Activation thresholds, not quitting thresholds, account for the low prevalence effect in dynamic search

Becker, Mark W.; Rodriguez, Andrew; Bolkhovsky, Jeffrey; Peltier, Chad; Guillory, Sylvia B · 2024 · Crossref

DOI: 10.3758/s13414-024-02919-1

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 the low-prevalence effect (LPE), a phenomenon where target detection rates decline as targets become less frequent in visual search tasks. While the LPE is well-documented in static search tasks, it is traditionally attributed to lowered "quitting thresholds," where observers terminate trials early when targets are rare. However, this explanation is theoretically inapplicable to dynamic, continuous search tasks—such as monitoring sonar or surveillance video—where there are no discrete trials to quit. The authors sought to determine if the LPE persists in these dynamic contexts and whether the quitting threshold model remains a valid explanation. The research employed two experiments using an online platform. Experiment 1 utilized a 2 (prevalence: 10% vs. 40%) x 2 (presentation: static vs. dynamic) between-subjects design. Participants searched for specific targets (Ts or Ls) among distractors. In the static condition, participants made present/absent responses for discrete trials. In the dynamic condition, participants continuously monitored a stream of moving items for 24 minutes, responding only when a target appeared. Experiment 2 replicated this design but manipulated prevalence by having two different targets appear at different rates (10% and 40%) within the same display, further isolating the prevalence effect from trial-level quitting decisions. The results demonstrated that the LPE persisted in both static and dynamic search conditions with similar magnitudes. In Experiment 1, target detection accuracy was significantly lower for low-prevalence targets compared to moderate-prevalence targets, regardless of whether the display was static or dynamic. Overall performance was worse in the dynamic task, but the prevalence effect was additive rather than interactive. Crucially, in the static condition, low-prevalence targets elicited faster target-absent reaction times, confirming the traditional quitting threshold mechanism. However, because the dynamic task lacked target-absent responses and discrete trials, the persistence of the LPE in this condition could not be explained by quitting thresholds. Experiment 2 confirmed these findings, showing that even when prevalence was manipulated within a single continuous stream, detection rates dropped for the rarer target. These findings challenge the sufficiency of the quitting threshold model as a universal explanation for the LPE. Since the effect persists in dynamic searches where quitting thresholds are irrelevant, the authors propose that an "activation threshold" explanation is more appropriate. This suggests that rare targets fail to sufficiently activate the observer’s decision criterion, leading to missed detections even when the target is inspected. The study highlights that the LPE is a robust concern for both static and dynamic real-world search scenarios, such as medical imaging and security screening, necessitating theoretical models that account for decision-making criteria shifts rather than just search termination behaviors.

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.

• visual search