Individual Differences Predict Low Prevalence Visual Search Performance And Sources Of Errors: An Eye-Tracking Study

Targets in real-world visual search tasks, such as baggage screening, may appear on as few as 2% of searches (Hofer & Schwaninger, 2005). Rare targets are missed more frequently than common targets, a phenomenon known as the low prevalence effect. Given the importance of rare target detection, researchers have sought to increase performance through technological improvements, experimental manipulations, and individual differences approaches. Here we focus on the individual differences approach, which has shown that it is possible to predict an individual's low prevalence search accuracy in a T among Ls search using basic cognitive tasks. Here, we address limitations of previous work by using both basic Ts and Ls and more representative baggage screening items. Results show we can account for 53% of variance in low prevalence search accuracy. Eye-tracking results show that fluid intelligence and near transfer search performance predict selection errors (misses caused by never inspecting the target) while working memory capacity and near transfer search performance predict identification errors (misses caused by misidentifying an inspected target). We conclude that the individual differences approach can be an effective tool to select who will perform well in real-world searches.