Modeling response times in the size-congruity effect: Early versus late interaction
The size-congruity effect occurs when numerical magnitude interferes with judgments of physical size. Various accounts propose that this interference is either encoding-related or decision-related, though at present a clear consensus is lacking. In our study, we administered a single-digit physical comparison task (i.e., which digit is physically larger?) and applied four different mathematical models (ex-Gaussian, ex-Wald, shifted Wald and EZ-diffusion) to the observed response times. The aim of this modeling was to index the underlying cognitive processes via estimates of drift rate, response threshold, and non-decision time. The collection of estimates for each individual was then subjected to Bayesian paired samples t-tests. We found that the drift rate for incongruent trials was smaller than for congruent trials, indicating that congruent trials had a faster rate of information uptake. The response threshold for incongruent trials was generally larger than for congruent trials, indicating that for incongruent trials more information needed to be accumulated before a response could be initiated. Critically, we found evidence of an invariance in non-decision times between incongruent and congruent trials. This combination of results provides support for a late interaction account of the size-congruity effect, shedding further light onto models of decision making in number processing.