The Influence of Clutter on Target Prevalence and Decision Making During Visual Search

Abstract

Participants are sensitive to target prevalence effects in visual search. Low prevalence of targets leads to increased miss rates and shorter response times, and high prevalence of targets leads to increased false alarm rates and longer response times. These effects have been explained using the Multiple-Decision Model (MDM), in which two decisions impact performance during serial visual search. The first decision is whether an inspected item is a target. The second decision is whether the search should be ended with a target-absent response. Target prevalence influences these decisions, evidenced by changes in miss rate, false alarm rate, and response time. These decisions can likely be further affected by the number of target decisions that need to be made and the crowding present for each decision. Experiment 1 examined the effect of the number of target decisions to be made, by varying set size, on the target prevalence effect. Experiment 2 examined the effects of the number of decisions and crowding on the target prevalence effect by manipulating the amount of clutter in aeronautical search displays. Varying set size altered participants’ quitting thresholds, and varying clutter altered participants’ quitting thresholds and 2AFC decision-making starting points. In Experiment 1, the shift in participants’ quitting thresholds in response to set size interacted with the target prevalence effect to alter participants’ response times. In Experiment 2, the shift in quitting threshold and 2AFC decision-making starting point in response to clutter interacted with target prevalence to alter participants’ response times and miss rates. These results suggest that the target prevalence effect can vary depending on the types of attention used. Additionally, this suggests that we can reasonably predict how participants will interact with a visual search display using the MDM model.