Australasian Mathematical Psychology Conference 2019

Modelling the speed and accuracy of continuous outcome colour decisions: Metric and categorical effects

Philip Smith
Melbourne School of Psychological Sciences, The University of Melbourne
Saam Saber
The University of Melbourne
Elaine Corbett
Trinity College, Dublin
Simon Lilburn
The University of Melbourne

We used the circular diffusion model to characterise the speed and accuracy of continuous outcome decisions about the hues of noisy colour patches in a small-N psychophysical design. Four subjects made eye-movement decisions about the hues of colour patches in an isoluminant, equi-discriminability colour space (CIELUV space) at three levels of stimulus noise. Heavy-tailed distributions of decision outcomes were found at high levels of noise, similar to those found in visual working memory studies at high levels of memory load. Decision times were longer for less accurate decisions, in agreement with the slow error property typically found in difficult two-choice tasks. Decision times were shorter and responses were more accurate in parts of the space corresponding to nameable colour categories, although the number of categories and their locations were idiosyncratic. We present data comparing the ability of three models of across-trial variability to account for the joint distributions of decision times and decision outcomes: Gaussian noise with different radial and tangential variances; probabilistic encoding failures; and nonlinear transformation of encoded stimulus identities. We show that the categorical effects can be modelled as stimulus bias effects, represented as vector sums of drift rate and category bias vectors.