Stimulus duration and partial report performance

This paper investigates the effect of stimulus duration on partial report performance. Two experiments showed that variations in stimulus duration from 50 to 500 msec had little or no effect on partial report accuracy, decay, or error patterns. These results contrast with previous research demonstrating that increasing stimulus duration decreases visible persistence duration. Our findings thus provide support for the argument (Coltheart, 1980) that informational persistence (knowledge about the visual properties of a stimulus) is a separate phenomenon from visible persistence (the phenomenal persistence of a stimulus), and suggest that the traditional view of iconic memory, which conflates these two forms of persistence, is incorrect. Furthermore, our results indicate that informational persistence is not merely a function of unnaturally brief stimulus exposures, but exists even after exposure durations as long as 500 msec. Possible mechanisms of informational persistence, as well as its potential role in everyday perception, are discussed.