Prevalence effects on visual search and haptic search

Abstract

In visual search tasks, relative proportions of target-present and target-absent trials have important effects on miss rates and search termination time (target-absent reaction times). Miss rates increase as target prevalence decreases (Wolfe, Horowitz, & Kenner, 2005). This phenomenon is called prevalence effect. Wolfe and Van Wert propose that target prevalence affects the observers’ criteria and the quitting threshold (Wolfe & Van Wert, 2010). The observers’ criteria govern decisions about individual items during search, while the quitting threshold is related to search termination time. Wolfe and Van Wert suggest that both of these will influence miss rates. Although prevalence effect is observed in vision, there are no studies in other sensory modalities. In this study, we tested whether or not there was the prevalence effect on haptic search and examined the cause of prevalence effect. To compare the prevalence effect on haptic search with visual search, we conducted haptic search tasks and visual search tasks at two levels of prevalence (10% and 50%). We used a tactile map as haptic search tasks and a simulated baggage-screening task as visual search tasks. In our results, prevalence had an effect on miss rates in both search tasks. In haptic search tasks, increase in miss rates in the 10% condition was strongly correlated with a decrement in search termination time. On the other hand, in visual search tasks, we did not find the tradeoff between miss rates and search termination time, as has been reported elsewhere (e.g., Ishibashi, Kita, & Wolfe, in press). Apparently, in these tasks the prevalence effect on haptic search was mainly caused by change in the quitting threshold, while the prevalence effect in the visual search task was mainly caused by a shift in observers’ criteria to a strongly conservative position at low prevalence. Meeting abstract presented at VSS 2012