Purchase this article with an account.
Raju Sapkota, Shahina Pardhan, Ian van der Linde; Change detection in visual short-term memory: The relative impact of pairwise swaps and object substitutions. Journal of Vision 2013;13(9):324. doi: https://doi.org/10.1167/13.9.324.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
The objects that feature in our visual environment undergo frequent changes. Object file theory (OFT) proposes that, across brief periods, we use visual short-term memory (VSTM) to detect these changes (Kahneman, et al., 1992). In this experiment, we investigated how two common types of change affected VSTM performance: object swaps (wherein two objects exchange locations), and object substitutions (wherein one or more objects are replaced by new objects). Fifteen participants completed two experiments. In both experiments, a study display comprised four 2° Snodgrass stimuli presented, for 400ms, at the corners of an imaginary 8° square. Concurrent verbal loading was used for articulatory suppression. Following a 1000 ms blank interval, a test display was presented. Participants indicated, via a yes/no button click, if study and test displays matched or did not match. In experiment 1, in 50% of trials, selected at random, a change occurred. This change was either a two-item swap, or a one-item substitution (fully counterbalanced). Experiment 2 was identical to experiment 1, except that in the object substitution condition, two objects (rather than one) were replaced. Percent correct performance data were analyzed using one-way repeated measures ANOVA, with change type (swap, substitution, or no change) as a within-subjects factor. In both experiments, a significant effect of change type on memory performance was observed [Experiment 1: F(1.19,16.66)=70.68, p<0.01; Experiment 2, F(2,28)=21.75, p<0.01]. In experiment 1, object swaps were detected more accurately than object substitutions (p<0.01). In experiment 2, an inverse pattern was observed: object substitutions were detected more accurately than object swaps (p=0.02). These findings extend OFT, suggesting that change detection when two extant object files are updated is easier than when one new object file is created. Conversely, change detection when two new object files are created is easier than detecting when two extant object files are updated.
Meeting abstract presented at VSS 2013
This PDF is available to Subscribers Only