The truth will set you free, but first it will piss you off: Why we need new approaches to youth mentoring

 

“The truth will set you free, but first it will piss you off.”
– Gloria Steinem

by Jean Rhodes

Note: In honor of the late, great Gloria Steinem, I have updated this this popular post with new research.

Over the course of my career, I have met many practitioners who have launched what sound like amazing mentoring programs. The practitioners are often new to the field, passionate about their mission and eager to join the broader mentoring movement. As much as I find their commitment inspiring, I also confess to occasional skepticism about the return on their investment of time and resources. My doubts are fueled by the somewhat sobering data which, when plotted over the course of a long career, emerge as an irrefutable trend line.

A quarter of a century ago, David DuBois et al. conducted a comprehensive meta-analysis of youth mentoring programs that showed disappointingly small effects (D. DuBois et al., 2002). Since then, at least five subsequent meta-analyses of youth mentoring have yielded broadly similar results (Burton et al., 2021; Christensen et al., 2020; Cuming et al., 2026; L. van Dam et al., 2020; D. DuBois et al., 2011; Poon et al., 2021; Raposa et al., 2019). Although there remains considerable variation across studies, the average effect size remains small by conventional benchmarks (Christensen et al., 2020; Raposa et al., 2019).

A 2026 systematic review of formal one-to-one mentoring for vulnerable young people likewise found generally modest effects, with more evident benefits when relational and organizational supports were well aligned (Cuming et al., 2026). This despite a rich and growing body of research and evaluation studies, a strong commitment on the part of MENTOR and other organizations to evidence, and advances in training and classification through the National Mentoring Resource Center.

The resistance to improvement is particularly humbling when we compare mentoring program effect sizes with those that have been found in meta-analyses of other prevention programs for children and adolescents. For example, a meta-analysis of 177 prevention studies found effects ranging from moderate to large, depending on program type and target population (Durlak & Wells, 1997). Likewise, meta-analyses of youth psychotherapy, encompassing hundreds of studies, have reported strong mean effects (Weisz et al., 2022). Interestingly, however, Weisz et al. note that, “in settings in which therapists were able to use their clinical judgment to deliver treatment as they saw fit, not constrained by evidence-based interventions or manuals, and in which there was a comparison of their treatment to a control condition. Meta-analyses of these studies of usual clinical care have found effect sizes averaging about zero (see, e.g., Weisz, 2004; Weisz, Donenberg, Han, & Weiss, 1995), indicating no treatment benefit” (Weisz et al., 2005). I’ll spare you more data, as your confirmation biases may have already kicked in. You may already be thinking “mentoring has strong effects, the data must be wrong!” Such biases are understandable as they are grounded in your lived experiences with caring adults. But there is often a gulf between the influence of a natural mentor and that of a typical volunteer in a typical program (L. V. Dam et al., 2018). A four-year randomized trial of 1,353 youth in the Big Brothers Big Sisters community-based program found significant effects on violence-related delinquency and recurring substance use, along with several secondary outcomes, but many outcomes were nonsignificant (D. L. DuBois et al., 2026).

There are several factors that have contributed to the state of affairs. One is the field’s uneven adherence to evidence-based practice. Earlier research showed that implementation of mentoring program practices was associated with longer-lasting relationships (Kupersmidt et al., 2017). More recent research links exposure to empirically supported program practices with mentor satisfaction, commitment, security, and lower negativity in the mentoring relationship, although this evidence is cross-sectional and focuses on relationship quality rather than youth outcomes (Keller et al., 2023).  A 2026 scoping review of 36 studies found that 16, or 44%, did not report using an established mentoring framework or guideline, and even when frameworks were mentioned, they were often used only to define mentoring rather than to shape program design (Lohmeyer et al., 2026). Despite the availability of evidence-informed practices, many programs continue to rely on friendship approaches and unvalidated curricula. A 2025 analysis of 47 mentoring programs found that programs were stronger when their goals matched what mentors were asked to do and how success was measured, when families and other supports were involved, and when mentors received support tailored to the young people they served (Griffith et al., 2025).

There is also the lack of specificity. Many programs will endorse the general goals of promoting healthy development, reducing risk, and ameliorating youth dysfunction and disorder. But those are tall and rather unspecific orders. Depending on their circumstances, youth need dramatically different things from their mentoring experiences. Nonetheless, many programs continue to embrace one-size-fits-all friendship approaches. In doing so, they are akin to a pediatrician who, irrespective of presenting problem, prescribes two aspirins to every child that visits. Children with low-grade fevers and muscle aches are likely to benefit from this generalist approach, and they may even be held up as anecdotal evidence that the treatment works. But, in the long-run, this cherry picking of positive effects ignores the many for whom aspirin had little or no lasting effect. Recent randomized evidence illustrates the value and limits of specificity. In a 2026 trial of Reach & Rise, a program in which mentors were supported in using cognitive-behavioral principles, youth showed reductions in delinquency and substance use and improved school connectedness and caregiver-rated academic performance, but not depressive symptoms, happiness, life satisfaction, or peer relationships (Jarjoura et al., 2026).

A more effective approach would be to first understand the needs of the youth and then provide targeted, intensive training to the volunteer or paraprofessional. This targeted, evidence-based approach will require precision, fidelity, and, perhaps most importantly, humility. Precision in identifying the needs or goals of the target population and the best strategies for addressing them; fidelity in implementing strategies that adhere as closely as possible to the program and conditions in which they were designed and evaluated; and the humility to refrain from reinventing the wheel when there are already rigorously developed and validated strategies (Axford et al., 2020; Christensen et al., 2020; McQuillin & McDaniel, 2020).