One of the best-known paradigms to study interference between cognitive processes is the Stroop task1. Over time, many variants of the task have evolved2, differing with respect to stimulus material (e.g. color-word, picture-word or number variants), control conditions (congruent or neutral), presentation design (mixed or blocked), as well as combinations with additional cognitive demands. The neural and behavioral impact of this variety, however, has never been systematically assessed. We performed a series of meta-analyses to synthesize behavioral and neuroimaging findings of studies implementing Stroop-type tasks and to investigate commonalities and differences between different versions. In total, the coordinate-based meta-analyses included 133 neuroimaging experiments, which reported 164 effect sizes for inclusion in the behavioral analyses. Results revealed little impact of task variations on the mean effect size of reaction time (g=0.64 across all 164 effect sizes, CI = 0.56-0.73). Neurally, incongruence processing in the classic color-word Stroop variant consistently recruited regions of the multiple-demand network, with some modulation of spatial convergence by stimulus material, control condition, design, and cognitive demand. In line with the view of a “many-to-one mapping”3, our results suggest that the seemingly unitary behavioral costs of Stroop-type conflicts may arise from partly different neural processing mechanisms, depending on contextual factors. The impact of different features of the task should therefore be carefully considered when planning or interpreting Stroop-type experiments, especially in clinical or other applied fields.

[1] Stroop, JR (1935). J.Exp.Psychol., 18:643-662.
[2] McLeod, CM (1991). Psychol. Bull., 109 (2):163-203.
[3] Westlin, C, et al. (2023). TiCS, 27(3):246-257.