Do the Spatial and Kinematic Properties of Facial Expressions Influence Emotion Recognition in Autism Spectrum Disorders?

Autism spectrum disorder (ASD) is a neurodevelopmental disorder, characterized by difficutlies in social communication, and restricted and repetitive interests (American Psychiatric Association, 2013). Given that the ability to infer emotion from facial expressions is crucial for social interaction, emotion recognition has long been suspected to be a core difficulty within ASD (Hobson, 1986). However, whilst many studies suggest a disparity in the facial expression recognition ability of autistic and neurotypical individuals (Ashwin, Chapman, Colle, & Baron‐Cohen, 2006; Dziobek, Bahnemann, Convit, & Heekeren, 2010; Lindner & Rosén, 2006; Philip et al., 2010), there have been inconsistent findings, ranging from no differences between autistic and neurotypical individuals to large disparities (see Keating & Cook, in press, and Uljarevic & Hamilton, 2012 for reviews). Despite these inconsistencies, the evidence largely suggests that there are differences in facial expression recognition between autistic and neurotypical individuals. Only recently has there been a shift towards using dynamic depictions of faces (which include kinematic information) as opposed to static stimuli to assess facial expression recognition. Indeed, studies that employ dynamic tasks are profoundly under-represented in meta-analyses investigating the emotion recognition of autistic individuals (Keating & Cook, in press.). This may be problematic given that naturally occurring facial expressions are inherently dynamic, and therefore may have greater ecological validity (Uljarevic & Hamiltom, 2013; Krumhuber, Kappas & Manstead, 2013). Amongst the few studies that have examined the influence of facial movement kinematics on emotion perception, there is a consensus that there are differences between autistic and neurotypical individuals. For instance, Sato and colleagues (2013) found that autistic children were more likely to rate slow-moving morphs as ‘natural’ looking than neurotypical children. Another study had a professional actress display emotional (joy, surprise, sadness and disgust), and non-emotional (porunication of A, O, I, and tongue protrusion) expressions slowly (Tardif et al., 2008). The speed of these videos were then manipulated to give three conditions (very-slow, slow, and normal). Importantly, in their post-hoc analyses, Tardif and colleagues (2008) identified that the autistic individuals had superior emotion recognition in the slow relative to the normal speed condition. The findings of these behavioural investigations resonate well with those from neurophysiological studies. Generally, this literature suggests that autistic individuals, relative to neurotypicals, exhibit slowed processing of static face images as indexed by N170 latency (see Kang et al., 2018 for a summary). As a whole, the evidence suggests that autistic individuals may experience improved emotion recognition for slow moving faces. However, more research employing dynamic stimuli is necessary to confirm the assertion that autistic individuals have better emotion recognition for slowed facial stimuli. A parallel literature concerns emotion recognition from body movements. Here it has been suggested that kinematic aspects of bodily movement contribute to emotion recognition, and may underpin the differences in emotion recognition between autistic and neurotypical individuals. Indeed evidence of this comes from studies using point-light displays (PLD)- a set of moving dots that convey biological motion. One study manipulated these PLDs in terms of acceleration and had participants rate the naturalness of the displayed movements (Lee & Chang, 2019). This study found a robust association between performance on this biological motion naturalness task and attention switching domain scores on the AQ (Lee & Chang, 2019). Therefore, it seems that kinematics may be implicated in the emotion recognition differences between those high and low in ASD traits. Indeed, recent developments in the face processing literature emphasize the importance of kinematic cues in emotion recognition. One study demonstrated this by utilisng point-light displays of the face (known as PLFs) which had been manipulated to achieve three spatial levels (S1 – 50% spatial; S2 – 100%; S3 – 150%) and three kinematic levels (K1 – 50% speed; K2 – 100%; K3 – 150%) (Sowden et al., under review). This study revealed that intensity ratings, given by neurotypical participants, were modulated as a function of both spatial and kinematic cues. Specifically, at the S1 and K1 levels (i.e. with less spatial movement/ lower speed), participants rated the PLFs as more intensely sad, and less intensely angry and happy, and at the S3 and K3 levels (i.e. with more spatial movement/ greater speed), they rated the PLFs as less intensely sad, and more intensely angry and happy (Sowden et al., under review). This novel PLF task has great utility for investigating facial expression recognition as it eliminates contrast, texture, colour and luminance cues (e.g. a flushed face or tears), and other information (e.g. identity) that can be seen in photograph stimuli. Hence, this task allows us to investigate the importance of the kinematic and spatial aspects of facial expressions without these factors confounding the results. To the best of our knowledge it is also the first task which can index independent contributions of both spatial and kinematic cues in facial emotion recognition. Therefore, the present study will utilise dynamic face stimuli, that have been manipulated kinematially and spatially (Sowden et al., under review) to identify whether, compared to neurotypical individuals, those with ASD exhibit differences in processing the kinematic and/or spatial properties of facial expressions. Alexithymia When discussing the facial expression recognition of autistic individuals, it is also crucial to consider the role of alexithymia. Alexithymia is a subclinical condition characterized by difficulties identifying and expressing emotions (Kooiman, Spinhoven & Trijsburg, 2002). Whilst the incidence of alexithymia in the neurotypical population is 13% (Salminen et al., 1999), in the autistic population these rates are elevated, with 40- 65% of autistic adults meeting criteria (Berthoz & Hill, 2005; Hill, Berthoz & Frith, 2004). Whilst some research has suggested that autistic individuals exhibit differences relative to neurotypicals in recognizing others’ emotion, there are considerable individual differences – not everyone with an autism diagnosis exhibit face processing differences (Harms, Martin & Wallace, 2010; Uljarevic & Hamilton, 2012). It has been proposed that alexithymia, and not autistic traits, may account for the reported impairments in facial emotion recognition, and this could explain the inconsistencies in previous research (Bird and Cook, 2013). Indeed the “alexithymia hypothesis” (Bird and Cook, 2013) postulates that alexithymia underpins the individual differences in emotion processing that are seen in the ASD population such that intact emotion recognition is typical in individuals that have an autism diagnosis without co-ocurring alexithymia . Indeed, this hypothesis is supported by the finding that there are no differences in facial expression recognition between autistic participants and neurotypicals when the two groups are matched in terms of alexithymia (Cook, Brewer, Shah, & Bird, 2013). Moreover when the variables of this study were examined continuously, it was found that possessing alexithymic traits, but not autistic traits, was predictive of poorer performance on emotion recognition and empathy tasks (Cook, Brewer, Shah, & Bird, 2013). Thus overall, it appears that alexithymia is heavily implicated in differences in facial expression recognition that are commonly (but not universally) documented within ASD populations. The current study First participants will complete various questionnaires including the Autism Quotient (AQ; Baron-Cohen, et al., 2006), and the Toronto Alexithymia Scale (TAS; Bagby, et al., 1994). Next, participants will complete Sowden et al’s (under review) facial expression recognition task. In this task, participants will rate PLF (point-light face) stimuli regarding the extent to which they appear happy, angry and sad. These PLFs vary in emotion (happy, angry and sad), and have been adapted to achieve three spatial movement levels (S1 – 50% spatial movement; S2 – 100%; S3 – 150%) and three kinematic (speed) levels (K1 – 50% speed; K2 – 100%; K3 – 150%). Then, participants will complete two adapted PLF tasks in which they rate the naturalness of facial expressions that have been manipulated spatially and kinematically. Finally, participants will complete the Matrix Reasoning Item Bank (MaRs-IB; Chierchia et al. 2019), which assesses non-verbal reasoning.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.