Target–distractor correlation does not imply causation of the Stroop effect

In the Stroop task, the identities of the targets (e.g., colours) and distractors (e.g., words) used are often correlated. For example, in a list in which 4 words and 4 colours are combined to form 16 stimuli, each of the 4 congruent stimuli is typically repeated 3 times as often as each of the 12 incongruent stimuli. Some accounts of the Stroop effect suggest that in this type of list, often considered as a baseline because of the matching proportion of congruent and incongruent stimuli (50%), the word dimension actually receives more attention than it does in an uncorrelated list in which words and colours are randomly paired. This increased attention would be an important determinant of the Stroop effect in correlated situations, an idea supported by the observation that higher target-distractor correlation lists are associated with larger Stroop effects. However, because target-distractor correlation tends to be confounded with congruency proportion in common designs, the latter may be the crucial factor, consistent with accounts that propose that attention is adapted to the list's congruency proportion. In four experiments, we examined the idea that target-distractor correlation plays a major role in colour-word Stroop experiments by contrasting an uncorrelated list with a correlated list matched on relevant variables (e.g., congruency proportion). Both null hypothesis significance testing and Bayesian analyses suggested equivalent Stroop effects in the two lists, challenging accounts based on the idea that target-distractor correlations affect how attention is allocated in the colour-word Stroop task.