Skip Navigation
Impact Evaluation of the U.S. Department of Education's Student Mentoring Program

NCEE 2009-4047
March 2009

Program Delivery Findings

Both grantees and mentors were surveyed to describe various characteristics of program delivery, including training and support for mentors, characteristics of mentors, matching of students with mentors, and mentor/student relationship duration and activities. On average, grantees in the study implemented the program following the guidance provided by the legislation and program office. Also, program delivery was, by and large, consistent with findings from previous studies of schoolbased mentoring.

Key findings on program delivery include the following:

  • Approximately one in ten mentors reported not having undergone a reference or background check despite being required by the program as a condition of the grant. Eleven percent of mentors reported not having had either a background or reference check conducted pre-match, despite the fact that all 32 grantees indicated requiring some form of background screening before matching mentors with students. Because only mentors and not grantees were asked this question, it is possible that some mentors were simply unaware (or had forgotten) that a background or reference check was conducted by the grantee.
  • The majority of mentors received pre-match training or orientation and had access to ongoing supports from the program. Ninety-six percent of mentors reported receiving an average of 3.4 hours of some form of pre-match training or orientation. Forty-one percent of mentors reported that ongoing training was available after they had begun meeting regularly with their students. Ninety-four percent reported having access to some kind of ongoing supports, consistent with legislative and program guidelines.
  • The majority of students were matched with mentors of the same race and gender. Fifty-five percent of matches in our study were between individuals who had the same racial status. Eighty-one percent of matches in our study were between students and mentors of the same gender.
  • The majority of mentors met with their students on a one-to-one basis. Mentors, on average, also reported meeting with their students, on a weekly basis for approximately one hour per meeting. This finding is consistent with findings from other studies (Herrera et al. 2007; and Karcher, 2008).
  • Seventeen percent of the students randomly assigned to the treatment group never received mentoring from the program. This includes 14 percent of students in the treatment group who were never matched with mentors and another 3 percent who were matched with mentors, but never actually met. However, the percentage of unmatched students in this study is within the range of past experience engaging mentors in randomized impact studies of mentoring.6
  • On average, the programs took a total of 81 days to match students and mentors, from the start of the school year. On average, there was a lag of 37 days between the date of random assignment and the time when the student was matched. This lag between the beginning of the school year and matching students is consistent with findings from previous research (Herrera et al, 2000, Hansen, 2005, and Karcher, 2008).
  • For students who were matched and met with their mentors, the average length of the relationship was 5.8 months. This finding, however, is consistent with previous research. For example, Herrera et al. (2007) in the impact study of the BBBS schoolbased mentoring program reported an average match length of 5.3 months.
  • Discussing relationships and future plans, and to a lesser extent, working on academics were the most frequent activities reported by mentors. Mentors and students worked together on a range of activities. Approximately half of the mentors reported frequently discussing relationships and future plans (52 percent and 48 percent, respectively). In contrast, 43 percent reported working frequently on academics, while 21 percent reported never working on academics. The greater focus on the social needs of the students compared to academic needs has been found in some, but not all, of the previous research.7
  • Approximately 20 percent of the mentors were of high-school age (18 years or younger) and an additional 23 percent were of college-age. However, this is still a smaller percentage than findings from previous research where the majority of mentors were of high school or college age.8


6 One school-based mentoring study and one community-based mentoring study (both random assignment of students to conditions) has found that the proportion of students slated to receive mentoring services that remain unmatched with mentors was 7 and 22 percent, respectively (Herrera, et al., 2007; Tierney and Grossman, 2000).
7 Less than a third of the mentors in the BBBS study (Herrera et al., 2007) reported spending a lot or most of their time on academic activities.
8 For example, Herrera et al. (2007) reported that half of the mentors in the BBBS study were 18 years old or younger, with an additional 17 percent 19 to 24 years old. In Karcherís 2008 study of school-based mentoring, 70 percent of the mentors were college students.