Ageing and feature binding in visual working memory: The role of presentation time

Abstract

A large body of research has clearly demonstrated that healthy ageing is accompanied by an associative memory deficit. Older adults exhibit disproportionately poor performance on memory tasks requiring the retention of associations between items (e.g., pairs of unrelated words). In contrast to this robust deficit, older adults’ ability to form and temporarily hold bound representations of an object's surface features, such as colour and shape, appears to be relatively well preserved. However, the findings of one set of experiments suggest that older adults may struggle to form temporary bound representations in visual working memory when given more time to study objects. However, these findings were based on between-participant comparisons across experimental paradigms. The present study directly assesses the role of presentation time in the ability of younger and older adults to bind shape and colour in visual working memory using a within-participant design. We report new evidence that giving older adults longer to study memory objects does not differentially affect their immediate memory for feature combinations relative to individual features. This is in line with a growing body of research suggesting that there is no age-related impairment in immediate memory for colour-shape binding.

Keywords:

• Cognitive ageing
• Feature binding
• Visual working memory

Supplemental material

Supplemental content is available via the “Supplemental” tab on the article's online page (http://dx.doi.org/10.1080/17470218.2015.1038571).

Notes

¹The term “visual working memory” is used to refer to rather different concepts in the literature on visual perception and attention compared with the literature on conceptual models of working memory or the literature on individual differences in working memory capacity. We assume that visual working memory comprises a modality-specific memory system that is part of a broader multiple component working memory (Logie, 2011), and this reflects our use of the term here. For present purposes this is broadly consistent with how the term “visual working memory” is used in the literature on temporary feature binding.

²We used the lmBF function from the BayesFactor package to set up these specific contrasts. We used the default settings, with the exception of changing the number of Monte Carlo integration samples used to compute the Bayes factors in order to keep proportional error below 2%.

³As shown in the Supplementary Material, the analysis of d′ revealed much stronger evidence against the age group by memory condition interaction (BF = 0.117). The reasons for this are unclear but it is important to note that the analysis of proportion correct and A′ yielded Bayes factors of a similar magnitude to corrected recognition (BFs = 0.502 and 0.55 respectively).