A systematic review by the Campbell evaluated the impact of 46 mentoring programs. It found significant results in terms of reducing juvenile delinquency, improving school performance, reducing repeat offenses, drug use, and aggressive behavior [1].
The Crime Reduction Toolkit platform classifies this type of intervention as generating reductions in the overall recidivism of young offenders. This classification is based on the fact that the meta-analysis carried out by the Campbell Collaboration review showed a statistically significant reduction in recidivism among young people who received mentoring compared to those who did not. However, the platform also notes that while 14 of the 25 studies included in the meta-analysis showed a significant reduction in recidivism, three of them showed a significant increase, and the other eight studies showed no significant effect on recidivism [1].
The Crime Solutions platform classifies this type of program as effective to reduce juvenile delinquency and promising in terms of reducing alcohol and drug use. There is also evidence of improvements in school attendance and performance, as well as in social skills and peer relationships [1] [2] [3].
According to the results of a fourth systematic review (older than those cited above), although when taken together the results suggest that mentoring reduces recidivism by approximately 4% to 10%, when only the studies of a higher methodological quality are considered, there is little evidence that mentoring reduces recidivism. This suggests that an inadequate control of the pre-existing differences between those who received and those who did not receive the treatment may have contributed to more positive results. In light of this finding, the authors argue that further high-quality evaluations will be necessary before the benefits of mentoring can be established with a greater level of confidence. However, the authors also found that mentoring was significantly more effective when implemented as part of a set of interventions, depending on the average duration of each contact between mentor and mentee [4].