Abstract
Evidence that people often fail to detect substantial visual changes to objects and scenes (i.e., change blindness) has been used to infer an overall lack of detailed internal visual representations of the visual world (e.g., O‘Regan & Noë, 2002). However, people can retain detailed visual information for extended periods of time and recall details of both pre- and post-change objects even when they fail to notice changes (e.g., Mitroff et al, 2004; Hollingworth, 2005). Such recognition in the face of change blindness implies either (a) that change blindness results from a failure to compare representations of the pre- and post-change objects or (b) that longer-term representations are inaccessible to shorter-term change detection processes. In this experiment, participants studied individual objects for a subsequent memory test and, after this study phase, performed a change detection task with arrays of objects that either did or did not include these familiar objects. Change detection was better for“old” than for “new” objects even though object familiarity was not predictive of the change or its location. The pattern held for both short and long initial viewing of the pre-change array, with somewhat better detection for the longer initial viewing. This finding suggests that long-term representations of objects can contribute to the short-term processes involved in change detection tasks (at least in a 1-shot task), and that change blindness in the face of persisting representations can better be attributed to a comparison failure.