Emotion recognition and expression in autism spectrum disorder : significance, complexity, and effect of training

• The overall scientific aim of this thesis is to examine emotion recognition in children with Autism Spectrum Disorder (ASD), its specificity to ASD and connection to other cognitive functions, as well as to map the effects of previous and novel emotion recognition and emotion expression training programs in children with ASD across cultures. Emotion processing training in ASD is a potentially valuable tool to improve the lives and outcomes of children with ASD, but it has been lacking an adequate scientific preparation, content and motivational design, as well as high-end technical expertise. In study I, we review the existing randomized controlled trials on emotion recognition training for children and adolescents with ASD, focusing external validity, a largely aspect in the area. External validity is significant to evaluate for several reasons. First, emotion recognition training approaches have been diverse, and ASD forms population a heterogeneous population. Second, children and adolescents with ASD often have difficulties with generalizing knowledge from the training context to new situations, and it is not clear how this is reflected in the results of the various emotion recognition training programs. Third, emotion recognition training is often performed using computerized programs in controlled settings, raising questions about the extent to which the effects translate to everyday situations. The systematic review demonstrated few indications to presume that current emotion training programs generalize outside of the training setting into everyday life social interactions. This review highlights the need to focus on external validity in future emotion recognition training studies, and to improve reporting of these aspects. Study II examined basic facial affect recognition in well-matched samples ofchildren with ASD, Attention-Deficit Hyperactivity Disorder (ADHD) and typical development using the computer-based FEFA-2 test. We examined accuracy and response times for general and specific facial affect recognition skills in whole face and eye-region stimuli. The ASD samples performed inferior to typical developing controls. There were no difference between the ADHD sample, on one hand, and the ASD and typical sample, on the other. In the clinical samples, particularly the ADHD sample, cognitive distractibility explained a substantial proportion of variance of basic facial affect recognition performance. This research largely confirms previous findings on emotion recognition in ASD, and aspects of specificity compared to ADHD, a neurodevelopmental condition overlapping with ASD. Importantly, the study shows that performance on emotion recognition tasks is not only depending on pure emotion recognition capacities, but is also largely influenced by other cognitive functions. Study III explored cultural differences in emotion recognition in children with ASD and typically developed children. Compared to many previous studies, differences in recognizing emotion expression was tested in three modalities of basic and complex emotion processing, namely in face, voice, and body language including gestures. These expressions were also examined in integrated form using complex social scenes. The study was conducted across three countries and cultural contexts: Israel, United Kingdom and, and Sweden. Children with ASD showed impairments in both basic and complex emotions on all three modalities and their integration in context compared to typically developing matched control children. Both children with ASD and typical development performed better on basic than complex emotions. Cross-cultural differences were limited to some face and body stimuli, indicating high cross cultural comparability of emotion recognition findings. Study IV included a cross-cultural evaluation of the effects of the serious game “Emotiplay” in ASD on emotion recognition skills, and parent-reported autism symptomatology and adaptive skills in United Kingdom, Israel and Sweden. Emotion recognition tasks comprised face, voice, body, and integrative social scenes. Children used Emotiplay 8-12 weeks in a home setting. In the United Kingdom children were tested prepost, while children in Israel and Sweden were randomized to training or a waiting-list control group. Results showed improvements in emotion recognition regarding body language and integrative tasks as well as adaptive socialization in the United Kingdom site. In Israel and Sweden, the active groups improved more than controls on all emotion recognition outcomes. There was also an effect on autism related symptoms in the sample from Israel. Findings support the feasibility and usefulness of serious gaming to enhance emotion recognition and possibly reduce autism symptoms and socialization issues in autistic children. Emotiplay was found useful across three cultures. Still, future research and follow-up studies are needed to determine long-term effects and evaluate its impact on real life situations.

Publication and Content Type

vet (subject category)

dok (subject category)