Computer-mediated communication in autism

The aim of this thesis was to examine linguistic and social processing in autism and Asperger syndrome (AS), through computer-mediated communication. The first investigation used conversational analysis, on a corpus of computer-mediated dialogue, generated by two adults with AS. The results revealed that one of the two individuals had problems asking questions. Hence, an inability to ask questions may be one aspect of AS communication, though it may be not universal in this population. The second study used a computer program called Bubble Dialogue (Gray, Creighton, McMahon & Cunningham, 1991) to investigate the working understanding of nonliteral language and responses to inappropriate requests in individuals with AS and high-functioning autism (HFA). The AS/HFA group showed poorer understanding of a figure of speech and were more likely to consent to socially inappropriate requests compared to their typically developing peers. In contrast, understanding of sarcasm was predicted neither by verbal ability, executive ability nor clinical diagnosis. The results suggest that having AS/HFA does not, a priori, dispose someone to having problems with communication and socialisation, and that verbal ability protects the individual to a certain extent. Additionally, executive ability also seems important in mediating socialisation and communication ability. The third experiment tested the hypothesis that an autistic preference for internet-based communication may be due to the absence of verbal and non verbal cues, physical distance, and slower rate of information exchange through that medium. To test this, participants worked out predetermined map routes by asking the experimenter closed questions either via text chat, or through telephone conversations. An initial examination of the results suggested that AS performance may in fact have been better via the telephone. However, a detailed look at the strategies employed by some individuals with AS suggests that their executive problems may have resulted in their use of a less than systematic way to solve the task in both media. The results of this study also indicate a relation between executive and mentalising ability because both are required to solve the task. Interestingly, many of the participants with AS could generate novel closed questions to successfully solve the map task in both media, though they were slower than controls. Using computer mediated communication has therefore given us greater detail into the nature of, and the factors that influence, communication in autism