The performance costs of interruption during visual search are determined by the type of search task

Alonso, David; Lavelle, Mark; Drew, Trafton · 2021 · OpenAlex-citations

DOI: 10.1186/s41235-021-00322-0

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 how the cognitive demands of a primary task influence the performance costs associated with interruptions, specifically within the context of visual search. Prior research has established that interruptions disrupt performance, but the nature of these disruptions—whether they cause errors or merely increase completion time—varies significantly and is poorly understood. The authors propose that working memory (WM) plays a critical role in this variability, particularly in maintaining the "goal" or target template of a task. This research is motivated by the need to understand how interruptions affect high-stakes professions like radiology, where clinicians must search medical images for specific pathologies. The study tests the hypothesis that interruptions will incur higher accuracy costs when the search target changes frequently (requiring WM maintenance) compared to when the target remains consistent (allowing reliance on long-term memory). The researchers conducted two online experiments using laboratory-based visual search tasks. Participants searched through arrays of 150 objects for a specific target. In Experiment 1, the target changed on every trial, imposing high WM demands to update the target template. In Experiment 2, the target remained the same across all trials, reducing WM load. On 10% of trials, participants were interrupted by a secondary task requiring them to type a random string of characters before resuming the search. The study included 62 participants total (31 per experiment), with data collection managed via Pavlovia.org and Prolific.co. The experimental design ensured that interruptions occurred after participants had begun searching but before completion, mimicking real-world disruptions. The results demonstrated that the type of search task significantly modulated the impact of interruptions on accuracy. In Experiment 1, where targets changed frequently, interruptions led to a significant decrease in search accuracy compared to uninterrupted trials. Conversely, in Experiment 2, where the target was consistent, interruptions did not significantly affect accuracy. However, interruptions increased response times in both experiments, indicating a universal time cost regardless of WM demands. The interaction between trial type (interrupted vs. uninterrupted) and search type (changing vs. consistent target) was significant for accuracy but not for the magnitude of the time delay. These findings confirm that while interruptions always slow down performance, they only degrade accuracy when the primary task requires active maintenance of goal information in working memory. The significance of these findings lies in their implication for designing work environments and understanding cognitive limits. The study suggests that the adverse effects of interruption are not uniform but depend on the specific cognitive processes engaged by the primary task. For fields like diagnostic radiology, this implies that interruptions are more likely to cause diagnostic errors when radiologists are searching for varied or novel signs of disease rather than consistent patterns. By identifying that WM demands determine whether an interruption leads to errors or just delays, the research provides a framework for predicting and mitigating interruption costs in applied settings. It highlights the importance of considering task design and memory load when evaluating the risks of interruptions in complex professional environments.

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 2026-06-25; verification: verified.

Topics

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

• prospective memory interruptions