Differential acquisition rates for different types of information from pictures

In three picture-recognition experiments, we investigated the characteristics of “specific feature” and “holistic” information hypothesized to be extracted from pictures. In each experiment, exposure time was varied at study. The recognition test was two-alternative forced-choice in which the target/distractor relationship was manipulated in such a way that responding on the basis of one type of information or the other was required. The results indicated that following one fixation on a picture at the time of initial study, performance based on holistic information was superior to performance based on specific feature information, whereas the reverse was true following sufficient study time for multiple fixations. These results support a view of pictorial information acquisition which postulates that most holistic information extracted from a picture is extracted during the first eye fixation on the picture, whereas subsequent fixations have the primary purpose of seeking out specific, informative features.