Ten years Diffusion Model for Conflict (DMC) tasks: Theoretical foundations, applications, practical recommendations,...

A central question in cognitive psychology concerns how humans selectively attend to task-relevant information while ignoring task-irrelevant information. This question is frequently studied using conflict tasks, such as the Simon, Eriksen flanker, and Stroop tasks that require responses to relevant stimulus features while ignoring irrelevant and potentially conflicting features. All conflict tasks produce better performance in congruent trials where task-relevant features and irrelevant features match compared with when they mismatch in incongruent trials: the congruency effect. However, they differ markedly in the temporal dynamics of this congruency effect and accordingly in their RT distributions. The Diffusion Model for Conflict tasks (DMC) was introduced as a model to formally account for all distributional patterns in conflict tasks. It integrates dual-route theories of cognitive control into the framework of sequential sampling models by positing two superimposed, independent evidence accumulation processes: a linear one for controlled processing of the task-relevant information, and a pulse-like one for automatic processing of the task-irrelevant information, in which activation first increases until a maximum and then decreases again. This review summarizes DMC's architecture, its core parameters, and its ability to account for various distributional patterns. We review and discuss applications of DMC across several psychological domains, and technical considerations such as parameter estimation and recovery. Limitations of the model are critically assessed, and fields of open research are outlined. Overall, DMC offers a general account of conflict processing. While a powerful tool for quantifying the dynamics of selective attention and cognitive control, there is still a limited standardization in its application, and more research is needed to extend DMC to other classes of conflict tasks.