Mentoring High-Risk Minority Youth: Evaluation Of The Brothers Project

Adolescence, Spring, 1998 by David Royse

ABSTRACT

Mentoring as an intervention for at-risk teens is becoming increasingly popular despite sparse evidence of its effectiveness. This research, part of a larger evaluation effort, reports on a four-year mentoring project developed specifically for African-American adolescents. Self-esteem, attitudes toward drugs and alcohol, grades, school attendance, and disciplinary infractions were examined using an experimental design. No significant differences were found between the control and intervention groups. However, multiple explanations are offered to account for why it is so difficult to document the positive benefits of mentoring.

Although it has been noted that mentoring is increasingly viewed as an intervention for youth considered at-risk, vulnerable, or likely to be unprepared for adult living (Mech, Pryde, & Rycraft, 1995) and as a way to fortify inner-city youth against delinquency, school dropout, teen pregnancy, unemployment, and other negative life situations (Blechman, 1992), formal matching of adult volunteers can be traced to the first Big Brothers agency in 1902 (Morrow & Styles, 1995). Despite this history, few empirical studies have documented the benefits of mentoring on the lives of young people. Big Brothers/Big Sisters, for example, have been providing adult support to youth from single-parent households for over 90 years without evidence that "conclusively demonstrated that youth who participate in Big Brothers/Big Sisters programs fare better than they would have had they not participated" (Furano, Roaf, Styles, & Branch, 1993, p. 6). While Big Brothers/Big Sisters has begun a four-year research initiative, the mo st recent report on this project stated, "It remains to be seen whether developmental mentoring relationships can produce real changes in the lives of youth, such as improved grades and more positive behavior" (Morrow & Styles, 1995, p. ix). Slicker and Palmer (1993) have also observed a "clear lack of research on the effects of a mentoring relationship with low-achieving potential high school dropouts" (p. 328).

While sparse, there are a few studies relative to mentoring. Frecknall and Luks (1992), for instance, reviewed surveys from 76 parents whose children participated in a Big Brothers/Big Sisters program in New York City. Sixty-three percent of the parents reported that their children "greatly improved" in attitudes or behavior. The authors reported a trend of greater parental perception of success when children spent a greater amount of time in the program and when parents had more contact with the Big Brother/Big Sister.

On the other hand, Nelson and Valliant (1993) did not find any statistically significant differences in the self-esteem of boys awaiting assignment of a Big Brother (n = 6), those from intact middle-class two-parent families (n = 27), those participating in Big Brothers/Big Sisters for at least three years (n = 9), and a group of young offenders residing in an open-custody facility (n = 18). However, depression scores for those awaiting a Big Brother and those in the group home were more elevated than the scores of boys from intact families and those involved with a Big Brother for at least three years. Boys in the young offenders group and those with no adult male substitutes scored significantly higher on assaultiveness than did boys with adult male substitutes.

Most students in Milwaukee's One-on-One program did not show improvement in grades during the academic year although that was a primary objective, and Baltimore's RAISE project found that students remained far below average in academic performance and above average in risk of dropping out. RAISE researchers concluded that gains in school attendance and grades were not "sufficient to eliminate the academic risks with which students entered the program" (as cited in Freedman, 1993, p. 83).

Slicker and Palmer (1993) studied 32 at-risk 10th-grade students who were paired with a mentor recruited from school personnel and 54 at-risk students in a control group. At the end of six months, there were no significant differences in drop-out rates between the two groups. While there was a significant difference in self-concept scores, it favored the control group. However, the authors concluded from the mentor logs that the quality of mentoring varied greatly. When they divided the students into those who received effective mentoring and those who were ineffectively mentored, the effectively mentored group demonstrated an increase in grade point average that approached statistical significance (p = .06). The authors concluded that "the key to an effective mentor program. ... is the close monitoring of the integrity of the treatment used" (p. 333).

The literature also suggests another explanation as to why mentoring programs produce mixed results. Hamilton and Hamilton (1992) observed that mentors often have different views of their purpose. What they called Level 1 mentors saw their primary purpose as developing a relationship. Level 2 mentors spoke of introducing options as their major purpose. Level 3 mentors stressed developing character, while Level 4 mentors focused on developing competence. Mentors who viewed their purpose as relationship-building worried about whether they were liked and if they had chosen the right activities. Hamilton and Hamilton concluded that an emphasis on learning to do something produced the most functional pairs.