This follow-up experiment will explore the effects of providing subjects with an explicit strategy on change detection tasks with familiar items. In the first follow-up study, data in the condition in which participants were told that they could identify the changed item by recognizing it as familiar appeared to be bimodal: Some participants performed almost identically to the change detection task with unfamilar objects, but a few participants outperformed even their 6-AFC performance. That pattern suggested that some participants defaulted to just treating the task as a change detection task, ignoring the usefulness of long-term memory. The performance of the other participants, though, suggested that the availability of both long-term memory and change detection performance allowed them to outperform either the 6-AFC or the change detection task alone. Their data were consistent with a strategy of first trying to detect a change. If they found it, they indicated it. If they didn't, they scanned the display for familiar objects and picked the familiar one. Their overalll performance in this task is comparable to what we would expect if we combined their performace on the other two tasks in this contingent way. This follow-up study will test whether we can encourage the use of both change detection information and long-term memory, effectively shifting more of the participants into this higher-performing group. The experimental procedure and data analysis will be similar to that of the main experiment and the first follow-up. Subjects will study a series of objects and then complete three tasks: change detection with unfamiliar objects, with instructions simply to find the change; change detection with familiar objects, with instructions to attempt to find the change, and if they cannot tell what changed, to pick the familiar item; and a 6AFC memory test. Results from the first follow-up study indicate that subjects do not make use of long-term memory information in change detection tasks, even when explicitly told that whenever a familiar item was present in the second display it would be the changed object. If an explicit, step-by-step strategy allows them to make use of both change detection performance and long-term memory, we expect accuracy rates higher than in either the change detection task or the memory task. If the strategy goes unused, however, we expect accuracy rates similar to those of the no-strategy change detection task. The data analysis procedure will examine the following comparisons: * Unstudied change detection accuracy vs. explicit-strategy change detection accuracy * Explicit-strategy change detection accuracy vs. 6AFC accuracy