Skip to content
  • Academic publication
  • 23rd Mar 2017

Increasing social trust with an ice-breaking exercise – an RCT carried out with NCS participants


This paper reports the results of a small scale randomised controlled trial carried out by the Behavioural Insights Team in partnership with the National Citizens Service (NCS) and The Challenge, a charity that acts as a delivery organisation for NCS.

NCS is the largest youth movement for 100 years and in 2016 more than 80,000 young people aged 16-17 took part in the programme. This trial was conducted as part of the Summer 2016 NCS programme. Young people are put  into small groups with whom they will spend the four weeks of the programme, and with whom they are expected to complete a social action  project. As part of this trial,  50 such groups, consisting of 750 young people, were assigned at random to receive one of four sets of instructions at the beginning of the course:

  • One quarter of the teams were given no specific instructions as part of this experiment. These young people serve as a control group for this trial.
  • Another quarter of the teams were given an “ice breaker” task and asked to discuss their similarities as a group.
  • A third quarter of the teams were given a similar ice breaker, in which they were asked to discuss their differences as a group.
  • A final group of teams were asked to discuss their strengths and weaknesses as a team.

Participants were then surveyed four weeks later to ascertain their level of social trust, and were asked to complete a task intended to measure their creativity.

Our main findings are:

  • There is a statistically significant positive impact overall of the similarities exercise on general social trust, relative to the control group.
  • The effect of the similarities exercise is particularly concentrated on people with lower levels of social trust, with no effect at all on participants with only average levels of social trust. The effect on low trust participants is the equivalent of moving from the 25th percentile to the 50th for social trust.
  • None of the other exercise have statistically significant effects overall.
  • The differences exercise has a statistically significant and positive effect for participants who have high levels of social trust.
  • None of the interventions affect participants’ creativity.

These effects are modest in size, but substantial in comparison to the level of trust in the population. Although further research is needed, we believe that these findings represent a valuable step towards developing a suite of interventions that can reliably increase social cohesion amongst young people.