The Telephone Game Research
Officially known as ‘Efficiency and Interaction during Information Transfer between Autistic and Neurotypical People’
On to the second piece of work, the telephone game. This research turned out to be not a paper, but a very new poster presentation at the 2019 International Society for Autism Research (INSAR) conference. This means the dataset they’re studying was collected recently – November 2018, in fact, barely a year old at the time I wrote this article. But it also means this dataset hasn’t been fully analyzed, written up and published yet. Overall it’s good news: we have a paper to look out for, and they’re sharing an interesting portion of their findings so far!
Anyway, according to the INSAR website’s record of this presentation, the authors, Catherine Crompton and Sue Fletcher-Watson, drew on ideas from the double empathy problem. They wanted to test if social cognition is truly impaired in autistic individuals compared to neurotypical individuals. In order to examine this, they set up three groups – one group of all autistic individuals, one group of all neurotypical individuals, and a mixed group of autistic and neurotypical individuals.
An illustrated example of how the groups worked: https://twitter.com/cjcrompton/status/1161292630795403264?s=20
These groups then played telephone (or as research calls it, conducted a ‘diffusion chain’) – a researcher gave one person in each group a story to repeat to one other person from their group, who would then repeat the story to the next person, and so on. The autistic-neurotypical group received and retold the story every other neurotype (e.g., one autistic person, one neurotypical person, etc.). From what I can tell, they had 2-3 trials per group type, so the data was collected on two or three ‘diffusion chains’ (or telephone game rounds) per group. They had 8 participants per chain, and they recorded the conversations for examination later.
A cute tweet showing what a diffusion chain is: https://twitter.com/cjcrompton/status/1161293699613188096?s=20
The stories given to participants by researchers had been pre-coded with 30 specific details that they would track as the stories were passed along. The results are preliminary, and eventually need to be further validated – but I found them exciting. Basically what happened: the entirely autistic and entirely neurotypical groups lost details at about the same rate as the story passed from participant to participant. The mixed autistic and neurotypical group lost details at a faster rate, especially at the beginning, if we look at the chart below. It is worth noting that all group averages appear to end up around 5-6 details recalled by the 8th and final retelling. It does not appear that any statistical ‘significant differences’ were reported on the INSAR’s summary. This more or less means I’m not sure if we can say this result is most likely not due to chance and not is so close as to be considered statistically the same.