Susceptibility to the Müller-Lyer Illusion in neurotypical individuals expressing autistic traits


The Müller-Lyer Illusion (ML) biases perceptual size estimation such that perceivers over- or underestimate the length of the central shaft when the arrowheads point inwards or outwards. Some literature suggests reduced susceptibility to the ML in autistic populations; however, it is unclear whether autistic trait expression corresponds with visual illusion susceptibility. Thus, task and/or sample differences may be responsible for equivocal findings. The present study investigated size estimation bias on four perceptual tasks in neurotypicals expressing varying levels of autistic traits. All participants (n=30) completed two questionnaires: the Autism-Quotient (AQ) and Systemizing-Quotient (SQ). Two forced-choice tasks presented two figures concurrently (Task1) or successively (Task2). Participants indicated which figure in the composite-pair contained the longer shaft. Participants also adjusted vertical lines (Task3) or ML-figures (Task4) to match shaft lengths of target ML-figures. Participants' AQ and SQ scores were significantly positively correlated (Pearson's r=.370, p<.05), but not with performance measures. Biased figure selections significantly increased when figures were presented concurrently, relative to successively (d=0.491, p<.01). Interaction effects showed adjustments were biased in the expected directions, with figure adjustments significantly more biased than vertical line adjustments (?p2=.621, p<.001). Figure adjustments were also significantly more inconsistent than vertical line adjustments (?p2 =.720, p<.001). The weaker bias when figures were presented successively, and when adjusting a vertical line, suggest viewing ML composite-pairs together strengthens the illusory effect. Thus, task details more strongly affected illusory bias than did autistic trait expression, with neither AQ nor SQ predicting the ML size estimation bias in the current sample.

Acknowledgments: Natural Sciences and Engineering Research Council