Mentoring Multi-Site Trial

What does this project involve?

This project used a multi-site trial (MST) to test the efficacy of 12-weeks of mentoring, delivered by 17 small organisations to 10-17 year olds at-risk of involvement in violence.  

The model of mentoring aimed to provide each child with a minimum of 12 one-to-one mentoring sessions, of at least 45 minutes each, from a paid, trained adult.

Key components of the mentoring included an initial meeting to be held prior to the start of mentoring, a written plan used during mentoring, and a defined closure process at the end of mentoring. Children’s participation was voluntary.

Why did YEF fund this project?

Most rigorous impact evaluations of youth sector provision focus on well-defined, manualised programmes delivered by single organisations. These evaluations require large sample sizes that are often only achievable by large organisations. This is likely to exclude small, community-led, organisations from participating in evaluations. Smaller organisations may be more likely to be led by individuals from Black, Asian and ethnic minority communities, and they may also provide specialist services not offered by larger organisations. In addition, most services provided to young people in England and Wales take place in small, local organisations and consist of non-manualised approaches. This creates a gap between the evidence and everyday practice.

This project aimed to test an approach to closing this gap, using a Multi-Site Trial (MST) with smaller organisations to test the effectiveness of a short-term model of mentoring. An MST combines the reach of multiple smaller organisations (Delivery Partner Organisations or ‘DPOs’), to create a sample large enough for an impact evaluation. Following a pilot study (that showed the MST approach to be feasible and acceptable when delivered by 9 DPOs), this efficacy study tested an approach to short-term mentoring delivered by 17 DPOs.

As our Toolkit demonstrates, mentoring has, on average, demonstrated a moderate estimated impact on reducing children’s involvement in violence; however, we lack evidence of impact from a UK context.
This evaluation, therefore, aimed to ascertain the impact of a short-term, shared-practice model of mentoring delivered by small organisations, on behavioural and emotional problems amongst at-risk 10-17 year olds (as measured by the self-reported Strengths and Difficulties Questionnaire (SDQ)). Experiencing these problems is associated with later involvement in crime and violence.

The evaluation included a wait-listed randomised controlled trial (RCT) involving 744 children from across England and Wales. 372 received mentoring immediately; 372 were assigned to the waitlist control group, and received mentoring after the evaluation. 63% of the children involved were White; 15% were Black, Black British, Caribbean or African; 11% Mixed or multiple ethnic groups; 7% Asian or Asian British; and 4% identified as ‘Other’.

The evaluation also featured an implementation and process evaluation comprising of a mentee feedback survey (with 299 mentees), mentor survey (with 26 mentors) and in-depth interviews with 16 DPO managers, 16 mentors and 19 mentees. These qualitative methods explored the implementation and feasibility of the shared practice model, the barriers and enablers to delivery, and mentors and children’s perceptions of the model. The trial began in February 2023 and concluded in March 2024.

Key conclusions

How secure is the evidence?

These findings have a moderate security rating (3/5 magnifying glasses). The trial was a well-designed RCT. The trial was large enough to detect meaningful impacts and larger than most previous evaluations of mentoring. 23.7% of the children who started the trial were not included in the final analysis, as they did not complete the follow-up survey. We do not know if the effect found for short-term mentoring would be the same if the children missing from the final analysis were included. The children in the intervention group were similar to those in the control group.

Should you still deliver mentoring?

The findings from this evaluation are only the findings from one study. Although the primary outcome finding shows no impact on reducing children’s self-reported behavioural and emotional problems, the international evidence base on mentoring still suggests that mentoring is a good bet for protecting children from involvement in violence. The secondary outcome findings also indicate a range of other positive outcomes that mentoring may have on children.

The study may suggest that mentoring should be delivered for longer than 12 weeks (and for at least 6 months as suggested by the YEF Toolkit). This will be tested in a range of other evaluations that YEF is conducting on mentoring programmes.

What will YEF do next?

YEF have no current plans for further evaluation of this mentoring model.

Download the report