The Impact of Autistic Traits on the Affective Prosody and Gesture Perception in Conversation

Background: Integration of speech and gesture is an essential ability that helps humans to participate in daily communication and social interactions. However, deficits in communication and social reciprocity are primary aspects of Autism-Spectrum Disorders (ASD) and several studies show that individuals with ASD have more difficulty in integrating audiovisual speech stimuli (Stevenson et al., 2014) and discriminating temporal asynchronies in linguistic situations (Bebko et al., 2006). The deficits in speech and gesture integration in ASD may reflect processes of multisensory integration that impact on social and emotional communication.

Objectives: This study is aimed to investigate the differences of emotional speech and gesture processing between individuals with low levels of autistic traits and high levels of autistic traits.

Methods: A total of 64 right-handed native English speakers between 18 and 40 years (25.20±6.11) participated in this study. The participants were divided into two groups: one group was a low autistic traits group (M:10, F:30) and one group was a high autistic traits group (M:8, F:16). We used a cutoff of below 18 to enter the Low AQ (LAQ) group and above 29 on the Autism-spectrum Quotient (Baron-Cohen at al., 2001) score to enter the High AQ (HAQ) group. Participants were presented with dyadic interactions and asked to respond about speakers’ emotions from prosody (Task1&Task2: Audio-only), gesture (Task3: Visual-only), and with either emotionally congruent or incongruent displays (Task4: Audio-visual). For the audio-only and the video-only tasks, the stimuli included three emotional categories (Angry, Happy, Neutral). For the audio-visual task, stimuli included happy and angry emotions that were either congruent or incongruent across modality (Piwek et al., 2015).

Results: In the results of the prosody experiment, the HAQ group showed significantly lower accuracy than the LAQ group in Task2 (F(1,186)=4.559, p=.034). There was a significant emotion effect in both tasks (F-test, all p<.001). The HAQ group showed lower accuracy in recognising the happy emotion in both Task1 and 2. An interaction between group and emotion was found in Task2 (F(2,186)=5.867, p=.003).

In the gesture experiment, the HAQ group exhibited significantly lower accuracy than the LAQ group (F(1,186)=11.272, p=.001), and a significant main effect of emotion (F(2,186)=260.001, p<.001). Also, there was an interaction between group and emotion (F(2,186)=12.135, p<.001). In both groups, the angry gesture had the lowest percentage of correct recognition (LAQ: 42.9%, HAQ:40.1%).

In the congruence experiment, there was no group difference. A significant congruence effect was found (F(1,124)=150.116, p<.001) and a marginal group by congruence interaction (F(2,124)=3.855, p=.052). In reaction times, the HAQ group reacted slower than the LAQ group in all tasks, in particular a significant group difference was found in the congruence experiment (t=-2.139, p=.036).

Conclusions: This study investigated how autistic traits impact on affective prosody and gesture perception. The results of this study show that significant performance differences exist between low and high autistic trait individuals with the HAQ group presenting lower accuracy in recognising affective prosody and gestures. It suggests that consideration of these differences can contribute to understand the impact of autistic traits on social communication.