Local probability effects of repeating irrelevant attributes.

Abstract

We present two experiments in which participants classify stimuli having two potentially conflicting attributes, one of which is response-relevant whereas the other ("irrelevant") attribute is logically and statistically independent of the response. We introduce a novel design not used with filtering tasks before in which the main factor is the local (i.e., one-step) transition probability π (= 0.25, 0.50, 0.75) that the irrelevant attribute is repeated from one trial to the next. Experiment1 involved a visual Simon task in which the color of the stimulus is relevant and its location is irrelevant. Experiment2 used a semantic classification task in which the parity of a digit presented is relevant and its numerical magnitude is irrelevant. The results of both experiments demonstrate that participants in the π = 0.75 group responded faster when the irrelevant attribute is in fact repeated rather than alternated; in contrast, participants in the π = 0.25 group responded faster (Experiment1) or equally fast (Experiment2) when the irrelevant attribute is alternated rather than repeated. These expectancy-related effects cannot be attributed to spurious design contingencies as the irrelevant attribute was independent of the relevant attribute (and thus of the response), of the congruency status, and also of their alternation/repetition. One interpretation of our findings is that information about the irrelevant attribute in the previous trial is used much as an informative central precue, so that participants can prepare early processing stages in the current trial, with the corresponding benefits and costs typical of standard cueing studies.