Study Highlights Five Ways Mentors Shape Teen Autonomy

Davis, A. L., & McQuillin, S. D. (2022). Supporting autonomy in youth mentoring relationships. Journal of Community Psychology, 50(1), 329–347. https://doi.org/10.1002/jcop.22567

Introduction

Adolescence brings a drastic increase in decision-making power, and how young people exercise that power carries real consequences for their long-term wellbeing. Davis and McQuillin (2022) note that while the developmental importance of autonomy is well established across disciplines, almost no prior work has examined the specific role mentors play in either fostering or threatening it. The authors frame autonomy as the freedom to make choices aligned with one’s own values, drawing on Deci and Ryan’s (1987) self-determination theory, which distinguishes between decisions driven by internal motivation and those driven by an external locus of causality, such as guilt or social pressure. They also address the double-edged nature of autonomy directly: while autonomous decision-making is linked to better mood, self-esteem, and academic performance, adolescents who exercise autonomy without adult guidance, or who are exposed to coercive control by adults trying to protect them, tend to show worse outcomes, including deviant behavior and strained relationships. Mentors, the authors argue, are uniquely positioned to find the middle ground because their role does not inherently require them to instruct or make decisions on a young person’s behalf.

Methods

This is a conceptual and theoretical paper rather than an empirical study, so there is no sample, data collection, or statistical analysis to report. Davis and McQuillin instead conducted a narrative synthesis of existing scientific literature spanning developmental psychology, motivational interviewing, parenting research, and youth mentoring evaluations. Their approach was to identify recurring patterns across these separate literatures and integrate them into a single organizing framework specific to mentoring relationships. The paper builds toward two visual models: a “harmful autonomy-seeking spiral,” illustrating how knee-jerk adult control can push adolescents toward rebellion and increased risk, and an “upward spiral of autonomy-supportive mentoring,” which starts with mentors understanding youth values and builds toward more active forms of influence.

Results

The author’s contributed a five-domain framework describing how mentors influence mentee autonomy, ordered from lowest to highest risk of backfiring:

• Role modeling is presented as the safest avenue of influence. Citing Reynolds and Parrish (2018), the authors note that mentors who modeled positive decision-making quadrupled mentees’ odds of college attendance, while offering direct advice showed no measurable effect in that same study.
• Encouraging desirable behaviors relies on contingency management, using praise and reinforcement rather than punishment, to shape choices without triggering feelings of being controlled.
• Providing access to resources, relationships, and experiences draws on Granovetter’s (1973) concept of weak ties, positioning mentors as conduits to social capital mentees may not otherwise access, provided mentors obtain consent rather than imposing resources unilaterally.
• Advocacy, or acting on a mentee’s behalf with institutions like schools, can open doors to opportunities but risks harm if it proceeds without the mentee’s knowledge or consent.
• Conversations about behavior change are identified as the highest-risk domain. The authors point to McQuillin et al. (2011), where mentors with only minimal training in motivational interviewing produced worse academic outcomes than no mentoring at all, a finding attributed to controlling language, or “righting reflexes,” that can damage the mentoring relationship.

The clear takeaway: influence strategies are not interchangeable in risk, and the domains that require the most direct verbal engagement with a young person’s choices are also the domains most likely to backfire without proper training.

Discussion

Davis and McQuillin argue that good intentions are not sufficient protection against causing harm. A mentor who notices a mentee taking real risks may feel pulled toward tighter control, but the literature reviewed here, including Moffitt’s (1993) theory of adolescent-onset conduct problems, suggests that this instinct often produces the opposite of its intended effect, pushing adolescents toward greater rebellion and risk exposure. The authors introduce “autonomy-connectedness” (Maas et al., 2019) as the theoretical target mentors should aim for: a state where mentees maintain genuine independence in their decisions while still feeling securely attached to a supportive adult. They also note that these five domains appear to build on one another, meaning mentors who skip the foundational, lower-risk work of understanding a mentee’s values and role-modeling healthy choices may be setting themselves up for failure when they attempt higher-risk interventions like behavior-change conversations.

Implications for Mentoring Programs

The authors offer several recommendations. Programs should sequence mentor training so that lower-risk skills, understanding mentee values, modeling healthy decisions, are established before mentors attempt higher-risk strategies such as motivational interviewing. Programs relying on conversations about behavior change should provide ongoing, not one-time, training and supervision, since brief or nominal training in this domain has been shown to backfire. Mentors should also be taught to seek explicit consent from mentees and guardians before sharing resources or advocating on a mentee’s behalf, since both actions involve a power differential that can tip into coercion if handled carelessly. Finally, the authors suggest that formal, program-sponsored mentors and informal mentors from within a young person’s family or community may need different guidance, since natural mentors already embedded in a mentee’s life may be operating more like parenting figures.

Read the full article here.