1. There is no evidence that the intervention had any benefit in terms of pupils’ academic performance. Although the attainment data suggests a small negative impact, the quality of this data is too low to draw this conclusion with confidence. The data quality was compromised due to changes in national testing.
2. Participation in the intervention is associated with a small improvement in self-reported non-attainment outcomes including self-confidence and teamwork. It is possible that these small effects are an underestimate due to technical issues regarding the groups of children that were compared in the analysis.
3. For pupils eligible for free schools meals, there is no evidence that the intervention had a positive impact on academic attainment or self-reported character attributes. Again, the attainment data suggests a negative impact, but the quality of this data is too low to draw this conclusion with confidence.
4. Almost a quarter of schools did not deliver the intervention due to issues such as lack of teacher volunteers, and other schools did not deliver a full programme of activity. Support from senior leaders, dedicated space, school staff time, and a dedicated slot in the school day or after school were all identified as necessary conditions for successful implementation.
5. Study participants were extremely positive about the intervention and many felt it had a positive effect on the behaviour and skills of participating pupils.

In general, the analysis to assess the impact of the intervention reveals small ​‘effect’ sizes. In the case of non-attainment outcomes these are small and positive. This is the case for almost all of the non-attainment outcomes measures, as well as the two identified as primary outcomes (teamwork and self-confidence). In the case of attainment outcomes they are small and negative. However, for these attainment outcomes, the reliability of the finding is undermined somewhat by the quality of the data. For reasons beyond the control of the evaluator, the KS3 attainment data collected at the end of the trial may not have been consistent across schools. Further evaluation would be required to understand whether the intervention did indeed have a negative impact on attainment. Future research could also assess the impact of full exposure to the programme over two to three years, to give adequate time for the uniformed organisations to establish themselves.

The process evaluation revealed that pupils, teachers, and parents all had very positive views about the intervention, and in general believed that it had a positive impact on the pupils involved. It also highlighted a number of factors that had prevented the intervention from being delivered as planned in some schools, including a lack of dedicated space and time, a lack of adult volunteers (including teachers) to support the uniformed youth group delivery staff, and sometimes a lack of support from school senior leaders. Schools should consider these factors when deciding whether to implement a similar intervention.

The findings related to non-attainment outcomes have medium to high security. The trial was a large, well-designed, randomised controlled trial, with schools randomly allocated to either receive the intervention or not. The pupils from the intervention schools were similar to those from the comparison schools who received no intervention, and relatively few pupils were lost to the analysis due to issues such as moving school. However, the non-attainment outcomes were measured using a bespoke survey. This survey was well-designed and based on well-established test items, but as it was developed in its current form for the purposes of this particular evaluation, it may not be able to provide a standardised measure of non-attainment outcomes.

The findings related to attainment outcomes have low to medium security. Although the trial was well-conducted as described above, it transpired that the Key Stage 3 data which was to be used as an outcome measure would not be available. The alternative outcome measures that had to be used were based on scores from internal school records and were very weakly related to previous standardised test scores. Different schools also used different tests and marking schemes to derive the scores. This makes it harder to estimate reliably the size of the intervention’s attainment impact. A more reliable estimate of the impact on attainment will be available after these pupils reach the end of Key Stage 4 and additional national test data is available.

The costs vary depending on the number of outdoor activities and the size of the groups. Assuming 20 pupils per school, the cost is estimated at £18 per pupil per year for St John Ambulance, £26 for the Sea Cadets, £245 for The Scout Association, and £420 for the Fire Cadets — an average of about £180.