The role of attention in binding visual features in working memory: Evidence from cognitive ageing

Abstract

Two experiments were conducted to assess the costs of attentional load during a feature (colour–shape) binding task in younger and older adults. Experiment 1 showed that a demanding backwards counting task, which draws upon central executive/general attentional resources, reduced binding to a greater extent than individual feature memory, but the effect was no greater in older than in younger adults. Experiment 2 showed that presenting memory items sequentially rather than simultaneously, such that items are required to be maintained while new representations are created, selectively affects binding performance in both age groups. Although this experiment exhibited an age-related binding deficit overall, both age groups were affected by the attention manipulation to an equal extent. While a role for attentional processes in colour–shape binding was apparent across both experiments, manipulations of attention exerted equal effects in both age groups. We therefore conclude that age-related binding deficits neither emerge nor are exacerbated under conditions of high attentional load. Implications for theories of visual working memory and cognitive ageing are discussed.

Keywords:

• Visual working memory
• Binding
• Ageing
• Attention
• Central executive

Acknowledgments

This research was funded by European Research Council Grant 201312 awarded to J.R.B. who is also an Honorary Fellow at the University of Edinburgh. We thank Ian Deary, Alan Gow, and Paula Davies from the Lothian Birth Cohort Project, as well as the Lothian Birth Cohort 1936 participants, for their assistance with this project.

Notes

¹ Because binding performance was numerically lower than shape memory performance in older adults, we carried out a t test to confirm that there was no reliable difference between shape and binding memory performance in older adults, t(23) = 1.57, p = .13. It could be argued that the lack of age-related binding deficit is due to a Type II error. Note, however, that the sample size was larger here than in previous reports (e.g., Brockmole et al., 2008; Parra et al., 2009) that also failed to find age-related binding deficits and was equivalent to that used in Experiment 2 of the current paper, which does observe a reliable age-related binding deficit.