Abstract
Is visual memory dependent on attention? We tested this basic question in a partial-report change detection task that gauges three types of visual short-term memory (VSTM): iconic memory, fragile VSTM (fVSTM) and working memory. Participants had to judge whether a memory display containing eight oriented rectangles was identical to a test display that was presented after a blank interval or whether one of the rectangles had changed its orientation between successive views. In both the iconic memory and fVSTM condition, a retro-cue appeared during the blank, while in the working memory condition a post-cue appeared together with the test display. In all cases, the cue validly indicated the location where the potential change would happen or had happened. In the iconic memory condition, the cue appeared just after offset of the memory display. In the fVSTM condition and the working memory condition, the cue appeared 1 second after offset the memory display, yet respectively before or after presentation of the test display. Crucially, each trial was preceded by a predictive cue (pre-cue) with 75% validity. Thus, if subjects would selectively encode the single pre-cued item, they would perform correctly on the majority of trials. Indeed, when the pre-cue was valid, performance in all memory conditions was nearly perfect. However, upon invalid pre-cuing, performance in the working memory (post-cue) condition collapsed nearly to chance levels, while performance was only slightly impaired in the iconic memory and fVSTM condition. Altogether, these results suggest that focal attention is crucial for visual working memory storage, while it is less important for iconic memory and fVSTM. In line with previous behavioral (Vandenbroucke et al., 2011) and magnetic stimulation studies (Sligte et al., 2010), this study again implies that fVSTM is a memory storage that can be dissociated from visual working memory.
Meeting abstract presented at VSS 2013