Intuitive Cooperation in Autism

Thursday, May 11, 2017: 12:00 PM-1:40 PM
Golden Gate Ballroom (Marriott Marquis Hotel)
M. Brosnan1 and C. Ashwin2, (1)Centre for Applied Autism Research, University of Bath, Bath, United Kingdom of Great Britain and Northern Ireland, (2)University of Bath, Bath, United Kingdom of Great Britain and Northern Ireland

Dual Process Theory of Autism (Brosnan et al., 2016) proposes that slower, deliberative (Type 2) processing dominates in people with a diagnosis of ASD, whereas rapid, intuitive (Type 1) processing typically dominates in the general population. Rand et al. (2012) propose that cooperation, as assessed through a ‘donating’ task, is enhanced through Type 1 processing (e.g. when forced to make a fast decision) and diminished under Type 2 processing (e.g. with more time to deliberate on the decision).


There were three objectives: 1) To replicate the ‘fast-cooperation’ effect; 2) To examine if autistic-like traits in the general population relate to cooperation (assessed through amount donated) and the fast-cooperation effect; and 3) To examine if a group on the autism spectrum differed from a typically developing (TD) control group on cooperation (assessed through amount donated) and the fast-cooperation effect.


1) 154 participants from a university open day volunteered to undertake a 5-minute online task, and were randomly allocated to either a fast (decide within 10 seconds) or slow (think about the decision for at least 10 seconds) condition. The decision was to identify how much of 40p would be donated to a collective pot to be doubled and distributed equally amongst all group members (and how much was to be kept for themselves, see Rand et al., 2012). 2) This task was repeated with 53 (33M, 20F), 17 year olds, who also undertook an assessment of autistic-like traits. 3) This task was repeated with 29 people with ASD (20M, 9F) and 32 TD controls (17M, 15F) aged 17.


1) Mean Fast condition (n=83) = 29.35 (SD 10.60); Mean Slow condition (n=71) = 26.06 (SD 13.00); t(134.991)=1.71, p<.05; Cohen’s d = .28. The effect was significant, but small (around 8%), and in line with prediction.

2) A Linear Regression revealed that greater autistic-like traits were a significant predictor of a lower amount donated (Beta= -.34, p<.02) but condition (fast/slow – the effect was 5%) and sex (male/female) were not (both p>.05).

3) An ANOVA with amount donated as the dependent variable with Diagnosis (ASD/TD) and condition (Fast/Slow), and sex as a covariate revealed no significant relationships (all p>.05). Examining the insignificant trends, the ASD group donated around 8% less than the TD group in both conditions, see Figure 1.


There is a relatively consistent effect in the manipulation of co-operation (around 8%) which may be affected by the speed of presentation (consistent with Rand et al.) or level of autistic-like traits (consistent with Brosnan et al.) – but these variables do not interact with each other. It would seem that those with ASD are as susceptible to the fast-cooperation effect as the TD group. Taken as a whole, the patterns in the data would suggest that higher autistic-like traits (including a diagnosis of ASD) are associated with less cooperation (8%, as measured by this task) and are as equally susceptible to the effects of time pressure as comparison groups.