Robust evidence for proactive conflict adaptation in the proportion-congruent paradigm

In the standard Proportion-Congruent (PC) paradigm, performance is compared between a list containing mostly congruent (MC) stimuli (e.g., the word RED in the color red in the Stroop task; Stroop, 1935) and a list containing mostly incongruent (MI) stimuli (e.g., the word BLUE in red). The PC effect, the finding that the congruency effect (i.e., the latency difference between incongruent and congruent stimuli) is typically larger in an MC list, has been interpreted by the popular conflict-monitoring account (Botvinick et al., 2001) as reflecting a proactive process whereby attention to task-relevant information is adapted based on how frequently conflict from task-irrelevant information arises. Recently, however, alternative accounts of the PC effect have emerged that assume either that the PC effect reflects processes other than proactive conflict adaptation (e.g., stimulus-response contingency learning) or that proactive conflict adaptation is only engaged as a last resort (e.g., when contingency learning cannot be used to minimize interference). We examined these ideas in three experiments in which proactive conflict adaptation could be evaluated independently from processes that are normally confounded with it in the PC paradigm, while still allowing those processes, particularly contingency learning, to be used to minimize interference. Consistent with the conflict-monitoring account of the PC effect, but inconsistent with all the alternative accounts of the PC effect, evidence for proactive conflict adaptation emerged in all experiments. Although multiple processes may be engaged in the PC paradigm, this paradigm remains a valid tool for examining proactive conflict adaptation, its typical use.