Purchase this article with an account.
Anna Shafer-Skelton, Timothy Brady; Evaluating the independence of working memory for scene layout and simple features. Journal of Vision 2020;20(11):1576. doi: https://doi.org/10.1167/jov.20.11.1576.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
A fundamental question in cognition is how our memory stores are structured. While neuroimaging evidence suggests that objects and the layouts of major surfaces (which we refer to as “scene layout”) may be processed separately, previous behavioral work has encountered mixed results in relation to working memory, possibly because this work has largely not isolated scene layout, instead using scenes that contain objects. To investigate scene layout vs. object memory stores, we measured dual-task costs when participants remembered three colored dots in addition to three other items: (1)a fourth, peripheral color; (2)the orientation of a peripheral gabor patch, (3)the layout of a full-screen background scene with no objects. We also collected single-task data for each of these three items on their own, attempting to match performance in these conditions. If object and scene memory rely on separate resources, remembering a scene should have a smaller cost to participants’ color memory than remembering an additional color.
In Experiment 1(N=27), performance in single-task conditions wasn’t successfully matched, with higher performance in the extra-color single-task condition compared to orientation and layout. However, adjusting for this, we find preliminary evidence of similar dual-task costs to extra-scene vs. extra-object performance, meaning we find no evidence of more independence for color vs. scene memory stores compared to color vs. peripheral gabor or color vs. peripheral color. Contrary to other paradigms that find relatively independent storage of orientation vs. color (e.g., Wang et al.2016), we also find little independence for these features. In addition, a post-hoc analysis collapsing with an additional Experiment 2 in order to better match single-task performance finds similar results. Future work that matches single-task performance can better disambiguate these hypotheses and examine under what conditions orientation and color are independently stored, providing critical insight into the independence of scene and feature memory stores.
This PDF is available to Subscribers Only