Meta-analytic evidence for the complex mechanisms underlying congruency sequence effect

The congruency sequence effect (CSE) refers to a reduction in the congruency effect after incongruent trials compared to congruent trials in a conflict-inducing task. There is an ongoing debate about the mechanisms underlying the CSE.

To help inform this debate, we conducted a meta-analysis of the relevant CSE studies published in the past 31 years (from 1992 to 2023). By measuring the mean effect sizes from various tasks and procedures, we examined to what extent the CSE results from top-down or bottom-up mechanisms and to what extent these mechanisms are local to the tasks performed or global to the experiment.

Results demonstrate that while the CSE was larger for studies that included bottom-up confounds (Hedges' g = 1.28), it was still robust and significant for studies that controlled for these confounds (Hedges' g = 0.95). Additionally, CSE was significant both within (Hedges' g = 1.54) and between tasks (Hedge's g = 0.27), but the effect was larger within a task. This suggests that the mechanisms driving the CSE affect both local and global control mechanisms. Furthermore, the current meta-analysis suggests that the congruency effect and CSE may not result from the same control mechanisms. Lastly, given that bottom-up confounds are easily controlled for in the prime probe and temporal flanker tasks, which produced a large CSE (Hedges' g = 1.13), these may be useful procedures to use to address future questions for CSE.

Overall, the present meta-analysis provides converging evidence for conclusions from previous studies of the CSE and highlights the complex factors that produce this effect.