The latest paper from Siobhan’s PhD work, published in Journal of Vision, suggests so. In this study, we examined the serial position effect in spatial working memory using a continuous report task to explore how sources of recall error change across different positions in a sequence. We compared the sources of recall error in full and partial report tasks. The key take-home for this paper is that the serial position effect in precision in spatial working memory depends on whether a full sequence or only one item from that sequence is recalled. We found that when a full sequence is recalled, items presented early in this sequence are recalled more precisely than subsequent items (primacy effect). In contrast, when only a single item is to be recalled, this serial position effect is reduced and reversed. This finding reconciles previously contradictory findings in visuospatial working memory, where differences in the nature of the recall task appeared to determine whether a recency or primacy effect was observed. The key data are captured in the figure below:
Mean imprecision for whole report (sequential presentation mode only) and single probe tasks as a function of serial position at each set size. Error bars represent the standard error of the mean.
This finding is important because it suggests that the primacy effect in the whole report task results from accumulation of noise/error due to performing multiple goal-directed actions during recall, whereas only a single action was required in the partial report task. Additionally, this finding has important implications for models of capacity, and suggests that spatial STM is flexible and dynamic, with resource allocation being determined by the nature of the recall task.
You can read the paper open access here: https://jov.arvojournals.org/article.aspx?articleid=2785388 and the research was funded by an ESRC NINEDTP studentship awarded to Siobhan McAteer.
Leave a Reply