In U.S. public schools, disruptive behavior (e.g., noncompliance, fighting) is the primary reason for disciplinary referrals, including suspension and expulsion (Gregory & Weinstein, 2008; McClay, 2019; Mendez & Knopf, 2003). Disruptive behavior problems in childhood are associated with a range of adverse outcomes through adolescence and beyond (Barnert et al., 2021; Dishion & Patterson, 2006; Magnusson & Laftman, 2019). Childhood disruptive behavior problems are linked to poor academic achievement (e.g., heightened risk of school failure or dropout), association with deviant peers, and disciplinary referrals at school during adolescence (Deighton et al., 2018; McEvoy & Welker, 2000; Reinke et al., 2008; Vitaro et al., 2018), as well as higher rates of unemployment and criminal involvement in adulthood (Barnert et al., 2021; Border et al., 2018; Colman et al., 2009; Magnusson & Laftman, 2019; Mohr-Jensen & Steinhausen, 2016). Given the negative trajectory of childhood disruptive behavior, effective intervention efforts have broader implications beyond the present; intervention programs could help shape more favorable life outcomes for these youth in the long term.

Schools are optimal settings for delivering interventions for disruptive behavior because they eliminate key barriers (e.g., lack of transportation, limited financial resources) that often prevent youth from receiving the services they need (Atkins et al., 2017). However, school-based interventions for disruptive behavior are limited in several ways. First, they are often time and resource intensive, and on average, show small effects for externalizing behaviors (Barnes et al., 2014; Durlak et al., 2011; Eiraldi et al., 2016; Waschbusch et al., 2019; Wilson & Lipsey, 2007). Second, like most psychosocial treatments for problem behaviors, school-based interventions typically position youth as treatment recipients – i.e., none of these interventions explicitly position youth as leaders or otherwise competent, knowledgeable individuals that have the potential to create positive change in others.

Experimental work with non-clinical samples shows the motivational benefits of helping others (Dunn et al., 2008; Eskreis-Winkler et al., 2018; Eskreis-Winkler et al., 2019). For example, a recent study by Eskreis-Winkler et al. (2018) found that individuals randomized to be advice “givers” reported increased motivation, prosocial behavior, and overall well-being compared with individuals assigned to be advice “receivers.” These findings were consistent across different advice categories (e.g., financial, interpersonal, health, work) and age groups (middle school students and adults). The authors hypothesized that these results were in part due to larger gains in self-confidence among advice givers compared with advice receivers post-intervention.

Additional work shows similar effects for youth who take on formalized helping roles in “real-world” contexts. For example, youth who engage as peer mentors and tutors often show significant improvements in self-confidence, personal growth, social skills, and interpersonal relationships (e.g., Beltman & Schaeben, 2012; Coyne-Foresi & Nowicki, 2021). Furthermore, research on “peer support” interventions indicates that peer leaders (compared to support group members) show higher levels of problem-solving and school connectedness at postintervention – which suggests that giving help can yield greater benefits than receiving help (Ellis et al., 2009).

The benefits of helping appear to extend to criminal offending populations as well. For example, Hanniball et al. (2019) found that both delinquent youth and adult ex-offenders who were randomized to a prosocial helping condition reported greater positive affect compared with those assigned to a personal benefit condition. Another qualitative study found that incarcerated adults who were assigned to act as caretakers for inmates with mental or physical impairments showed improved relationships with prison staff and reduced levels of self-reported violent or aggressive behaviors. Participants cited the “sense of purpose” and “meaning” they gained from being caretakers as motivation to engage in fewer antisocial behaviors while in prison (Einat, 2017). Relatedly, a study of previously incarcerated adults found that those with stronger “helper” orientations had higher levels of psychological well-being, lower levels of pro-criminal attitudes, and lower expectations of recidivism (LeBel, 2007).

In this study, we sought to extend this work to disruptive youth in middle-school contexts by developing an intervention that addresses some limitations of many existing school-based interventions. We piloted Peer Coach Training (PCT; Huey & Galbraith, 2020), a brief, strengths-based intervention to remediate disruptive behavior in middle school contexts. PCT deemphasizes the youth’s existing problems and focuses instead on training youth to act as “coaches” to help their peers engage in prosocial behaviors. PCT is modeled on Ross and McKay’s (1976) Peer Therapist Program, which encouraged females in the juvenile justice system to help their peers by serving as informal therapists to those struggling with disruptive behavior. Ross and McKay found that the Peer Therapist Program was more effective than alternative interventions at reducing recidivism (1976). They argued that by labeling these girls as “therapist” and persuading them to act as such, the girls came to view themselves as prosocial change agents rather than as troublesome youth. Our PCT intervention adopted Ross and McKay’s (1976) general approach, while integrating evidence-based strategies from the peer mentoring (e.g., DeMarco, 1993; Raposa et al., 2019), social skills training (e.g., Dryburgh et al., 2020; Jackson et al., 1983), and behavior modification (e.g., Axelrod & Hall, 1991; Azrin & Besalel-Azrin, 1999; Kazdin, 2001; Martin & Pear, 2019) literatures.

Mixed methods were used to test the feasibility and acceptability of PCT. We used a pre-post-follow-up design with assessments at baseline, posttreatment, and three-month follow-up. Intervention acceptability was assessed with surveys and qualitative interviews with participating students and teachers. We predicted that youth who received PCT would show decreases in externalizing behaviors at posttreatment, and that these gains would be largely maintained at three months post intervention. In accordance with previous findings showing high satisfaction ratings for strengths-based interventions (e.g., Craig & Furman, 2018; Yuen et al., 2020), we anticipated that the strengths-based orientation of PCT would appeal to students and teacher informants. As such, we predicted that quantitative and qualitative intervention acceptability data would show that youth and teachers have positive impressions of PCT – i.e., they find it to be a satisfying and acceptable intervention.

Method

Participants and Recruitment

Participants were nine 7th and 8th grade students from a public middle school located in a low-income, urban setting in Los Angeles County. Of the nine participating youth, 77.8% were male, and the average age was 12.4 years (SD = 0.5). 77.8% identified as Latinx, and 22.2% as Black/African American. Over half (55.5%) spoke English as a second language. A third (33.3%) of the youth endorsed lifetime gang involvement, and 22.2% of the sample reported that they were currently gang-involved.

PCT participants were selected by the Assistant Principal, who was asked to refer students who exhibited the most extreme (either in frequency or type) disruptive behavior in classroom contexts. For eligibility, youth must have received at least one disciplinary referral for disruptive behavior between the first day of school and recruitment, which began approximately one month into the semester. However, youth who were unable to speak or understand English proficiently were ineligible, as we only had the capacity to lead the intervention in English.

Procedures

Ten eligible youth received a take-home consent form that explained the study and expectations for participation, and informed consent was completed by their caregivers. After the Assistant Principal received signed consent forms, undergraduate research assistants (RAs) met with the youth during the last class period of school to complete the assent and baseline assessment. During the assent process, youth provided names of each of their academic teachers and gave permission for investigators to contact them for assessment data. We successfully engaged nine of the original ten youth referred to our study.

From each student’s list of classroom teachers, one was randomly selected and contacted to complete youth assessments. Each teacher was first sent an email with an information sheet describing the nature of the PCT program and was asked to participate. If the teacher did not respond after 24 hours, the first author visited the teacher to explain the study and request their involvement. If no contact was made after 48 hours (via e-mail, phone call, and school visit) or the teacher declined before that time, another teacher was randomly selected from the remaining teachers on the student’s list (and contacted in the same fashion) until a teacher agreed to participate. After agreeing, each teacher completed a consent form and baseline assessment.

Two groups received PCT, with 4–5 participants in each group. Youth were assigned to each group by grade level (i.e., 7th graders in one group, 8th graders in another group) to maximize attendance, as 7th and 8th grade students at this school often had scheduling conflicts that were specific to grade level (e.g., state-wide exams). Group sessions for the first PCT cohort were co-facilitated by the first and senior authors, whereas sessions for the second cohort were co-facilitated by the first author and another graduate student. PCT sessions were held weekly after school over five weeks. The first session was a lengthier orientation session lasting approximately three hours. The subsequent four sessions were approximately one hour each. All sessions were guided by an intervention manual that evolved over the course of the evaluation (Huey & Galbraith, 2020). On average, participants attended 80% of PCT sessions.

Youth and teachers were asked to complete a post-treatment assessment approximately one week after PCT ended. During the final PCT session, youth were instructed to practice the peer coaching skills they had learned over the previous weeks on their friends, write about their experience on a worksheet, and turn in this worksheet to one of the group facilitators approximately one week post intervention. As such, we assessed students exactly one week following the intervention, rather than immediately after the intervention ended, to capture any changes in behavior that may have occurred from completing this final assignment. Approximately three months post-intervention, teachers and youth completed follow-up assessments. Participants received $10 for each assessment they completed.

After completing follow-up assessments, youth and teachers were asked to participate in a qualitative interview to provide their overall impressions of PCT as well as specific feedback about the program. Due to COVID-19 shelter-in-place orders implemented shortly after follow-up data were collected, interviews were conducted on an encrypted, HIPAA compliant, university IRB-approved video conferencing application. Without the school as a hub for connecting with students and teachers, we were only able to contact a subset of the enrolled youth (n = 6) and teachers (n = 6) to complete interviews. All study procedures were approved by the University of Southern California Institutional Review Board (IRB).

PCT curriculum

PCT is modeled on Ross and McKay’s (1976) Peer Therapist program, an intervention offered to adolescent females in a juvenile detention setting. PCT retains three core features of the Peer Therapist curriculum. First, instead of directly “treating” youth, PCT trains youth in behavior change strategies, which youth subsequently use to affect change in their peers. Second, youth serve as “coaches” by using their newly acquired skills to influence their close peers in a prosocial direction. In PCT, youth are not asked to formally mentor or “coach” one specific peer but are instead encouraged to use their skills as much as they can with any of their peers who might benefit. Finally, our PCT model adopts a strengths-based, de-pathologizing approach that eschews the use of punishment, criticism, or confrontation. Instead, the focus is on reinforcing youth competencies, prosocial skills, and personal strengths.

The first three sessions were each dedicated to teaching the youth a new skill: the introductory session focused on positive reinforcement, the second focused on critical feedback, and the third focused on active listening. In the final two sessions, youth integrated the skills they learned in the first three sessions. They were also asked to practice using these skills in and outside of session to solidify their roles as peer coaches. All sessions were designed to maximize youth engagement throughout the intervention. Didactics were kept to a minimum, and sessions were structured to be as interactive as possible. For example, rather than asking youth to simply describe examples of the concepts they were taught, group facilitators would ask youth to act out examples of each concept to assess their comprehension (e.g., youth role-played examples of positive reinforcement).

Youth were assigned brief exercises to complete between sessions to encourage practice and retention of their peer coaching skills. These assignments were typically worksheets that instructed youth to practice what was taught during the previous session and report on the outcomes of this practice. See Table 1 for a brief overview of PCT session themes and homework assignments. Sessions occurred over the course of five weeks (one session per week) with a final, brief check-in regarding the last homework assignment one week after the fifth and final session. Additionally, one of the intervention facilitators (first author) checked in with each youth once monthly via phone call or text message to remind youth to use their peer coaching skills; these check-ins ended at the final follow-up assessment.

Table 1 Summary of Peer Coach Training (PCT) Session Structure and Content
Full size table

Intervention adaptations

Although the session goals and content were identical for both cohorts, we made various adjustments to program structure throughout the intervention to optimize youth engagement, most of which were based on experiences the facilitators had conducting the cohort one sessions. Throughout session one, especially during the didactic portions, youth in the first cohort were relatively disengaged and easily distracted (e.g., playing on their phones during the session, making inappropriate comments about other group members, getting in and out of their seats, climbing on furniture). Thus, after that initial session, we instituted a set of “group rules” for subsequent sessions to which the facilitators and participants could refer to curb future disruptions. We also minimized the didactic portion of the curriculum and increased the amount of interactive role play. Additionally, during the first session, several youth wanted to help videotape the skits and asked if they could share clips of the skits with their friends. In response, we added an extensive videotaping component to the curriculum, which involved having youth record each other perform the skits in each session. At the end of the program, each participant was given a short video that compiled the skits their group acted out throughout the first four sessions. Each participant was given a digital copy of the video to have and share with family and friends as they wished. The purpose of these two adjustments was to increase youth participation during session and to address the youth’s requests to potentially share what they were doing with others outside of the program (rather than “sharing” via social media). Finally, it was decided that to maximize homework completion (which served as prompts to “practice” as peer coaches outside of session), group-level contingencies (e.g., movie tickets, soda) were awarded if the majority of group members returned their completed homework. In sum, although the content of session one (and subsequent sessions) was similar for both cohorts, the intervention developers adjusted the structure of each session after cohort one’s first session to better facilitate youth engagement.

Assessment Measures

Disruptive behavior

Youth rated their own behavior problems using the Youth Self-Report form (YSR; Achenbach, 1991) at baseline, posttreatment, and three-month follow-up. The YSR has 102 items for which youth provide ratings of “not true” (0), “somewhat true” (1) or “always or often true” (2) about their own problem behaviors.

Teachers rated youth behavior problems using the Teacher’s Report Form of the Child Behavior Checklist (TRF; Achenbach, 1991) at baseline, posttreatment, and three-month follow-up. The TRF consists of 118 items. Teachers provide ratings of each item with either “not true” (0), “somewhat true” (1) or “always or often true” (2).

The TRF and YSR are well-validated assessment tools that have high reliability, criterion validity, discriminant validity, and convergent validity across a diverse range of populations (Achenbach, 2019; Raines & Crumpton, 2017). Each measure produces syndrome scales, DSM-oriented scales, broadband scales, and a total problems scale, with internal consistencies ranging from α = 0.67–0.95 (Achenbach, 2014). Given the target of our intervention, we reported only the YSR and TRF scales relevant to disruptive behavior. The syndrome scales used in this study include attention problems, rule-breaking behavior, and aggressive behavior. The DSM-5 oriented scales used include attention-deficit/hyperactivity problems, oppositional defiant problems, and conduct problems. The broadband externalizing problems scale was also used, which includes the three syndrome scales noted above. All scale scores are normed based on nationally representative samples. T scores below 65 are within the normal range, T scores between 65–69 fall within the borderline range, and scores of 70 or higher fall in the clinical range (Achenbach, 1991).

At each assessment period, teachers also completed the Disruptive Behavior Disorders Rating Scale (DBDRS; Pelham et al., 1992), a 45-item survey that assesses for disruptive behavior disorder symptoms. Respondents are asked to rate each item based on what best describes the youth’s behavior on a four-point likert scale (from 0 “Not at All” to 3 “Very Much”). It is a well-validated and reliable assessment tool used across diverse youth populations (Erford, 1997; Hambly et al., 2017; Pelletier et al., 2006). Internal consistency for the three DBDRS scales – Oppositional Defiant and Conduct Disorder (OD/CD), Inattention, and Impulsivity – ranges from 0.75 to 0.96 (Hambly et al., 2017; Pelham et al., 1992).

Youth satisfaction

A brief satisfaction survey was developed based on Attkinson’s Client Satisfaction Questionnaire, a well-validated measure used to assess client satisfaction in health and human services settings. The eight items on this scale were adapted to fit the intended goals of the PCT intervention, i.e., to provide youth with skills to effectively help their peers. A sample item includes “Have the services you received improved your ability to help your friends in need?” Youth provided ratings for each item on a scale of 1–4, with higher scores indicating higher levels of satisfaction.

Statistical Analyses

The primary goal of the study was to assess the feasibility and acceptability of this novel intervention. Thus, despite our small sample, we felt that nonparametric statistical tests were appropriate given our study goals. A priori power analyses using G*Power (Franz et al., 2009) suggested that with three time points and nine participants, we would have 80% power to detect medium or larger effect sizes. We used Friedman’s one-way analysis of variance (ANOVA; Friedman, 1937) by ranks to examine changes in disruptive behavior across the three assessment time points. Friedman’s ANOVA is a nonparametric statistical test used to assess changes in single samples across three or more time points; it is an extension of the sign test and involves ranking each row of data. It is often used with small samples, as the data is less likely to be normally distributed (Zimmerman & Zumbo, 1993). Effect sizes for Friedman’s ANOVA can be calculated using Kendall’s W tests (Friedman, 1937). Kendall’s W values range from 0 to 1, and the effect size categorizations are as follows: small effect (0.1), moderate effect (0.3), and large effect (0.5 and above; Tomczak & Tomczak, 2014). We conducted post-hoc Dunn-Bonferroni tests to reduce likelihood of Type I error (Dinno, 2015).

Youth and Teacher Interviews

The interview scripts were developed by study authors to assess youth and teacher impressions of PCT. The qualitative interviews were conducted by a trained research assistant, under the supervision of the study investigators. The interviews lasted approximately 15–30 min, and all were audio or video recorded. Participants (youth and teachers) were compensated $25 for completing both the interview and satisfaction survey (youth only).

The interview guides included questions about the following: 1) impressions of PCT, 2) overall impressions of PCT facilitators, 3) most and least helpful parts of the PCT program (youth only), 4) changes in behavior, 5) changes in academic performance, 6) student-teacher relationships, and 7) whether respondents would recommend PCT to peers/other schools.

Interview Data Analyses

Interview recordings were reviewed by the first author and a research assistant for completeness and accuracy. The first author extracted central themes using thematic analysis (Braun & Clarke, 2006; Clarke & Braun, 2014; Terry et al., 2017) for both teacher and student interviews. Thematic analysis is a flexible qualitative analytic strategy for identifying, analyzing, and reporting patterns within data. We used inductive thematic analysis (i.e., data driven rather than theory driven) to generate data on the youth’s experiences with and teachers’ perspective on PCT. Thematic analysis, as outlined by Braun and Clarke (2006), includes the six following phases: 1) familiarizing oneself with the data, 2) generating initial codes, 3) searching for themes, 4) reviewing themes, 5) defining and naming themes, and 6) producing the report. In phase 1, a research assistant transcribed the data and the first author read and reread all interviews. In phase 2, the first author manually and systematically coded the data and used an inductive approach to generate codes (i.e., developed codes based on interview data rather than interview questions). In phase 3, after all data were coded and collated, codes were sorted into potential themes. In phase 4, these codes and themes were reviewed and refined (e.g., themes that were originally too broad, such as “improved classroom behavior” were expanded into more specific themes including “increased collaboration with peers” and “improved class performance”) based on discussions with an expert in qualitative data coding and analysis. In phase 5, these themes were then formally defined and named. The final themes that emerged from the data were generated by the first author and are reported in the results section (i.e., phase 6).

Results

Assessment Data

Disruptive behavior

Youth

Table 2 summarizes outcome results for the YSR from baseline to follow-up among youth enrolled in the PCT program. Significant effects were found over time for externalizing problems and for the following syndrome scales: attention problems, rule-breaking behavior, and aggressive behavior. A Dunn-Bonferroni post hoc test indicated significant reductions in externalizing behavior from baseline to posttreatment (p = 0.003), and from baseline to follow-up (p = 0.018). Similarly, post hoc tests indicated significant reductions in attention problems from baseline to posttreatment (p = 0.025), and from baseline to follow-up (p = 0.045). Dunn-Bonferroni post hoc tests indicated significant reductions in rule-breaking behavior (p = 0.013) and aggressive behavior (p = 0.034) from baseline to posttreatment, but no significant effects for either from baseline to follow-up.

Table 2 Youth Self-Report Scores
Full size table

There were also significant effects over time for two of the DSM based scales: attention-deficit/hyperactivity problems and conduct problems. Post-hoc tests showed significant reductions in attention-deficit/hyperactivity problems from baseline to follow-up (p = 0.034), and in conduct problems from baseline to posttreatment (p = 0.025) among PCT participants. All other pairwise comparisons were nonsignificant.

Effect sizes for each YSR scale ranged from small (0.1) to large (>0.5). Of the YSR scales with significant results, externalizing behavior and rule-breaking behavior yielded large effect sizes. Attention problems, aggressive behavior, attention-deficit/hyperactivity problems, and conduct problems showed moderate effect sizes.

Teacher

Table 3 summarizes outcome results for the TRF from baseline to follow-up. All TRF scales on PCT youth’s behavior yielded null findings.

Table 3 Teacher Report Form Scores
Full size table

Table 4 summarizes outcome results for the DBDRS from baseline to follow-up. All DBDRS scales assessing changes in the PCT participant’s behavior yielded null findings.

Table 4 Teacher Disruptive Behavior Disorder Rating Scores
Full size table

Youth satisfaction

Table 5 summarizes acceptability outcomes from the Youth Satisfaction Survey. Results show that on average, youth participants were satisfied with PCT. Only one item “Have the services improved your ability to help your friends in need?” yielded an average below 3. Means, standard deviations, and median scores for each PCT Satisfaction item are included in Table 3 below, with higher scores indicating higher satisfaction ratings on a scale of 1–4.

Table 5 Peer Coach Training Youth Satisfaction Ratings
Full size table

Interview Data

Youth Interview Data

Four central themes emerged from thematic analysis of youth interviews with respect to youth behavior change. Youth participants stated that they had 1) improvements in their self-confidence, 2) increases in their helping behaviors to their friends, 3) improvements in their own and their peers’ behavior as a result of their coaching, and 4) high satisfaction with their experience in PCT.

Theme 1: Improvements in Self-Confidence. Half of the youth interviewed endorsed that their participation in PCT led to improvements in their self-confidence with respect to social interactions. For example, one youth stated:

I’m shy but PCT made me feel confident enough to open up in class.

Another student noted:

PCT gave me the skills to use words in conflict. It made me feel better about myself to be able to use words instead of getting physical.

Other youth participants said that PCT helped them feel more confident in handling conflict as well. For example, one student indicated:

PCT made me feel more confident in handling situations that got out of control.

Theme 2: Increases in Helping Behavior. Every PCT participant interviewed indicated that PCT motivated them to increase their helping behavior toward their friends. Specifically, youth participants stated that PCT motivated them to use their peer coaching skills to help their friends act in more prosocial ways. For example, one youth noted:

PCT made me want to do what I could to keep my friends on the right track.

PCT students also increased their efforts to help their peers act in fewer disruptive ways in class. One youth stated:

I wanted to make sure I used what I learned from PCT to help my friends stay out of trouble.

Finally, another PCT youth mentioned:

After PCT I used my leadership to de-escalate situations a lot more than I did before.

Theme 3: Improvements in Own and Peers’ Behavior. The majority of youth reported changes in their own and their peers’ behavior from learning peer coaching skills. With respect to behavior changes in their peers, one PCT youth said:

Using my coaching skills made my friends act more nicely to me and to others.

Another PCT participant reported a similar experience:

My friends followed the rules more when I used my peer coaching skills to encourage them.

And with respect to changes in participants’ own behavior, one youth in the PCT program reported:

I have a lot of anger, but PCT helped me learn to control my anger and use my words when I have to give someone constructive feedback.

Finally, one youth highlighted how PCT produced improvements in his own and his friends’ behaviors simultaneously:

PCT helped me use my words and helped me get my friends to use their words more instead of getting physical like we normally do.

Theme 4: Satisfaction with PCT. Aside from the three prior interview themes that emerged regarding self-reported behavioral changes, interviews from participating youth indicated high acceptability of the PCT program, which is congruent with findings from the PCT Satisfaction Survey. All youth stated that they would participate in the intervention again if it was offered. Most youth highlighted that their favorite part of the program was the role play and videotaping aspects of the sessions. For example, one youth stated:

I liked when we acted everything out and got to videotape what we did.

And another mentioned:

I really liked when we could record each other and direct the scenes that were “real-life” scenarios that happened in school to help kids act better.

Youth also indicated that they found the coaching skills they learned in PCT to be quite useful. Every youth participant stated that they would recommend the program to their friends. In general, interviewed youth struggled to come up with negative qualities about the program when asked, although two youth did note that they wished the counselors were “stricter” with one of the more disruptive students in their cohort.

Teacher interview data

Four central themes emerged from teacher interviews. Teachers primarily endorsed 1) improved student effort on classroom assignments and class participation more broadly, 2) increased collaboration with peers, 3) increased self-confidence among PCT youth, and 4) high satisfaction with PCT.

Theme 1: Increased Self-Confidence Among PCT Youth. Almost every teacher commented that PCT helped boost the PCT participant’s confidence in the classroom. For example, one teacher stated:

I saw a positive change in [the PCT student’s] demeanor and how he carried himself.

A different teacher stated that they saw their student “brighten up.” A third teacher indicated:

I noticed she [PCT student] was smiling a lot more. She seemed to be more confident to speak up in class.

Theme 2: Improved Class Performance. Four of the six teachers interviewed indicated that they had witnessed improvement in the PCT student’s academic performance, especially with respect to assignment efforts and class participation. Specifically, teachers stated that they saw an increase in the number of homework assignments turned in, and that their PCT student spoke up more in class. For example, one teacher reported:

[PCT student] showed a bump in his homework completion.

Another teacher stated about their PCT student:

I saw a big improvement in classroom engagement.

And a third teacher indicated about their PCT student:

I saw a big change in effort. She put in her best effort much more consistently.

Finally, a fourth teacher indicated that her PCT student’s grades improved from failing to a B + over the course of PCT, which she attributed to the student’s participation in the program.

Theme 3: Increased Collaboration with Peers. The majority of teachers witnessed increases in their PCT student’s positive collaborative efforts in the classroom. For example, one teacher stated that:

I saw [the PCT student] collaborate more with other students in the class.

Another teacher noted:

[The PCT student] seemed to make more effort to help his peers in class when they needed it.

Furthermore, one teacher indicated that:

[The PCT student] was often antagonistic and standoffish, but then that started to level off and she began to collaborate much more with her classmates.

Theme 4: Satisfaction with PCT. In general, teacher informants reported that they liked PCT’s strengths-based approach. Every teacher noted that they would like to see PCT implemented at their school the following year, and that they would recommend the program to other schools with similar populations (e.g., low-income). Only two teachers reported that they did not witness any positive behavioral changes in their students participating in PCT, but both teachers noted that they still had positive impressions of the program.

Discussion

This paper evaluated the acceptability and feasibility of PCT, a brief strengths-based intervention adapted for disruptive youth in an urban, middle school setting. Throughout the program, youth were encouraged to view themselves as coaches with the capacity to influence their peers to act in more prosocial ways. During the implementation process, several key adjustments were made to improve youth engagement and enthusiasm for the program, which included adding: (1) group rules to clarify and cue acceptable in-session behaviors, (2) a protocol for performing and videotaping behavioral practice skits, and (3) group contingencies to encourage the youth to practice their peer coach training skills outside of group sessions. Quantitative analyses showed promising improvements for youth report of externalizing behavior, rule-breaking behavior, aggressive behavior, attention-deficit problems, and conduct problems, whereas teacher report indicated no changes in behavior over time.

According to qualitative interview data, youth unanimously reported that they enjoyed participating in PCT, and they would highly recommend it to their friends. Youth also noted that the intervention led them to increase their helping behavior toward their friends (e.g., helping their friends use “words” instead of physical force), which ultimately led to improved behavior among their peers. Teachers appreciated the strengths-based orientation of PCT and witnessed improvements in the participating youths’ classroom participation and self-confidence. All youth and teachers indicated that they would like to see PCT implemented again the following school year. Overall, we found that PCT was a promising and feasible intervention for disruptive, predominantly Latinx, middle-school youth.

We believe that youth adoption of a “peer coach” identity could be one potential mechanism explaining post-PCT reductions in problem behaviors. PCT youth were encouraged to support prosocial behaviors in their peers by using positive reinforcement, constructive feedback, and focused listening skills. In doing so, youth may have internalized new identities as “coaches” to their peers, which may have caused a shift in their behavior to better reflect their new identities (Ross & McKay, 1976). Indeed, studies show that having a positive self-identity is linked to higher levels of prosocial behavior (e.g., Crocetti et al., 2012; Crocetti et al., 2014; Patrick et al., 2018), and lower levels of antisocial behavior (e.g., Bruner et al., 2014; Kavussanu & Al-Yaaribi, 2021; Shields et al., 2018). However, changes in youth behavior over the course of the program were either not noticed by teachers or did not occur in the presence of their teachers.

There are several possible explanations for this discrepancy between teacher and youth reports of problem behavior. First, teachers were reporting on student behavior in their classes and youth were likely reporting on their behavior in general (i.e., in all classes and outside of school). Supporting this hypothesis, although youth reported reductions in their disruptive behaviors more broadly in qualitative interviews, they did not specify that these reductions occurred in classroom contexts per se – e.g., these reductions may have been more salient in recreational contexts, but not necessarily in the classroom. Thus, it is possible that overall, youth did reduce their disruptive behavior, but that the skills they learned were perhaps more salient in settings outside of the classroom.

A second explanation for the discrepancy in perceptions of disruptive behavior could be that teachers and students often disagree on reports of problem behavior. ASEBA cross-informant research shows low correlations between youth and teacher ratings of youth externalizing behavior (De Los Reyes et al., 2019; Youngstrom et al., 2000), which could indicate that youth behaviors are different across contexts (e.g., home versus school; Achenbach, 2014, De Los Reyes et al., 2015; Rescorla et al., 2017; Santos et al., 2020). Observations of youth behavior during the intervention lends credence to this explanation. Although many within-session role plays reflected school themes, when youth were asked to generate their own peer coaching scenarios, they often offered examples that occurred outside of the classroom (e.g., with siblings at home, at the park with friends). Furthermore, when youth were asked to report on instances in which they practiced their peer coaching skills, they often wrote about coaching behavior outside of school. It is important to note, however, that we do not have collateral data from other sources to confirm this hypothesis, as we were unable to gather data from caregivers or peers that could provide insight on the youth’s behavior outside of school (e.g., Dodge et al., 2015; Rescorla et al., 2017).

There are other possible explanations for the discrepancy between youth and teacher assessment data. For example, it is possible teacher assessment measures were less sensitive in detecting youth behavior change relative to qualitative interview questions. This could explain why youth and teacher qualitative findings converged much more than quantitative data from youth and teacher assessments. Moreover, it is possible that differences in the salience of the “peer coach” identity between youth and teachers may have contributed to discrepancies. PCT youth were actively encouraged to self-identify as “helpers” throughout the five-week intervention, and thus may have been more likely than teachers to rate themselves as less disruptive because of a prosocial identity shift (Ross & McKay, 1976).

Limitations and future directions

There are important limitations to consider. Given the feasibility focus of this study, our sample size was small, no comparison group was used, and the intervention was carried out at a single middle school in South Los Angeles. Although we did find significant improvement for some youth-reported problem behaviors, results may not generalize beyond Latinx youth, and we cannot say that our intervention necessarily caused these changes. Moreover, increased variability in the data due to small sample size may have contributed to discrepancies between self and teacher reports of youth behavior; it is possible that a larger sample may have yielded results for which youth and teacher data converged. A randomized trial with a larger and more diverse sample is needed to demonstrate that PCT is effective in reducing disruptive behavior in middle school youth from different backgrounds.

Despite these limitations, this small pilot trial has important implications for future research. First, this brief intervention is feasible to implement in a low-income middle school. In addition, these preliminary results suggest that PCT may be a promising approach to facilitating prosocial behavior, although additional refinements might be needed given the disappointing teacher assessment findings. Future trials of PCT should consider collecting data on the behavior of the enrolled youth’s peers as well, as it would allow us to examine whether this intervention has impacts beyond the individuals participating in PCT, as youth qualitative interview data suggested. In short, it appears that PCT is a promising intervention that should be tested on a larger scale with additional resources.