Unpacking the task details behind when autistic trait expression varies Müller-Lyer susceptibility


The Müller-Lyer Illusion (ML) is a well-known visuospatial illusion that biases perceptual size estimation. There have been mixed reports of reduced susceptibility to the ML in both autistic and neurotypical participants expressing autistic traits. The present study investigated illusory bias on three perceptual tasks in neurotypical participants (n=11). Participants also completed the Autism-Quotient (AQ) and Systemizing-Quotient (SQ). Participants indicated by keypress which of two figures contained the longer central shaft when the figures were presented either concurrently (Task1), successively with a discrete response slide (Task2), or successively with responses made on the second figure's presentation (Task3). Questionnaire scores revealed task specific correlations. Task1 RT was negatively correlated with AQ (r(9)= -0.606, p=.048), while Task2 RT was negatively correlated with SQ (r(9)=-0.642, p=.033). Task3 biased figure selection was positively correlated with AQ (r(9)=-0.674, p=.023). Participants showed expected biases in figure selection for all three tasks. Given the lack of difference in illusory bias found, the negative RT:AQ correlation in Task1 could reflect a speed-accuracy trade-off according to autistic traits. The negative RT:SQ correlation during Task2 is consistent with a benefit of systemizing ability when a memory-based comparison of both figures was required. The positive correlation of biased figure selection with AQ during successive presentation (no response slide) is contrary to the predicted negative correlation. Differences in the task details may have led to the opposing effect of autistic trait expression. Future studies should replicate and extend this finding to determine why this specific task resulted in a positive correlation.

Acknowledgments: Natural Sciences and Engineering Research Council