Abstract
Estimates of visual working memory capacity vary considerably across individuals. These estimates appear to reflect a stable cognitive trait of the observer because they are consistent over time and are strongly predictive of performance related to high-level cognitive processes such as abstract reasoning and fluid intelligence. One common measure of capacity is the change detection procedure (e.g., Luck & Vogel, 1997) in which observers are shown an array of simple objects that must be remembered over a retention period. Memory for the array is then measured with the presentation of a “test array” in which the observer must detect a change in the identity of one of the objects from the original array. In the present study, we examined the psychometric properties of two variations of the test array presentation: a whole-probe test in which all of the items from the memory array are present; and a single-probe test in which only a single item from the memory array is evaluated. We measured memory capacity in these two conditions across two separate sessions separated by a one-week interval. At the group level, we found that both whole-probe and single-probe conditions yielded comparable memory capacity estimates as well as strong test-retest correlations across the two sessions. However, at the individual level, we observed that low memory capacity subjects showed a large benefit in performance for single-probe conditions relative to whole-probe conditions, while the high capacity subjects showed a slight cost in performance for single-probe test arrays. Together, these results are consistent with previous work that has suggested that low memory capacity individuals are highly susceptible to attentional distraction and interference.