Processing stages in overlapping tasks: Evidence for a central bottleneck.

Pashler, Harold · 1984 · Journal of Experimental Psychology Human Perception & Performance

DOI: 10.1037//0096-1523.10.3.358

archive: archived pipeline: cataloged verified

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

Summary

This paper investigates the cognitive mechanisms underlying interference in the overlapping tasks paradigm, specifically testing whether a central bottleneck or capacity sharing explains the slowing of the second response (R2). The author contrasts bottleneck models, which posit that certain processing stages cannot occur simultaneously, with capacity-sharing models, which suggest that concurrent tasks compete for limited resources, leading to graded performance reductions. To distinguish between these theories, Pashler employs an additive-factors analysis, manipulating stimulus factors that affect specific processing stages—encoding versus decision/response selection—and observing how their effects change when a task is performed alone versus as the second of two overlapping tasks. The study reports five experiments, with detailed results provided for Experiment 1. In this experiment, eight participants performed a visual search task where they determined the presence or absence of a target letter in a four-letter array. This task was performed either in isolation or as the second task (Task 2) overlapping with a first task (Task 1) involving a simple discrimination of a light bar’s position. Two factors were manipulated in Task 2: stimulus contrast (high vs. low), which primarily affects the early encoding stage, and target presence/absence, which affects later decision or response selection stages. The design allowed for the comparison of factor effects in single-task versus dual-task conditions to identify which stages were postponed by the overlap. The results from Experiment 1 demonstrated significant R2 slowing, with the second response delayed by an average of 179 ms compared to single-task performance. Crucially, the effect of the contrast manipulation was reduced in the dual-task condition. While low contrast slowed responses by 58 ms in the single-task condition, it caused only a 28 ms delay in the dual-task condition, resulting in a significant interaction between task overlap and contrast. In contrast, the effect of target presence/absence remained largely unchanged, with a 31 ms difference in the single-task condition and a 27 ms difference in the dual-task condition, showing no significant interaction with task overlap. These findings indicate that the early encoding stage of the second task proceeds in parallel with the first task, while later stages are postponed. The findings support bottleneck models of attention, specifically suggesting that the central bottleneck occurs after stimulus encoding but before or during response selection. The reduction in the contrast effect under dual-task conditions aligns with the prediction that factors affecting stages prior to the bottleneck will have diminished effects when that bottleneck postpones subsequent processing. Conversely, the unchanged effect of target presence/absence suggests that decision and response selection stages are subject to the bottleneck. The absence of overadditive interactions argues against capacity-sharing models, which would predict that resource depletion affects all stages proportionally. These results imply that divided attention limitations are not due to general resource scarcity but rather to specific sequential constraints on central processing stages.

Key finding

In overlapping tasks, the slowing effect of factors affecting early automatic processing stages is reduced, while factors affecting later nonautomatic stages remain additive, providing evidence for a central bottleneck in response selection rather than capacity sharing.

Methodology

lab_experiment

Sample size: 8

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 openalex_abstract on 2026-05-08 (2 acquisition events logged).

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.

• dual task multitasking
• attention
• capacity resource theory