Three research-based recommendations for improving youth and mentors’ engagement in programs

by Jean Rhodes

Youth can’t benefit from mentoring programs they (or their mentors) don’t attend, and poor attendance is a common problem in many programs. One factor that may discourage attendance is the threat of violence and crime. In this important new study, researchers studied this issue and make valuable suggestions for how to increase attendance in programs that offer services to youth of color in high poverty, high crime neighborhoods.

Selected sections excerpted from Wathen, M. V. et al. (2021), Toward improvement engagement of youth of color in cross-age mentoring programs in high poverty, high crime neighborhoods. Children and Youth Services Review, 12. 


“This study examines predictors of attendance in a cross-age youth mentoring program offered in four high-poverty, high-crime communities. Youth in greatest need of after school and summer interventions are those residing in such communities, but programs are scarce. More specific to mentoring programs that do exist, past research has demonstrated the significance of attendance as a predictor of positive outcomes. Two datasets were combined for this study: The Saving Lives, Inspiring Youth (SLIY) mentoring program dataset and a neighborhood database. OLS regression results show that for all participants, traveling from a lower-crime home area to a program in a higher-crime area was negatively related to attendance, as was age. In addition to crime, variables related to attendance for mentors included stress, perceived family resources, and race, whereas age and having a sister were related to attendance for mentees. Implications for program designers and policymakers are discussed.

Excerpted from Intro

A critical challenge in empowering youth of color in low-income environments is engaging them in available programs and services (Deschenes et al., 2010). For example, the hallmark evidence-based program, “Becoming a Man,” reported only 50% attendance (Heller et al., 2013). What might be contributing to this low rate of engagement? This issue is important in part because youth of color in low-income areas are already deprived of supportive social services and after-school programs much more than their privileged peers (Bringewatt and Gershoff, 2010Bruce and Bridgeland, 2014). For example, only 10–15% have after school and summer programs available to them (Deschenes et al., 2010). A 2014 report by the Afterschool Alliance states that, “More than 2 out of 3 parents living in communities of concentrated poverty (67 percent) report that finding an enriching environment for their child in the afterschool hours was a challenge, compared to 46 percent of parents living outside of these areas” (2014). This report concludes that access to quality programs for children and youth is still lacking, and that affordability for low-income families exacerbates the problem (Afterschool Alliance, 2014). Thus, accessibility and engagement remain important issues…

Barriers to attendance

One barrier to participation is that youth in low-income neighborhoods often have limited or no access to mentorship programs (Deschenes et al., 2010). Research demonstrates that youth in neighborhoods with greater resources have more opportunities to participate in mentorship programs, whereas youth in neighborhoods with higher community violence traditionally have limited programs available (Wesely et al., 2017).

The negative repercussions of lack of access to mentoring programs are compounded by challenging environments. Youth living in neighborhoods with higher rates of community violence and poverty are often subject to more complex and ongoing trauma than those in lower crime neighborhoods (Beardslee et al., 2019Richards et al., 2016). Oftentimes, the trauma experienced by youth in these environments has lifelong effects on their personal and professional development (Bulanda et al., 2013). Youth and their caregivers must consider violence and safety as they make decisions about where and whether youth travel outside the home (Jacoby et al., 2017Shedd, 2015). Youth who live in neighborhoods with high poverty are more likely to engage in violence as witnesses, victims, and, more rarely, perpetrators, experience more health and mental health challenges, have lower academic achievement, and are more likely to live in poverty in adulthood (Beardslee et al., 2019Duncan and Kawachi, 2018; N. R. Karcher et al., 2021Machell et al., 2016Murry et al., 2011Nieuwenhuis et al., 2017Smart et al., 2021). These issues illuminate the importance of providing youth from lower resource communities access to youth mentorship programs…

The importance of attendance

Past research has demonstrated the significance of ‘dosage,’ operationalized as attendance, as a predictor of positive outcomes in peer-mentoring programs (M. J. Karcher et al., 2006). Mentoring program infrastructure plays a role in youth engagement and attendance (M. J. Karcher & Berger, 2017). Organizational considerations include screening, ongoing training, structured activities, explicit feedback, frequent check-ins, and other supports for mentors (Jarjoura et al., 2018).

Demographic factors may also impact participation. One important factor is violence in a youth’s neighborhood (Harden et al., 2015Mikhail and Nemeth, 2016). Oftentimes, youth identify safe and unsafe areas in their home neighborhoods, which impact their interactions with neighbors, relatives, and community members and help them navigate violence (Deane et al., in preparationDill and Ozer, 2016Shedd, 2015). Current research on this topic suggests that youth who experience community violence may utilize cross-age and adult-youth mentorship programs to access places and people they consider safe (Deane et al., in preparationDill and Ozer, 2016). Transportation issues, such as access to public transportation and the funds to afford it, may impede participation of youth (Coller & Kuo, 2014). One study found that competing family responsibilities such as babysitting or chores may influence attendance for mentees (Terzian et al., 2009). Additionally, youth who identify as male and who also experience adolescent social pressures may be less likely to participate as frequently in youth mentorship programs as they age than youth who identify as female (Coller and Kuo, 2014Sánchez, Pinkston, et al., 2018). However, exactly why programs fail to engage male youth of color needs further understanding (Terzian et al., 2009)..

Factors known to improve youth attendance

Research shows that several factors may enhance youth attendance in a mentorship program. One is community involvement. When families, parents, and caregivers of youth are invited to social events sponsored by a mentorship program, the participation of youth improves (Coller and Kuo, 2014Garringer and MacRae, 2008; M. J. Karcher & Berger, 2017). The mentor relationship is equally important for increasing participation, as mentors and mentees receive mutual peer support (Kiyama & Luca, 2014).

Research highlights the importance of recognizing differences between mentors and mentees in cross-age peer mentorship programs, primarily in their stage of development (M. J. Karcher and Berger, 2017Schenk et al., 2020). Youth social and behavioral maturity benefited both mentor and mentee attendance (Garringer & MacRae, 2008). In addition, mentors participate less consistently if their respective mentee possessed behavioral issues (M. J. Karcher & Berger, 2017). Mentors and mentees who score higher on the Social Interest Scale are also more likely to regularly participate in mentorship programs (Garringer, 2015). For mentee participation, it is important for parents or guardians to be well informed about the program (Garringer, 2015).

The current study addresses the gap in the literature around correlates of attendance in peer mentoring programs for youth of color in under-resourced communities.


Two datasets were combined for this study: The Saving Lives, Inspiring Youth (SLIY) mentoring program dataset and a neighborhood database created by the authors. A sample of 132 low-income, urban mentees aged 9–16 (M = 11.53 years, 56.6% female) was recruited to participate in SLIY mentoring program treatment group. Similarly, a sample of 148 high school mentors aged 14–21 years (M = 16.72 years, 60.6% female) was recruited for the treatment group.


This study examined the relationships between attendance in a cross-age youth mentoring program and participant baseline psychosocial and demographic data, along with zip and census tract data on crime, poverty, and other characteristics of the youth’s home neighborhoods and respective program site locations.

While it may appear that the poverty and crime experienced in these neighborhoods was markedly similar, there was enough variation to see significant effects. Even within high crime areas there were sub-areas of increased crime that influenced attendance for participants. Additional factors influenced attendance for mentors and mentees. Mentors were less likely to attend if they viewed the income of their household to be sufficient, or if they scored high on the Stress Scale. Mentees were less likely to attend if they had a sister in the household, but more likely to attend if they were older or if the program site was in a location of high public program use. It is notable that other than stress, psychosocial factors did not predict attendance for the mentors and mentees in these mentoring programs. The results point to important areas for further research on attendance in cross-age youth mentorship programs, with the goal of increasing their impact.

Census tract and zip code variables for poverty, use of public assistance, and educational attainment were taken from the American Community Survey (ACS). Zip code level crime indices were taken from the SimplyAnalytics database. This dataset was merged onto the SLIY dataset by either zip code or census tract for each participant’s home address at baseline. In addition, these data were merged for each participant’s mentoring program site census tract or zip code. In short, these data are linked to both the place where a participant lived and to the mentoring site location.
•For all participants, traveling to a higher-crime program area was negatively related to attendance.
•Mentor variables related to attendance for mentors included stress, perceived family resources, and race.
•Variables related to attendance for mentees included age and having a sister in the home.
•Funding for programs in high-poverty, high-crime neighborhoods must budget for transportation.

The findings of this study illustrate that both geographic factors and family characteristics of the peer-mentoring program participants influenced the proportion of sessions they attended, and that factors for mentors and mentees are different.

For regressions with all participants, and for regressions run separately for mentors and mentees, the level of crime in the geographic area in which a program site was located was negatively related to attendance.

For mentees, in addition to crime at the program site location, age was positively related to attendance. Older mentees may have been able to travel to the program site independently, whereas younger mentees needed someone to take them (Deane et al., in preparation). Program staff knew that some mentees were escorted to the program by mentors, supporting their attendance but also indicating that it could be difficult for families to transport younger mentees.

It is important to also acknowledge the factors that did not predict attendance. The crime index in participants’ home neighborhood was not associated with attendance, nor was the rate of public program use such as SNAP, TANF, or SSI in households with children. Poverty in either the participants’ home or program site census tracts was not associated with attendance. Aside from self-reported stress for mentors, no other psychosocial measures were significantly related to attendance.

Implications for programs and practitioners


Policymakers and program providers can make use of these findings to enhance engagement by improving safety, providing resources, and carefully choosing locations within neighborhoods. Specific suggestions include:

1) Travel Safety:

•Engage youth as experts on safety planning for program attendees. Youth given that degree of respect have, in the past, been effectively protective of themselves, peers,  mentees, and even instructors (McCrea et al., 2019);
• Provide safe transportation (not just money but vans) to and from programming.
• Build cohorts of youth to accompany each other to and from programming, including on public transportation.
• Ensure young mentees have back-up plans for getting to and from the program site.

2) Resources:

• Include continual stress and trauma management in programming so that when programs are disrupted by drive-by shootings, etc., youth feel heard and comforted.
• Provide stipends for mentors that pay at least as much as minimum wage.
• Partner with a counseling program to provide outreach to highly stressed youth.

3) Program Location:

• Get community input. Consider the level of crime around potential program site locations relative to the larger neighborhood, and when possible, choose a site that participant families perceive to be safe and convenient.”