The Effect of a Verbal Concurrent Task on Visual Precision in Working Memory

Abstract

By investigating the effect of individualized verbal load on a visual working memory task, we investigated whether working memory is better captured by modality-specific stores or a general attentional resource. A visual measure was used that allows for the precision of representations in working memory to be quantified. Bayesian analyses were employed to contrast the likelihood of our data assuming a small versus a large effect, as predicted by the differing accounts. We found evidence that the effect of verbal load on visual precision and binary feature recall was small. The results were indeterminate for the size of the dual task effect on verbal accuracy and the probability of recalling a continuous target feature. These results, in part, support a multiple component account of working memory. An analysis of how the chosen effect intervals affect the results is also reported, highlighting the importance of making specific predictions in the literature.

Keywords: cognitive load; dual task; short-term memory; visual memory; working memory.

Figure 1. A trial of the staircase span/letter recall tasks.

Figure 2. A trial of the orientation recall task. For the dual task, such a trial was completed in place of the 8-second retention interval of the letter recall task. Shades of gray represent different colors. Not to scale.

Figure 3. The two effect size intervals for the small and large effect models. As can be seen, for the large effect model the majority of the prior density was distributed over values of less than 2.

Figure 4. Violin plots for the distribution of posterior estimates of the cross-modal interference effect size. Horizontal lines represent 2.5, 50, and 97.5th quantiles.