I have just finished a six week, LuCiD summer internship at The University of Manchester. The project looked at category typicality and its role in how children with and without ASD store semantic information, and was supervised by Katie Twomey and Alexandra Sturrock.
Background and Aims
Children with Autism, even those without learning difficulties, have been shown to process and use language in different ways to their typically developing (TD) peers. One area of investigation is into differences in semantic (word meaning) skills and in particular, semantic features associated with category membership. Research suggests difficulties here are caused by the autistic child’s focus on specific, localised details, and inability to process information globally, meaning they struggle with more abstract concepts like category formation (Klinger & Dawson, 2001; Fiebelkorn et al, 2013). However, little research has directly assessed the structure of the categories children with autism learn, and in particular, whether children with autism are more likely to learn words for unusual, or atypical, category members (Gastgeb et al., 2012). For example, car is a very typical example of the category TRANSPORT, whereas a unicycle is a very atypical example.
To better understand any link between Autism and category typicality, a word association task (i.e., “name as many animals as you can in one minute”) was carried out as part of a larger PhD study. This study investigates language differences between school-aged children with Autism, their TD peers, and the effect of gender on language processing within both groups (Sturrock et al, 2019). The word association task was included four categories (animals, food, occupations and emotions) and a word list of nearly 1200 unique words was generated.
To determine how typical the words produced were, we would normally use existing typicality ratings, which are usually on a scale of one to seven (Djalal, Ameel & Storms, 2016). However, some of the words in this word list were so unusual that a typicality rating did not exist for them. I was therefore asked to collect new ratings using an online questionnaire for each of the 1200 words as the main aim of this internship. By having these ratings we would be able to establish if children with Autism were more likely to produced atypical examples of categories; if so, this would further support the theory that this group store and use categories in a different way to their TD peers.
First, I adapted the data from the original PhD study so that it could be input onto questionnaires. This process also required me to make supervised methodological decisions, which was a thought-provoking insight into how to organise the plan for a study and the importance of not rushing decisions.
To ensure the items were fairly distributed amongst the participants, we randomised the word list within categories and then Latin square counterbalanced the order of the categories. The total word list was then split across eight different questionnaires, to make it more realistic for a volunteering participant to complete. I conducted a brief overview of previous typicality research and this highlighted that we would need at least twenty five people to rate each word to give us a meaningful average typicality rating.
We then piloted these questionnaires on friends and family and made sure they were ready to go out to the general public. Participants could then contact the researchers and would be sent a link to one of the questionnaires.
Throughout these different tasks I was given the freedom and responsibility to independently problem solve and contribute to making important decisions. This has not only increased my knowledge around research and data manipulation, but also my confidence in myself to undertake future research.
Findings and the next steps for the project
Within the timespan of this internship I was not able to collect any data and therefore see the final stage. However, I was able to see the data collected from some of the previous work done as part of the larger PhD study, and even carry out some SPSS calculations, allowing me to complete the full research cycle.
Whilst I may have finished my internship and contribution to the study, Katie and Alex will still go on to collect the data and using the mean typicality ratings to determine if there is a link between typicality and Autism. I will be paying close attention to find out!
I thoroughly enjoyed my experience on this LuCiD internship, both working on this project and engaging in other activities. Having the chances to attend a lab meeting and meet other researchers, being shown exciting new technology (such as eye-trackers), supporting recruitment for the LuCiD participant database and even presenting my study informally to other researchers has been really engaging and rewarding. I’ve thoroughly enjoyed my internship and I’m interested to see how I can involve research in my future endeavours. I would like to thank both of my supervisors for their advice, patience and humour and I would recommend the experience to anyone interested in research!