Incentive Value of Social Signals in Typical Development and Autism.

Background:  It has been proposed that typical development of social orienting is based on the learning of the association between ‘engaging’ social signals, such as positive emotion and gaze cueing, and subsequent positive outcomes such as an enjoyable interaction or the opportunity for social learning. By contrast, atypical social orienting in autism could result from the difficulty to learn such association, resulting in not assigning the reward value to 'engaging' social stimuli.

Objectives:  A novel interactive eye-tracking task was developed to investigate visual orienting towards social signals associated with a rewarding stimulus in typical developing toddlers, children and adults as well as in children with autism.

Methods:  A group of typically developing toddlers (n = 32, 3-4 years old) and a group of typical adults (n = 32, 18-45 years old) (study 1) as well as two groups of typical and autistic children (n = 58, 6-18years old, study 2) completed the interactive eye-tracking task. The participants observed a stimulus display consisting of two peripherally presented dynamic social signals (faces) and a centrally presented reward. Participants' eye movements were concurrently recorded with an eye-tracker, and the location of participant's fixation triggered the delivery of corresponding stimuli on-line. Fixation on each face triggered a dynamic sequence of signals and subsequent delivery of a reward, which was a popular animated cartoon, or a penalty, a blank screen. Two types of social signals were presented. An engaging social signal consisted in a person greeting and turning towards the centre of the screen while the other non-engaging social signal consisted in a person moaning and turning away from it. Engaging social signals triggered reward delivery for half the participants of each group, and non-engaging social signals for the other half of the participants of each group.

Results:  Preliminary analyses revealed that both typically developing toddlers and adults (study 1) and both groups of children with and without autism (study 2) were able to learn the association between either of the social signals and the subsequent reward delivery, and preferentially orient to the reward-predictive cues. Importantly, all the groups acquired stronger preference when the reward-predictive social signal was engaging than when it was non-engaging.

Conclusions:  These preliminary results show that typically toddlers, children and adults as well as children with autism were able to learn the reward value of social signals (either engaging or non-engaging) when such signals were associated with a rewarding stimulus. Moreover, these studies showed that the engaging nature of social signals facilitates the learning of social signal-reward association both in typically developing children and in adults, as well as in children with autism. The implications of these findings are discussed.