On the separability of target and distractor processing in conflict tasks

Abstract

By orthogonally pairing relevant and irrelevant information, conflict tasks entail an inverse problem: Are congruency effects shaped by distractor strength, target strength, or both? This fundamental question has received mixed empirical answers, leaving it unclear whether congruency effects only increase with stronger distractor processing or also with weaker target processing. We hypothesize that the discrepant results were partly driven by perceptual target–distractor grouping: Selective target and distractor stimulus manipulations in flanker-like displays can exert non-selective influences on target- and distractor-based activation in a dual-route architecture of response selection, if such manipulations alter the perceptual target–distractor organization and thus the perceptual quality of the non-manipulated stimulus. Three large-scale experiments selectively manipulated target and distractor strength (Experiments 1–2: stimulus discriminability, Experiment 3: stimulus intensity) in separate blocks of flanker-like conflict tasks. The strength of the non-manipulated stimulus was equidistant to the weak and strong levels of the manipulated stimulus to minimize perceptual-grouping confounds. All three experiments consistently revealed that congruency effects not only increase with stronger distractor processing but also with weaker target processing. These grouping-confound-minimized data allowed a first experimental validation of the separation between sustained target processing (controlled component) and transient distractor processing (automatic component) in the diffusion model for conflict tasks. Model comparisons and significance tests on parameter estimates suggest that the controlled component is sensitive and mostly specific to target manipulations, while the automatic component is sensitive (but only in the parameter estimates) and largely specific to distractor manipulations.