Dataset for paper entitled: "The Congruency Conundrum: The Mechanisms of Facilitation in the Stroop task"
Description
Congruent trials represent a conundrum for selective attention mechanisms: relying on the irrelevant dimension helps responding quickly and accurately but also contravenes the overall task goal of ignoring the irrelevant dimension. Here we investigated the role of learning in producing this relatively understudied phenomenon. In standard Stroop task designs, incongruent and congruent trials are often presented in equal proportions, leading to congruent word-colour pairings occurring more frequently than any individual incongruent pairing (i.e., the word red is presented in red more times than the word blue). It has been argued that these word-colour correlations lead to Stroop effects. By comparing the Standard design to alternative designs in which correlations are absent, our results show Stroop facilitation remains in the absence of correlation (Experiments 1 and 2), and that facilitation is absent in the presence of (negative) correlation (Experiments 4). In contrast, facilitation was eliminated when the incongruent-to-congruent trial ratio was at least 3:1 (Experiments 3 and 4) and when positive correlation was absent. Furthermore, in Experiments 3 and 4 the proportion congruency effect arose from changes to facilitation, not interference, posing a challenge to accounts of the effect based on conflict monitoring or a more general response contingency. It is argued that learning associations between relevant and irrelevant dimensions in selective attention tasks occurs only when an attentional limit is reached (i.e., when there are at least three non-congruent response options) and is only important for congruent combinations; representing a response set size effect that is unique to facilitation effects.