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

Lee, Yunji; Verhaeghen, Paul; Hazeltine, Eliot; Schumacher, Eric H. · 2025 · Psychological Research

DOI: 10.1007/s00426-025-02093-5

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This meta-analysis investigates the mechanisms underlying the congruency sequence effect (CSE), a phenomenon where the magnitude of the conflict effect is reduced following incongruent trials compared to congruent ones. The study addresses ongoing debates regarding whether the CSE arises from top-down attentional control or bottom-up associative processes, and whether these mechanisms are local to specific tasks or global across domains. By synthesizing data from 146 studies published between 1992 and 2023, comprising 311 independent experiments with healthy young adults, the authors aim to clarify the contributions of these competing accounts and identify experimental factors that modulate the effect size. The researchers employed strict inclusion criteria, focusing on standard conflict tasks (Stroop, flanker, Simon, prime-probe) with manual key responses and a 50/50 proportion of congruent and incongruent trials. Effect sizes were calculated by transforming reported F-statistics into Hedges’ *g*. The analysis compared studies that included bottom-up confounds (feature repetition and contingency learning) against those using confound-minimized designs. Additionally, the study examined within-task versus between-task CSEs in dual-task paradigms to assess domain specificity, and analyzed the correlation between the initial congruency effect and the CSE. Results indicated a large overall CSE (Hedges’ *g* = 1.11). While the effect was larger in studies with bottom-up confounds (*g* = 1.28), it remained robust and significant in confound-minimized studies (*g* = 0.95), supporting the role of top-down control mechanisms. The CSE was significantly larger within tasks (*g* = 1.54) than between tasks (*g* = 0.27), suggesting that control mechanisms are primarily local to the task set, though a smaller global effect persists. The analysis also revealed that the CSE decreases as the number of stimulus-response mappings increases, likely due to reduced bottom-up confounds in more complex designs. Crucially, the study found no significant correlation between the magnitude of the congruency effect and the CSE, implying they are driven by distinct control processes. The findings provide converging evidence that the CSE is a complex phenomenon influenced by both top-down and bottom-up mechanisms, with control being predominantly task-specific. The robustness of the CSE in confound-minimized designs validates top-down accounts like conflict monitoring theory. Furthermore, the lack of correlation between the congruency effect and CSE suggests that within-trial conflict and trial-to-trial adaptive control are dissociable. The authors recommend prime-probe and temporal flanker tasks as optimal procedures for future research, as they effectively control for bottom-up confounds while producing large, reliable CSEs.

Key finding

The congruency sequence effect is robust and significant even when bottom-up confounds are controlled, and it operates through mechanisms that are both local to specific tasks and global across domains, with a stronger effect observed within tasks.

Methodology

meta_analysis

Sample size: 311

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed. Discovered via author_sweep_intake on 2026-05-28.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-10; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• task switching