Abstract
In visual search, observers often miss rare targets (e.g., threats in luggage). This low prevalence effect (LPE) is resistant to many cognitive interventions. However, a recent study showed that having participants identify the item that was most similar to the target (i.e., similarity search) eliminated the LPE. As real-world searches often require binary decisions (e.g., is there a threat in this bag?), we tested whether the benefits of similarity search generalize to binary decisions and to more naturalistic stimuli. Participants searched for T shapes amongst T-like objects, and the prevalence of True-T’s was manipulated. Participants clicked on the “most T-like object in the array” and reported whether the chosen item was a True T (yes/no). Participants were given feedback only on their binary decision. We found an LPE for this search task (miss rates for True T’s were higher under low prevalence). Next, we removed the binary decision task and only provided feedback based on participants’ search performance. Here, we found that the LPE was greatly attenuated. Our results suggest that feedback based on the visual search task, and not similarity search per se, is what breaks the LPE. We tested this further in an additional experiment using naturalistic stimuli. Participants watched clips of road videos at different hazard prevalence rates (high/low) and clicked on the “most hazardous location” in the video, and then made a binary decision (“would you need to respond to that hazard? yes/no”) and were given feedback. While participants accurately located the hazards in the low prevalence condition, there was an LPE for the binary decision task. Together, these results indicate potential limitations in applying similarity search outside the laboratory; the LPE is still seen if participants adjust either their quitting threshold or response criterion depending on the feedback provided.