Recent meta-analysis highlights the promise of cross-age peer mentoring

by Jean Rhodes

Although several meta-analyses of intergenerational mentoring programs have been conducted, my colleagues and I recently conducted the first comprehensive meta-analysis of cross-age peer mentoring programs for youth. Based on the doctoral dissertation of Samantha Burton, Ph.D., the study shows some potentially promising results. Indeed, analyses revealed a medium-sized overall effect of cross-age peer mentoring programs (g = .45), which is more than double that observed in past meta-analyses of  intergenerational youth mentoring programs in which an adult mentor is paired with the youth mentee  (i.e., ranging from g = .18 to .21; Dubois et al., 2002; Dubois et al., 2011; Raposa et al., 2019). Given the far greater emphasis on intergenerational mentoring programs compared to cross-age peer mentoring programs in practice and in the research literature, these findings are notable and may lead to increased interested in that latter. Cross-age peer mentoring can offer feasible and efficient opportunities to have older peers mentor youth (e.g., by pairing up youth from different grades within the same school) with the potential for mutual benefit.

But, as with what we’ve learned in intergenerational mentoring meta-analyses–it has to be done right! Consistent with some previous findings, programs with moderate to high levels of adult oversight and supervision had larger effects compared to programs with low levels of adult oversight and supervision. Elements of strong adult oversight and supervision included:

• mandatory training for mentors; supervision to support intervention delivery; videotaping select mentor-mentee interactions to monitor intervention quality and provide additional support as needed; and program staff and parent participation in program activities, all of which UTSA Professor Michael Karcher deploys in his excellent mentoring program (Karcher et al., 2002; Karcher, 2005b; Smith, 2011; Tomlin, 1994).

In contrast, programs with low adult oversight and supervision provided less than two hours of training, with a significant portion of youth mentors reporting receiving no training at all  Adult oversight and supervision may improve treatment fidelity, a construct that has repeatedly been linked to better treatment outcomes. Adult supervision also provides opportunities for verifying that mentoring interventions are carried out as intended, while also ensuring appropriate mentor-mentee interactions and preventing obstacles to program success (e.g., mentors not understanding their role or carrying out their role ineffectively). This oversight may be particularly important for youth mentors, given their developmental stage, maturity level, and the potential for negative peer influence.

Results also demonstrated that peer mentoring programs in urban settings yielded larger effects than those in rural settings. Although speculative, it is possible that programs in urban settings may serve youth who experience increased environmental risk factors such as exposure to high rates of crime, violence, delinquency, substance use, and poverty. Some meta-analytic evidence from intergenerational mentoring suggests that programs serving youth with greater levels of these kinds of environmental risk factors may tend toward larger effect sizes (Dubois et al., 2011). In general, programs that work with youth who experience more acute behavioral, emotional, and academic problems tend to show stronger positive effects because, like a pendulum, there is more room for an upswing.

Additional moderators of program effect size were also observed. Larger effects were observed for programs operating in the community or outside of the school day (i.e., weekend or summer) compared to those that were school-based (i.e., in schools, during or after the school day). It is possible that space and time constraints within school settings may restrict the types of activities in which matches can engage. School-based programs operating during or after the school day may be more likely to meet within a group format in a space like the school cafeteria, which could distract from intended one-on-one interventions . In addition, school-based programs that meet after the school day ends may be negatively affected by youth fatigue that limits mentees’ ability to fully engage in program content, as well as mentors’ capacity to deliver interventions successfully. Future research should replicate and extend these findings, perhaps exploring the interacting factors that may influence the impact of program setting on effect size. This area of research is particularly important given that many youth access older peer mentorship within school settings, and school-based mentoring has become one of the most popular contexts in which youth receive mentoring services (Garringer et al., 2017). As this literature grows, future meta-analyses should explore additional factors that improve peer mentoring.

Taken together, the current findings provide support for the efficacy of programs that foster one-on-one, caring relationships between youth and older peers who are closely supervised by adults, particularly as an intervention with the potential to reach large groups of youth and prevent more intensive treatments. They highlight opportunities for improving the quality and rigor of mentoring practice, particularly moving toward interventions with strong attention to oversight of mentors. Mentoring programs should  creating supportive, scaffolding environments for young mentors, thereby facilitating increased mentor competency and fostering mentee success. Increased implementation and evaluation of cross-age peer mentoring programs in this way is a promising path to scale supplemental and preventative services to youth.

Want to understand more about meta-analysis? See this excellent, approachable overview.