Abstract:  Working memory (WM) is the memory system to store and process information shortly for the goal-directed behavior. The resource of WM is extremely limited. It is necessary to selectively maintain relevant information and inhibit interference from distracting information. Previous studies have found that individuals with high WM capacity would be able to suppress interference more efficiently. However, it’s still unknown whether the precision of WM representation will be influenced by distractors in a similar way as WM capacity. Perceptual representation as the first stage of information processing, its precision will highly influence the quality of information stored in WM. Here, we adopted an independent perceptual task and a variant version of color recall task to examine (1) the effect of interference on WM capacity as well as precision and (2) whether the precision of perceptual representation would influence the ability to inhibit interference during WM processing. Fifty-three undergraduate students participated in the experiment. In the first perceptual task, participants reported the color of a cued item while the item remained presented until a response was made. Afterwards, color recall tasks were operated including a ‘distractor-present’ and a ‘distractor-absent’ condition. Each condition contained three levels of WM load: low (2 targets), medium (4 targets) and high (6 targets). In ‘distractor-present’ condition, target items were presented along with two more distracting colored stimuli with a different shape. Participants were asked to remember only the color of the target items and report the remembered color of a cued item by clicking on a color wheel. From the distribution of errors between the reported color value and the original color value, we could obtain measures for both capacity and precision via standard mixture model used by Zhang and Luck, 2008. To explore the effects of distractor on WM capacity and precision and whether these influence will vary with WM load, repeated measures analysis of variance (ANOVA) was performed on WM capacity and precision. For WM capacity, main effects of distractors, WM load and their interaction were significant. WM capacity decreased with increasing WM load. And WM capacity in ‘distractor-absent’ condition was significantly higher than performance in ‘distractor-present’ condition. Interaction of distractor and WM load mainly reflected that distractors decreased WM capacity under the medium and high loads. For WM precision, main effect of load was significant and distractor affected precision on low load condition. When all participants were divided into two groups according to their performance in the perceptual task, the distractor effect only existed in the group with lower perceptual precision. Pearson correlation analysis further revealed that perceptual precision could predict the ability to inhibit distractors, in either WM capacity or precision. The quality of the perceptual representation in the lower group negatively correlated with the distractor effect in WM capacity while the perceptual precision in the higher group negatively correlated with the distractor effect for WM precision. The present study has shown that interference influenced the quantity and quality of WM representations differently under certain WM load. And individuals with lower perceptual quality would be more susceptible to interference. These findings imply that the performance and the ability to suppress distraction during WM may be enhanced via training of perceptual precision.

Key words:  visual perception, working memory, representation precision, selective attention