Australasian Mathematical Psychology Conference 2019

Updating judgement contexts with extreme stimuli

Simon Farrell
School of Psychological Science, University of Western Australia
Greta Fastrich
University of Reading

A number of theories assume that objects are not judged in isolation, but are compared to other objects. In many experiments the context is the objects seen in the experiment; for example, if judging the size of squares, the judgement context would be the set of all (or some of) the squares seen so far in the experiment. We ask what happens when an extreme stimulus is only occasionally presented. Does it enter into the judgement context, or is it effectively discounted? Across two experiments involving magnitude judgements on squares and numbers, we find little effect of the outlier on following judgements. Nonetheless, we show that people used the experiment context to form their judgements, by showing sensitivity to the skew of the distributions. Fitting two models of context-based judgement—Parducci’s range-frequency theory, and Haubensak’s consistency model—suggests the combined effects of overall context and individual items is challenging.