Temporal Dynamics of Visual Short-Term Memories

Using visual displays consisting of single items to be stored in visual short-term memory (VSTM), Jacob, Breitmeyer & Treviño (2013) demonstrated three stages of information processing in VSTM: iconic visible persistence, iconic informational persistence, and visual working memory. To investigate the effect of higher memory load on these proposed VSTM stages, several measures of VSTM scanning and visual search efficiency including VSTM capacity, VSTM scanning and visual search slopes, and stimulus comparison effects, were obtained in the first part of the study, using 1, 3, and 5 display items. Results again revealed three stages of VSTM processing, but with a second phase growing longer as memory load increased, suggesting a need for a longer VSTM encoding process in this phase, and VSTM maintenance in the third stage. The second part of the study investigated whether these three stages of information processing are due to task-dependent strategic effects. Separate groups of observers were run in stimulus priming and comparison tasks using (1) a short range of SOAs exploring the first two seconds and (2) a longer range of SOAs exploring the first four seconds of post-stimulus processing. Half the trials were blocked by SOA, and in the other half SOAs were tested in a randomly intermixed order. Priming and comparison effects in the short-range experiment demonstrated, as before, three stages of post-stimulus processing in VSTM, and comparison effects in the long-range experiment showed evidence for an additional fourth stage of VSTM processing. Furthermore, in the short-range experiment, higher priming effects occurred when trials were blocked by SOA. In contrast, the priming effect across SOAs in the long-range experiment and the comparison effect in both SOA ranges were not influenced by SOA-blocked or -intermixed trials. This is indicative that the initial three stages as well as the later stages of VSTM are not an artifact of the manner in which observers are tested. These findings and their implications are related to other paradigms and methods used to investigate post-stimulus processing.