Mentor-Mentee Relationship Quality and Problem Behavior in a Group Mentoring Program for Minority Youth: Measures and Moderation Effects

Mentoring programs are increasingly likely to serve youth disproportionately at-risk for engaging in problematic behavior (Jekielek et al., 2002). Benefits youth glean from mentoring are evidenced; e.g., a meta-analysis of program evaluations found modest effect sizes for improvements in youth aggressive behaviors (d=0.29) and academic achievement (d=0.11) (Tolan et al., 2014). Mentor-mentee relationship quality is an influential factor thought to explain how mentoring programs produce positive outcomes (Whitney et al., 2011), such as has been found with intentional self-regulatory socioemotional skill development (Bowers et al., 2016).

Although traditional mentoring programs’ ratio of mentors to mentees is one-to-one, certain settings are amenable to the use of small group mentoring to reap demonstrated benefits (Herrera et al., 2002). Literature about small group mentoring, also known as “pyramid mentoring,” points to its utility and benefits with high-risk youth populations (i.e., youth in a “system” of juvenile justice or outside home care). Whereas the ideal mentor-mentee relationship consists of equilibrium between vertical (knowledge and experience of mentor with mentee as beneficiary) and horizontal (friendship, mutuality) aspects (Keller & Pryce, 2012), small group mentoring provides balance with youth as mentoring recipients while horizontal interactions with peers are fostered.

This paper describes an evaluation of a small-group mentoring program (N = 225 youth, 55.5% male, 38.9% African-American, 54.1% Latino/Hispanic, mean age 13.74 years) with over 50 adult mentors for 52 or more hours per year. Significant pre-program, as compared with current reported, score differences were found for mentor-mentee relationship quality (increase) and problematic behaviors (decrease). Also, significant associations were found between youth reports of relationship quality experienced with their mentors during the program (consisting of how emotionally engaged mentors appear to be in mentoring relationships and how youth-centered mentors are perceived as) and problem behaviors that were indicative of early stages of Loeber’s (2014) three types of delinquency pathways. Gender moderated the association with females’ results significant and males’ non-significant. A regression model, testing for mediation by youth program satisfaction to explain the relation between mentor-mentee relationship quality and youth problem behavior, found a suppressor effect. This was interpreted as a classic suppressor effect as scales used to measure mentor-mentee relationship quality appear to reflect client satisfaction with program provider, versus mentor-mentee emotional attachment or attachment security, as has been measured in other studies. Implications for youth-centered mentor training, characteristics of successful small group mentoring programs, and choice of evaluation measures are included in the discussion.