Abstract
There exists a long history of capacity limitations cited in the visual working memory (VWM) literature. It is less clear however, whether the types of limitations commonly observed reflect unique properties of the visual system per se or rather a more general property of perceptual stimuli encoded in working memory. In other words, should the study of VWM be replaced by the more encompassing concept of perceptual working memory? We make a qualitative comparison of memory performance between two perceptual modalities: haptic working memory (HWM) for object width, and VWM for line length. In the haptic memory experiment, participants were presented with a sequence of 1–4 object widths (experienced with the thumb and pointer finger) and made a 'wider or narrower' judgment against a probe item. We compared performance in the haptic memory experiment with previous data examining visual working memory for line length (Sims et al., 2012). The analyses from both studies were approached using rate–distortion theory, a branch of information theory. The results yielded several qualitatively similar findings. First, HWM performance like VWM performance showed a set size effect. This decline was notably more substantial in HWM performance than that of VWM, particularly when more than one item was held in memory. Additionally, discrimination for haptic width degraded in a manner similar to visual line length discrimination as set size increased. Lastly, both analyses revealed biases to the mean of the stimuli distribution and the magnitude of this bias increased as set size increased. This finding suggests that implicit statistical learning plays a central role in both visual and haptic working memory. The relatedness of the findings between the two sensory modalities suggests that characteristics of performance commonly observed in VWM may be better understood as characteristics embodied by a larger perceptual working memory system.
Meeting abstract presented at VSS 2016