificpraise such as "Good," "Nice job," or "Welldone!" Behavior-specific praise statements were measured usingevent recording and reported as number per session. On-task behavior. On-task behavior was defined as engagement withthe learning process and occurred when a student focused on the teacheror materials, participated in the lesson by responding to and askingquestions, and performed specific activities/assignments related to thelesson (Moore, 1983). On-task behavior was measured using momentary timesampling at one-minute intervals and reported as a percentage ofintervals on-task. On-task behavior was measured by observing a randomlyselected student and then observing the next student (a total of 15different students observed per session). Design and Independent Variable A multiple baseline across teachers design was used to examine theeffect of teacher coaching with written performance feedback for highschool teachers on the number of BSPS and opportunities to respond (OTR)and the percentage of on-task student behavior (Alberto & Troutman,2009; Kennedy, 2005). Baseline. This study was conducted on alternating days during15-minutes of interactive math instruction. Each observation includedinstruction, guided and independent practice, and review of studentanswers. The general education teachers conducted the class-wideinstruction while the special education teacher monitored and assistedindividual students as needed. Only one teacher in each class wasobserved. During independent practice students were allowed to workindependently or with peers. All three classes experienced off-task anddisruptive behaviors which were of concern to the teachers. Studentswere observed to wander within the classroom, come and go from theclassroom, talk among themselves, shout across the classroom to peers,and use their cell phones which negatively affected their overallparticipation. 'Teacher coaching intervention. Once consistency wasdemonstrated by a minimum of five consecutive observations with noevidence of BSPS during baseline, the first teacher was trained in theintervention of teacher coaching. The teachers were initially trainedduring an independent 45-minute training session which included (a) apower point presentation to define teacher coaching and BSPS, (b) therationale and benefits of teacher coaching and BSPS, (c) examples ofBSPS, (d) a discussion of teacher coaching (i.e., goal setting, preobservation conferencing, and written post-observation feedback), and(e) an opportunity for questions. At the end of the baseline condition,the teacher was told how many BSPS were observed during baseline. Afterdiscussion regarding how to provide BSPS during math instruction, theteacher set a goal for the number of BSPS to deliver within each15-minute observation. Following the initial training, the researcher provided a 5-minuteteacher coaching conference prior to every third intervention session.The researcher reminded the teacher of the BSPS goal for the session,provided an example BSPS, and gave the teacher an opportunity to ask anyquestions regarding BSPS. After every 15-minute intervention session,regardless of whether there was a teacher coaching conference, writtenperformance feedback was left in a folder on the teacher's desk.Performance feedback included the first two BSPS recorded, the totalnumber of BSPS observed, and BSPS from the researcher for theteacher's use of BSPS. Maintenance. Teachers were observed 2 weeks and 3 weeks after theremoval of the teacher coaching intervention to determine if their useof BSPS were maintained. No teacher- coaching took place prior to theobservation and no written performance feedback was provided afterwards. Social Validity. Social validity was evaluated using a modifiedversion of the Treatment Acceptability Rating Form-Revised (TARF-R:Reimers & Wacker, 1988). Teachers were given the modified TARF-Rwith the written performance feedback on the last day of intervention.They were asked to complete the form individually before the firstscheduled maintenance observation. The TARF-R consisted of 17 questionswith space for additional comments. Fifteen questions assessed thepracticality and effectiveness of the teacher coaching intervention andof the use of BSPS as a classroom management strategy. The questionswere scored using a 7-point Likert scale. Teacher perception regardingstudy participation was assessed using two open-ended questions. Results The purpose of this study was to determine if teacher coaching withwritten performance feedback would effect the number of BSPS duringinstruction and if increased BSPS positively affected on-task behavior(see Figure 1). During baseline, Kelly delivered no BSPS. At the initial teachercoaching training session, Kelly set a goal of 10 BSPS for each15-minute observation. The intervention BSPS mean was 9.7 (range, 4 to12) and during maintenance, across two sessions the BSPS mean was 9(range, 8 to 10). In terms on-task behavior, baseline to interventionmeans showed a slight decline with a return to baseline level duringmaintenance (baseline M = 51% to intervention M = 41% to maintenance M =50%). During baseline, Jaime delivered 2 BSPS on one occasion, M = .33(range, 0 to 2). At the initial teacher coaching training session, Jaimeset a goal of 7 BSPS for each 15-minute observation. The interventionBSPS mean was 8.5 (range, 6 to 11) and during maintenance, across twosessions the BSPS mean was 9.5. In terms of on-task behavior, baselineto intervention means showed a decrease in on-task behavior with areturn to baseline during maintenance (baseline M = 60%. to interventionM = 45.7% to maintenance M = 63.3%). During baseline, Chris delivered no BSPS. At the initial teachercoaching training session, Chris set a goal of 5 BSPS for each 15-minuteobservation. The intervention BSPS mean was 3.75 (range, 3 to 5).On-task behavior increased from baseline M = 44% to intervention M =51%. Due to the change in teaching structure as the school year came toan end no maintenance data were collected for Chris. Fidelity Fidelity was assessed on the accuracy of the teacher coachingintervention by two trained graduate assistants, each of whom wastrained to mastery on all study fidelity and dependent variables. Eachteacher's initial teacher coaching training session was implementedwith 100% accuracy per a checklist with the training components. Toassess treatment fidelity of the remaining teacher coaching interventionsessions a fidelity checklist with ten components was used.Interobserver agreement (IOA) was assessed on fidelity of teachercoaching sessions for each teacher using point-by-point agreement bydividing the total number of agreements by the total number ofagreements plus disagreements multiplied by 100%. For Kelly, fidelitywas assessed for 100% of the teacher coaching intervention sessions,with fidelity at 99% (range, 90% to 100%); 60% of sessions were assessedwith IOA at 100%. For Jaime, fidelity was assessed for 100% of theteacher coaching intervention sessions, with fidelity at 95% (range, 86%to 100%); 75% of sessions with IOA at 100%. For Chris, fidelity wasassessed for 33% of the teacher coaching intervention sessions, withfidelity at 100%; 33% of sessions were assessed with IOA at 100%. Inter-Observer Agreement Two trained graduate assistants conducted interobserver agreementassessments for all teachers during classroom observations across allphases of the study. For BSPS, total agreements of observed behaviorswere calculated by dividing the smaller number by the larger number ofobserved behaviors and multiplying by 100%. For the percentage ofintervals of on-task behavior, point-by-point agreement was calculatedby dividing the number of agreements by the number of agreements plusdisagreements and multiplying by 100% (Kennedy, 2005). For Kelly, IOA was assessed for 42% of classroom observations, withBSPS at 97% (range, 82% to 100%) and percentage of on-task intervals at92.56% (range, 87% to 100%). For Jaime, IOA was assessed for 47% ofclassroom observations, with BSPS at 96.44% (range, 83% to 100%) andpercentage of on-task intervals at 91.33% (range, 87% to 100%). ForChris, IOA was assessed for 44% of classroom observations, with BSPS at97.5% (range, 80% to 100%) and percentage of on-task intervals at 93.37%(range, 80% to 100%). Social Validity The results of the TARF-R indicate that the teacher coaching withperformance feedback was easy to understand and acceptable for theseinclusion teachers. One teacher stated it was difficult to be observedwhen students were behaving inappropriately, but it was nice to havepositive feedback for a job well done. All three teachers reported thatBSPS are effective in promoting positive student behavior regardingacademics. One teacher wrote that the students seemed to take moreinitiative and participate in the lessons during intervention. There wasan increase in the students' willingness to offer math solutionsand explanations, as well as to solve problems at the board in front ofthe class. Students noticed when others were recognized for doingsomething well and then they also behaved in ways that would berecognized. Discussion The results of this study extend the research of teacher coachingto general education math inclusion teachers for students with andwithout high incidence disabilities. The findings are consistent withprevious studies indicating that teacher coaching with performancefeedback can have a direct and immediate impact on teachers' use ofBSPS (Hawkins & Heflin, in press; Rathel et al., 2008; Sutherland,2000). The methods of this study differed from previous studies in twoways. First, after initial training, teacher coaching was only providedprior to every third session with written performance feedback followingeach session. Teacher coaching was reduced to every third session ratherthan every session as in previous studies to examine efficacy. Based onblock scheduling, every third session allowed one teacher coachingsession to occur each week. Second, feedback was only provided for BSPSeven though on-task behavior as well as opportunity to respond (OTR)data was collected. Data were collected on OTR in an attempt to extendthe findings reported by Sutherland, Wehby, and Yoder (2002). Sutherlandet al. (2002) found when BSPS increased OTR also increased and suggestedfurther research in manipulation of these variables. However, uponexamining the data from this study the OTR data collected suggested nodiscernable relation to the increase in BSPS. Likewise, the effects ofincreased BSPS for on-task behavior were inconclusive. Several factorsmay contribute to the lack of change in OTR. One factor may be due tobaseline frequencies. For example, two teachers had a high frequency ofOTR during baseline (M = 22 and 28 per 15-minutes) and their frequencyof OTR did not increase throughout the intervention; however, oneteacher had a low frequency of OTR during baseline (M = 8.7 per15-minutes) and increased OTR during intervention (M = 30 per15-minutes). Future researchers may further examine the relation betweenBSPS, OTR, and on-task behavior at the high school level. A secondpossible explanation may be due to the lack of feedback on theteachers' performance related to OTR and on-task behavior; or apossible explanation presented by Joyce and Showers (1982) that whileconcentrating on learning to integrate a new skill other skills maytemporarily suffer. Future studies of teacher coaching withobservation-feedback should investigate providing written feedback oneach of the dependent variables (i.e., BSPS, OTR, and on-task behavior)to investigate if feedback on more variables would lead to more teacherchange. Additionally, the lack of change in student on-task behavior whenBSPS increased may be due to the lack of sensitivity of thedata-collection method. On-task behavior was measured for a randomsample of 15 students per session. Although no specific data werecollected on how BSPS were distributed, anecdotal records indicate thatBSPS were provided to a range of students for both academic and socialbehaviors. In addition, BSPS was provided to the entire class onoccasion as well as to individual students both publicly and privately.A more accurate snapshot of student on-task behavior may be measured inother ways. For example, future researchers might consider identifyingspecific students to observe consistently throughout the study usingpartial interval recording with reduced interval lengths (i.e., 30seconds). This will allow frequent and repeated recording of specificstudents' behavior throughout the study to detect change inindividuals as opposed to the random one-time snapshots of 15 differentstudents during each observation as was collected during this study. Inaddition, data could be collected in regard to whether BSPS weredelivered publicly or privately as well as which students received BSPS. In addition to not providing feedback for OTR and on-task behavior,and a lack of sensitivity in measuring on-task behavior, anotherconsideration is the uniqueness of each tier related to teachers'teaching style and behavior management practices. There were anecdotalobservable differences between each class related to student-teacherinteractions during instruction. For example, Kelly modeled mathprocedures, offered frequent OTR regarding computation, and repeatedlydirected students to be quiet during instruction, waiting until talkingstopped to resume teaching while Jaime maintained a brisk pace, invitedvolunteers to demonstrate solutions on the board but rarely addressednoise levels. Chris was observed teaching individuals or small groups,talking students through problem-solving, and redirecting students tothe lesson. When students were uncooperative or disruptive, he usedhumor, talked quietly to individuals, removed students from the room,wrote office discipline referrals, or called parents. Future researchersshould consider teaching styles and behavior management practices whenselecting participants to allow for systematic replication across tiersin a multiple baseline design. Two other variables to consider in future research are disruptivebehavior (Sutherland et al., 2000) and work-completion (Kirby &Shields, 1972). Based on anecdotal records from all three classes, thelevel of disruptive behavior declined during intervention with areduction in students throwing items during instruction, usingelectronics, and walking around and/or leaving the classroom. Besidesthe observed reduction of disruption, the researchers noted throughoutthe intervention students complimenting peers when they received BSPSand telling the teachers "See I'm working too!" or"I got that one right!" Therefore, rather than focusing onlyon student time on-task behavior, future researchers may want to collectdata on disruptive behaviors, work completion, accuracy of work, andperformance on quizzes and/or tests. As a final point, many scheduled sessions did not occur for variousreasons affecting tier entry. For example, of the 35 scheduledobservations, five early sessions were missed due to schedule changesbased on inclement weather and district testing. Jaime missed anadditional two days because of independent work and student conferences,and Chris missed an additional 11 days because of student discipline,health-related absences, and program changes. The study concluded sevensessions earlier than scheduled because the team changed theirinstructional approach to prepare for the end of course test reducingthe number of possible intervention sessions for Chris. Futureresearchers may consider beginning data collection earlier in the schoolyear to allot more time for unexpected interruptions in data collection.However, these disruptions in the study sessions are an artifact ofconducting studies within applied settings, and may lend itself togeneralizable statements of the effectiveness of coaching on BSPS. The increased BSPS for all three teachers using teacher coachingwith written performance feedback and the positive responses by theteachers on the TARF-R indicate this intervention may be useful inassisting high school collaborative inclusion teachers in improvingclassroom strategies. Considering the confounding variables experiencedthroughout this study, the immediate improvement of integrating BSPSthrough teacher coaching with performance feedback suggest that otherpractical teaching strategies could be learned using this method.Further research regarding teacher coaching at the high school level maysupport this intervention as professional development for theimplementation and improvement of a variety of teaching strategies toincrease student engagement (Cwikla, 2004; Hasbrouck & Christen,1997; Hendrickson et al., 1988; Joyce & Showers, 1982; Villa et al.,2005). An extension of the literature may include allowing teachers tochoose which strategies they would like to learn through teachercoaching. References Alberto, P. A., & Troutman, A. C. (2009). Applied behavioranalysis for teachers (8th ed.). Upper Saddle River, NJ: PearsonEducation. Baloglu, N. (2009). Negative behavior of teachers with regard tohigh school students in classroom settings. Journal of InstructionalPsychology, 36, 69-78. Bost, L. W., & Riccomini, P. J. (2006). Effective instruction:An inconspicuous strategy for dropout prevention. Remedial and SpecialEducation, 27, 301-311. Brophy, J. E. (1979). Teacher behavior and its effects.(IRT-OP-25). Lansing, MI: Michigan State University Institute forResearch on Teaching. Brophy, J. E. (1980). Recent research on teaching. (IRT-OP-40).Lansing, MI: Michigan State University Institute for Research onTeaching. Carr, E. G., Taylor, J. C., & Robinson, S. (1991). The effectsof severe behavior problems in children on the teaching behavior ofadults. Journal of Applied Behavior Analysis, 24, 523-535. Colvin, G. (2004). Managing the cycle of acting-out behavior in theclassroom. Eugene, OR: Behavior Associations. Conderman, G., & Katsiyannis, A. (2002). Instructional issuesand practices in secondary special education, Remedial and SpecialEducation, 23, 169-179. Conroy, M. A., Sutherland, K. S., Snyder, A., Al-Hendawi, M., &Vo, A. (2009). Creating a positive classroom atmosphere: Teachers'use of effective praise and feedback. Beyond Behavior, 18, 18-26. Conroy, M. A., Sutherland, K. S., Snyder, A. L., & Marsh, S.(2008). Classwide interventions: Effective instruction makes adifference. Teaching Exceptional Children, 40, 24-30. Council for Exceptional Children (1987). For effective instruction:Working with mildly handicapped students. Academy Participant'sManual. Reston, VA: Council for Exceptional Children. Cwikla, J. (2004). Less experienced mathematics teachers reportwhat is wrong with their professional support system. Teachers andteaching: Theory and practice, 10, 181-197. Ferguson, E., & Houghton, S. (1992). The effects of contingentteacher praise, as specified by Canter's Assertive Disciplineprogramme, on children's on-task behavior. Educational Studies, 18,83-94. Garmston, R. J. (1987). How administrators support peer coaching.Educational Leadership, 44, 18-26. Gunter, P., & Jack, S. (1993). Lag sequential analysis as atool for functional analysis of student disruptive behavior inclassrooms. Journal of Emotional and Behavioral Disorders, 1, 138-152. Gunter, P. L., Jack, S. L., DePaepe, P., Reed, T. M., &Harrison, J. (1994). Effects of challenging behaviors of students withEBD on teacher instructional behavior. Preventing School Failure, 38,35-39. Handler, B. R. (2006). Two acts, one goal: Meeting the sharedvision of No Child Left Behind and Individuals with DisabilitiesEducation Improvement Act of 2004. The Clearing House, 80, 5-8. Hasbrouck, J. E., & Christen, M. H. (1997). Providing coachingin inclusive classrooms: A tool for consulting teachers. Intervention inSchool & Clinic, 32, 6-13. Hawkins, S., & Heflin, L. J. (in press). Increasing secondaryteachers' behavior-specific praise using a video self-modeling andvisual performance feedback intervention. The Journal of PositiveBehavioral Interventions. Hendrickson, J. M., Sroka, K., & Gable, R. A. (1988). Peercoaching: A specific approach for improving teacher performance andtrainee competence. Paper presented at annual Florida Staff DevelopmentLeadership Conference. St. Petersburg, FL. ED 303942 EC 212138. Joyce, B., & Showers, B. (1982). The coaching of teaching.Educational Leadership, 40, 4-10. Kennedy, C. H. (2005). Single-case designs for educationalresearch. Boston, MA: Pearson Education. Kirby, F. D., & Shields, R. (1972). Modification of arithmeticresponse rate and attending behavior in a seventh grade student. Journalof Applied Behavior Analysis, 5, 70-84. Maeda, J. K. (2001). Teacher coaching in physical education: Areview. Physical Educator, 58, 140-150. Mastropieri, M. A., & Scruggs, T. E. (2001). Promotinginclusion in secondary classrooms. Learning Disabilities Quarterly, 24,265-274. Miles, S. B., & Stipek, D. (2006). Contemporaneous andlongitudinal associations between social behavior and literacyachievement in a sample of low-income elementary school children. ChildDevelopment, 77, 103-117. Moore, J. E. (1983, October). Assessing time on task: Measurementproblems and solutions. Paper presented at the Joint Meeting ofEvaluation Network and Evaluation Research Society, Chicago, IL. Murray, C, & Pianta, R. C. (2007). The importance ofteacher-student relationships for adolescents with high incidencedisabilities. Theory into Practice, 46, 105-112. No Child Left Behind Act of 2001, Pub. L. No. 107-110, 115 Stat.1425 (2002). Retrieved July 16, 2009, fromhttp://www.ed.gov/policy/elsec/guid/states/index.html. Putnam, J. W., Spiegel, A. N., & Bruininks, R. H. (1995).Future directions in education and inclusion of students withdisabilities: A Delphi investigation. Exceptional Children, 61, 553-576. Rathel, J. M., Drasgow, E., & Christie, C. A. (2008). Effectsof supervisor performance feedback on increasing pre-serviceteachers' positive communication behaviors with students withemotional and behavioral disorders. Journal of Emotional and BehavioralDisorders, 16, 67-77. Reimers, T. M., & Wacker, D. P. (1988). Parent's ratingsof the acceptability of behavioral treatment recommendations made in anoutpatient clinic: A preliminary analysis of the influence of treatmenteffectiveness, Behavioral Disorders, 14, 7-15. Reinke, W. M., Lewis-Palmer, T., & Merrel, K. (2008). Theclassroom check-up: A classwide teacher consultation model forincreasing praise and decreasing disruptive behavior. School PsychologyReview, 37, 315-332. Reschly, A. L., & Christenson, S. L. (2006). Prediction ofdropout among students with mild disabilities: A case for the inclusionof student engagement variables. Remedial and Special Education, 27,276-292. Schumaker, J. B., Deshler, D. D., Bulgren, J. A., Davis, B., Lenz,B. K., & Grossen, B. (2002). Access of adolescents with disabilitiesto general education curriculum: Myth or reality? Focus on ExceptionalChildren, 35, 1-16. Scott, T. M., Nelson, C. M, & Liaupsin, C. J. (2001). Effectiveinstruction: The forgotten component in preventing school violence.Education and Treatment of Children, 24, 309-322. Shores, R. E., & Jack, S. L. (1993). Classroom interactions ofchildren with behavior disorders. Journal of Emotional and BehavioralDisorders, 1, 27-40. Shores, R. E., & Wehby, J. (1999). Analyzing the classroomsocial behavior of students with EBD. Journal of Emotional andBehavioral Disorders, 7, 194-199. Showers, B., & Joyce, B. (1996). The evolution of peercoaching. Educational Leadership, 53, 12-16. Simonsen, B., Fairbanks, S., Briesch, A., Myers, D., & Sugai,G. (2008). Evidence-based practices in classroom management:Considerations for research to practice. Education and Treatment ofChildren, 31, 351-380. Stichter, J. P., Lewis, T J., Richter, M., Johnson, N. W., &Bradley, L. (2006). Assessing antecedent variables: The effect ofinstructional variables on student outcomes through in-service and peercoaching professional development models. Education and Treatment ofChildren, 29, 665-692. Sutherland, K. S. (2000). Promoting positive interactions betweenteachers and students with emotional/behavioral disorders. PreventingSchool Failure, 44, 110-116. Sutherland, K. S., Alder, N., & Gunter, P. L. (2003). Theeffect of varying rates of opportunities to respond to academic requestson the classroom behavior of students with EBD. Journal of Emotional andBehavioral Disorders, 11, 239-248. Sutherland, K. S., Lewis-Palmer, T., Stichter, J., & Morgan, P.L. (2008). Examining the influence of teacher behavior and classroomcontext on the behavioral and academic outcomes for students withemotional or behavioral disorders. The Journal of Special Education, 41,223-233. Sutherland, K. S., Wehby, J. H., & Copeland, S. R. (2000).Effects of varying rates of behavior-specific praise on the on-taskbehavior of students with EBD. Journal of Emotional and BehavioralDisorders, 8, 2-9. Sutherland, K. S., Wehby, J. H., & Yoder, P. J. (2002).Examination of the relationship between teacher praise and opportunitiesfor students with EBD to respond to academic requests. Journal ofEmotional and Behavioral Disorders, 10, 5-13. Thomas, D. R., Becker, W. C., & Armstrong, M. (1968).Production and elimination of disruptive classroom behavior bysystematically varying teacher's behavior. Journal of AppliedBehavior Analysis, 1, 35-45. Van Acker, R., & Grant, S. H. (1996). Teacher and studentbehavior as a function of risk for aggression. Education and Treatmentof Children, 19, 316-335. Van Reusen, A. K., Shoho, A. R., & Barker, K. S. (2000). Highschool teachers' attitudes toward inclusion. High School Journal,84, 7-21. Villa, R. A., Thousand, J. S., Nevin, A., & Liston, A. (2005).Successful inclusive practices in middle and secondary schools. AmericanSecondary Education, 33, 33-52. Ysseldyke, J. E., Christenson, S. L., & Thurlow, M. L. (1987).Instructional factors that assess student achievement: An integrativereview [Monograph No. 7]. University of Minnesota InstructionalAlternatives Project. Ellen L. Duchaine, Kristine Jolivette, Laura D. Fredrick GeorgiaState University Correspondence to Ellen L. Duchaine, Georgia State University,Dept. of Educational Psychology and Special Education, PO Box 3979,Atlanta, GA 30302.