부산대학교 도서관

Perhaps most interesting within autism research is the focus on emotion processing and facial emotion recognition (FER) specifically because difficulties in recognising and responding appropriately to others emotions are part of the diagnostic criteria of autism outlined in the DSM-V (American Psychiatric Association 2013). The research surrounding emotion processing abilities in autism has shown conflicting results (Harms, Martin & Wallace 2010). The thesis presented here aims to explore emotion processing in autism with the use of new stimuli that better reflects the demands of everyday life by simultaneously incorporating tasks that require a greater understanding of socially appropriate information. Chapter 1 introduces autism in more detail and covers the theories that attempt to explain its aetiology, before briefly introducing emotion processing. Chapter 2 expands into the research area of the thesis covering theories of emotion and emotion processing research in much greater detail, both generally and within autism including a review of current methodologies used to explore FER in autism. Chapters 3 and 4 report the development and validation of the new stimuli set. The completed stimulus set contains 133 posed expressions and a matching 133 spontaneous expressions captured from 19 individuals in response to seven social prompts. Analysis of expression formation through FACS data with principle component analyses showed consistent display patterns were more frequent across spontaneous expressions than posed. In chapter 4 results showed significantly better recognition of spontaneous expressions than posed, participants could identify the correct answer significantly more than alternate options for six of the seven expressions. Participants also gave systematic emotion state attributions to those responses significantly more than chance would predict. Chapter 5 reports the results of displaying a subset of expressions with better recognition rates to typically developing children, and recording their visual attention patterns with an infra-red eye tracker. As was the case in chapter 4 differences in spontaneous and posed expressions extended to viewing styles; significantly more fixations were made to posed expressions and significantly more time was spent fixated to posed expressions but this didn't appear to influence recognition rates where no differences 3 were present between posed and spontaneous expressions. The viewing styles in combination with performance on a retrodictive mindreading task significantly predicted autism traits. In chapter 6 differences between an ASC sample and typically developing control sample on a retrodictive mindreading task were investigated. No group differences on overall task performance were shown; this was true for both the retrodictive mindreading task and the reading the mind in the eyes task. The previous effect in chapter 4, of spontaneous expressions recognised significantly better than posed was repeated, both groups also made significantly more consistent emotion to retrodiction attributions than a chance model would predict. Investigation of the role of Alexithymia showed no effect of Alexithymia severity on retrodictive mind reading ability. Investigating the role of alexithymia within the ASC sample showed no association between alexithymia sores and retrodictive mindreading ability. The results are discussed in respect of three themes in the thesis. Firstly the impact of posed and spontaneous expressions are considered in expression production (chapter 3), viewing styles (chapter 5) and recognition (chapters 4 and 6); suggesting that spontaneous expressions elicited during a real life social interchange under the methods outlined in this thesis are quantitatively different to posed expressions, are viewed differently, and have different recognition rates. Secondly the impact of Alexithymia on expression production and recognition was also considered. As an emerging theory that attempts to explain the emotion and social difficulties in ASC, Alexithymia's comorbidity was an important finding throughout this thesis, its role is discussed specifically its influence over retrodictive mindreading performance but not eyes task performance. Finally the results are discussed in respect of FER in autism. The lack of a significant difference between groups could have implications for the topic; many of the inconsistent results rely on different methods some of which are more artificial and lack validity of emotion processing in a real world sense. The discussion continues considering the limitations of the research, potential practical applications and future research directions.