Proactive interference from items previously stored in visual working memory

Makovski, T., & Jiang, Y. V. (2008). Proactive interference from items previously stored in visual working memory. Memory & Cognition, 36, 43-52.

Abstract:
This study investigates the fate of information that was previously stored in visual working memory but that is no longer needed. Previous research has found inconsistent results, with some showing effective release of irrelevant information and others showing proactive interference. Using change detection tasks of colors or shapes, we show that participants tend to falsely classify a changed item as "no change" if it matches one of the memory items on the preceding trial. The interference is spatially specific: Memory for the preceding trial interferes more if it matches the feature value and the location of a test item than if it does not. Interference results from retaining information in visual working memory, since it is absent when items on the preceding trials are passively viewed, or are attended but not memorized. We conclude that people cannot fully eliminate unwanted visual information from current working memory tasks.

Note:
This paper addressed the issue whether previously attended and memorized items could influence the visual memory of the subsequent item (so-called proactive inference). If there is a proactive inference, the memory performance in the n trial will be affected by the item presented in the n-1 trial.

Exp 1 showed that memory performance of color item in the n trial decreased when the probe’s color matched that of the n-1’s item at the same location (spatial specificity). Exp 2 generated the effect to the shape feature. Also, the proactive interference was only found when the item in the n-1 trials was actively attended and memorized. Results also suggested that the VSTM is configuration-based. A very clever method used in Experiment 3 showed that both spatial and temporal interference contributed to the interference effect. 4AFC memory test in which the choices included: correct answer (true color), novel color, the color in n-1 trial at the probe location (temporal interference), and the color item adjacent to the probe (spatial interference) was used. It is suggested that the VWM is temporally and spatially imprecise.