Coltheart, V.
Macquarie Centre for Cognitive Science Macquarie University
Brief unmasked (50-100 ms) visual displays of alphanumeric characters are retained in a short-lived form of visual memory referred to as iconic memory. Although only 3-4 items can be reported from such displays when full report is attempted, the use of partial report cues presented after display offset has shown that many more items are briefly available in memory and can be selected for report even when the cue is delayed by 50 ms or more (Sperling, 1960; Averbach & Coriell, 1961). However a different technique has been used more recently to measure memory from a brief visual display. The display is shown again with one or more missing items. The task is to report the missing items from the initial display. Using this technique, the results are not consistent with the traditional picture of iconic memory. Several experiments contrasted report from brief letter displays interrogated by missing letters and other cues. The implications of results obtained with different partial report techniques for an understanding of properties of iconic memory are considered.