Task Uncertainty Can Account for Mixing and Switch Costs in Task-Switching

Cooper, Patrick S.; Garrett, Paul Michael; Rennie, Jaime L.; Karayanidis, Frini · 2015 · OpenAlex-citations

DOI: 10.1371/journal.pone.0131556

archive: archived pipeline: cataloged verified

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

Summary

This study investigates whether cognitive control during task-switching can be conceptualized as an information-processing mechanism designed to resolve environmental uncertainty. Building on the hypothesis that cognitive control prioritizes goal-relevant information to manage entropy, the authors examine if variations in task uncertainty account for typical behavioral effects, specifically mixing and switch costs. The research posits that higher information entropy—quantified via information theory—increases computational load, leading to slower reaction times and reduced accuracy. The researchers conducted two experiments using a cued task-switching paradigm involving three tasks (letter, digit, color). Experiment 1 manipulated proactive control by varying the informativeness of cues preceding the target, creating conditions with full, partial, or no advance information about the upcoming task. Experiment 2 introduced reactive control demands by adding a bivalent distractor at target onset, increasing interference. Participants performed mixed-task and single-task blocks. The study quantified information entropy for each trial type using algorithms that distinguished between contextual control (information from cues) and episodic control (information from previous trials). Additionally, the authors applied this entropy model to data from previous literature to test its generality. Results demonstrated that increasing uncertainty was consistently associated with less efficient performance. In both experiments, reaction times were slower and accuracy lower on trials with higher entropy. Specifically, fully informative cues (low entropy) yielded faster responses than partially informative or non-informative cues (high entropy). Switch trials, which inherently carry higher uncertainty than repeat trials, showed significant costs. The analysis revealed that both mixing costs (the cost of performing in a mixed block versus a single block) and switch costs were associated with a common episodic control process. The entropy model successfully predicted behavioral performance across the experimental conditions and when applied to prior studies, showing that reaction times increased linearly with calculated information entropy. These findings support the notion that cognitive control functions as an information processor that resolves uncertainty. The study suggests that mixing and switch costs are not distinct phenomena but rather reflect the computational load required to manage uncertainty through episodic control processes. By linking behavioral costs directly to information entropy, the research provides a unified framework for understanding how cognitive control adapts to varying levels of uncertainty in both proactive and reactive contexts. This implies that the efficiency of task-switching is fundamentally determined by the amount of information processing required to prioritize relevant goals amidst conflicting or ambiguous signals.

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.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-26; verification: verified.

Topics

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

• task switching
• executive function