24 Studies Later: The Real Elements of Effective Youth Mentoring
Cuming, T., Hoffman, L., Biles, B., & Verdon, S. (2026). The impact of formal one-to-one mentoring upon vulnerable young people: A systematic review. Youth & Society. Advance online publication. https://doi.org/10.1177/0044118X261469877
Introduction
Why do mentoring programs succeed in some circumstances and fall short in others? Cuming and colleagues (2026) argue that most prior reviews focused on aggregate outcomes, such as whether mentoring improved grades or reduced risk behavior on average, while overlooking the relational and organizational conditions that actually determine whether a given mentoring relationship takes hold in the first place. The authors base their approach in Bronfenbrenner’s Ecological Systems Theory, treating formal mentoring as a relationship embedded within embedded systems, family, school, community organization, and broader culture, that shape whether and how that relationship can function. The aim of the study was to describe how the literature conceptualizes the role of formal mentoring for vulnerable youth, and to synthesize the conditions that explain variation in how these relationships form, endure, and produce change.
Methods
The team searched six databases for peer-reviewed studies published between 2015 and 2025 examining formal one-to-one mentoring with young people aged 8 to 18 identified as vulnerable or disadvantaged. From 1,336 initial records, 24 studies met inclusion criteria, spanning the United States, Australia, the United Kingdom, Ireland, Canada, and South Korea, using qualitative, quantitative, and mixed-methods designs. Coding proceeded inductively, generating 113 preliminary codes that were consolidated into three higher-order themes.
Results
Trust, compatibility, and sustained engagement emerged as imperative to effective mentoring relationships, but these relational qualities depended on organizational scaffolding rather than chemistry within the mentoring relationship alone. Demographic matching, based on either gender or ethnicity, showed limited predictive value; it was mentor empathy, cultural responsiveness, and attunement to a young person’s life that mattered. Because mentors often became family-like figures for young people with unstable home lives, abrupt endings risked reproducing damage that mentoring was meant to repair. Programs that built in supervision, consistent staff contact, and planned closure protocols saw fewer premature endings, while under-resourced programs with high staff turnover saw more. Salaried, professionally supervised models like Friends of the Children showed greater reliability than volunteer-based programs, though at higher cost. Cultural responsiveness and trauma-sensitive practice, even when never named as such, appeared repeatedly as the conditions under which trust actually took hold. Outcomes across psychosocial, educational, and social domains were generally modest, but effects were strongest for young people with higher baseline risk and when program structures aligned tightly with relational quality. The impact of mentoring ultimately depended less on the mentor-mentee bond itself than on the organizational and cultural systems supporting it.
Discussion
The authors interpret these results through an ecological lens, arguing that the modesty of mentoring’s effects is not a failure of the relationship itself but a predictable consequence of everything surrounding it. Mentoring cannot compensate for weak program infrastructure or unaddressed structural disadvantage A mentor can offer stability, warmth, and consistent attention at the microsystem level, but if the exosystem around that relationship (i.e program supervision capacity, funding, staff retention) is unstable, those relational gains have little scaffolding to hold onto. This provides an answer to why the same intervention produces meaningful change in one context and insignificant change in another. The authors also draw attention to their findings about trauma-informed practices: none of the 24 included studies explicitly identified as trauma-informed, yet the relational conditions most associated with success, safety, predictability, attunement to a young person’s history of adversity, are precisely the conditions trauma-informed care prescribes. This suggests mentoring programs may already be doing trauma-sensitive work without the language or formal training to reinforce and scale it deliberately. Additionally, cultural humility and responsiveness to community norms strengthened engagement more reliably than demographic matching did.
Implications for Mentoring Programs
Programs serving vulnerable youth should treat organizational infrastructure as a precondition for relational success rather than an afterthought secondary to the significance of a good match. That means investing in manageable caseloads, regular supervision, and coordinator availability, since these were the factors that most consistently predicted whether a mentoring relationship survived its early, fragile months. Programs should also build planned closure procedures into their design from the outset rather than treating endings as an afterthought, given how consistently abrupt endings were linked to diminished trust and reluctance to re-engage among young people who had already experienced relational disruption. Rather than prioritizing demographic matching, programs should invest training resources in cultural humility, attunement, and trauma-responsive practice, the qualities this review found to matter more than shared identity between mentor and mentee. This includes screening for trauma exposure at intake and ensuring ongoing supervision draws on staff with relevant clinical grounding, rather than relying on brief pre-match training alone. Finally, programs working with culturally distinct communities, including Indigenous communities, should resist assuming a one-to-one dyadic model is the right fit without genuine community consultation and leadership in program design.
Read the full article here
