@misc{cogprints2189, volume = {118}, number = {2}, title = {Partial report: Iconic store or two buffers?}, author = {Siu L. Chow}, year = {1991}, pages = {147--169}, journal = {The Journal of General Psychology}, keywords = {Iconic store, partial report}, url = {http://cogprints.org/2189/}, abstract = {ABSTRACT. Selective recall of a subset of letters from a multiletter array declines systematically with increases in the delay of the partial report cue. The issues addressed were (a) whether such a decline is due to progressive loss of location information or to systematic loss of features and (b) whether partial report is the result of a select-then-identify or an identify-then-select process. Instructing the subjects to guess or not to guess had an effect of array, displacement, and extra-array errors. Emphasizing on recall location affected both intra- and extra-array errors. The interstimulus interval manipulation had an effect on extra-array errors as well as on intra-array errors. These observations are contrary to the suggestions that intra-array errors are due to loss of location information and that extra-array errors are indicative of a joint effect on misidentification due to chance and the ratio of extra-array errors to intra-array errors. Some other results are difficult for a dual-buffer model but can readily be accounted for by the orthodox view of the iconic store. THE INITIAL STAGE OF VISUAL PERCEPTION is generally characterized as a transient veridical representation. Some investigators have suggested that only sensory, noncategorical information such as location, shape, color, and size are available in the transient representation (Haber, 1969; Neisser, 1967; Sperling, 1960). Selective recall is achieved by choosing some parts of the veridical representation for further processing on the basis of sensory information (e.g., Sperling, 1960; Turvey \& Kravetz, 1970; von Wright, 1968). This position is called the "select-then-identify" view. } }