Abstract
Most modern devices include a user interface that employ symbolic icons to represent information to the user. Previous studies relating visual features of icons to response time (RT) differences in visual search have produced ambiguous findings (Arend, Muthig, & Wandmacher, 1987; Byrne, 1993). In the current study, our goal was to quantify concreteness and visual complexity of icons in a search task. We distinguished between concrete and abstract icons by creating our own set of icons meant to correspond to common applications encountered on an Apple iPad. Concrete icons were operationalized as an image of a real object (e.g., a camera); abstract icons were simple graphical depictions of the concrete icons. To validate our item categories, 52 participants rated each icon on several characteristics, including concreteness and complexity. In a separate study, 20 new participants performed a visual search-and-match task for a target icon among four different target/distractor arrays (concrete/concrete; abstract / abstract; concrete /abstract; abstract/concrete) on a background similar to an iPad screen. Eye movement data were also collected. Subjective ratings confirmed that observers considered concrete icons (x = 4.15) more concrete than abstract icons (x = 2.49). Accuracy in the search task was high (~ 98%) and did not differ significantly across conditions. Contrary to previous findings, we found that search RTs did not differ between concrete (2028 ms) and abstract icons (1969 ms). However, participants were significantly faster to locate the icon when the distractor category was different from the target category (215 ms and 173 ms, respectively). Interestingly, although RT in the concrete and abstract conditions did not differ, when factoring in complexity ratings we found that for abstract icons higher complexity led to longer RTs (~215 ms), suggesting that when icons are more ambiguous visual complexity plays a larger role in driving performance.