Mentor’s Self-Efficacy Trajectories during a Mentoring Program for at-Risk Adolescents

Introduction: Mentor self-efficacy has been shown to be positively associated with the quality of youth mentoring relationships and mentee adjustment. Mentor self-efficacy, however, may fluctuate overtime, potentially affecting the impact of mentoring. Identifying factors that may contribute to mentor self-efficacy trajectories, especially among mentors of at-risk youth, who may be more likely to report decreases in perceived self-efficacy overtime, is needed in order to increase the robustness of programs.

Method: The current study includes 239 adolescents (11-18) and their mentors who were recruited for a randomized controlled trial of a mentoring intervention for at-risk youth, known as Campus Connections (CC). Mentor’s perceived self-efficacy was measured at five time points (i.e., baseline, week 3, week 6, week 9, and week 11). Latent class growth analysis (LCGA) was used to estimate an intercept and slope for each mentor. The LGCA model then tested whether these intercepts and slopes could be categorized into meaningful subgroups using the following indices: bayesian information criterion (BIC), adjusted BIC, and Lo-Mendell-Rubin adjusted LRT. Multinomial logistic regression was used to assess whether mentee (i.e., risk status, age, and gender) and mentor characteristics (i.e., personality traits, positive and negative affect, role satisfaction, and perceived social support from CC staff) at baseline were significant predictors of mentor self-efficacy groups.

Results: Three subgroups emerged: Mentors relatively high in self-efficacy throughout the mentoring relationship (i.e., stable group; n = 100), mentors high in self-efficacy at the beginning of the relationship and increasingly so as the relationship continued (i.e., increasing group; n = 92), and mentors moderately high in self-efficacy and decreasingly so (i.e., decreasing group; n = 47). The decreasing group was selected as the comparison group, since it had the lowest levels of perceived self-efficacy throughout the mentoring experience. Regression analysis indicated that greater mentor conscientiousness (OR = 1.8) and extraversion (OR = 1.7), as well as less negative affect (OR = 0.68) were associated with a greater likelihood of belonging to the increasing group relative to the decreasing group. Only mentor extraversion (OR =1.4) significantly predicted the likelihood of belonging to the stable group versus the decreasing group.

Conclusions: Findings suggest that mentor characteristics play an important role in how mentors perceive their ability to maintain a positive relationship with their mentees. These findings have significant implications for mentoring programs designed for at-risk youth, particularly in terms of mentor selection and training.