Abstract
Working memory research has largely focused on the visual domain, however, a full picture of working memory requires understanding its properties in other domains (e.g., auditory). Recently, working memory tasks have focused on circular reproduction tasks, since these allow the separation of putatively distinct mechanisms. Critically, such tasks have been leveraged only to study visual features (e.g., color & orientation) and have not been expanded to the auditory domain. Here we developed a new methodology: Auditory reproduction task (ART). To overcome the challenge of creating a circular space with auditory information, we relied on Shepard tones which create an illusion of infinite rising or falling tone frequency. We converted an octave range of Shepard tones to a 0-360° circular space. In the first experiment, we validated the perceptual circularity of the Shepard tones with multidimensional scaling. The perceptual space of the tones resembled an almost perfect circle. In the second experiment, we demonstrated that auditory working memory shows set-size effects, similar to visual working memory. Specifically, reproduction errors increased when participants were required to retain two (sequentially presented) versus one tone. Further, the preliminary results, when subjected to mixture modeling, revealed that precision decreased as a function of set size, demonstrating that the theoretical models on visual working memory can be used to study auditory working memory. Similar to visual working memory, we found evidence that the non-tested item influenced responses to the tested item—repulsion when the distance between the two random tones was small and attraction when the distance was large. Taken as a whole, these findings validate the ART task as a useful tool to study the properties of auditory working memory and how they may be similar to (or differ from) visual working memory.