Switching to an Allocentric Spatial Reference Frame Incurs Distinct Cognitive Processing Costs

The ability to acquire, recall, and flexibly use information from first-person (egocentric) and map-like (allocentric) reference frames is essential for human wayfinding and thus, daily life. Prevailing models of the neural systems that support the representation and coordination between reference frames suggest that switching back-and-forth should result in either lower accuracy, slower response time (RT), or both. However, predictions of these models have not been adequately tested for spatial memory of large-scale, naturalistic environments. Inspired by previous work on task switching, we developed a novel task in which participants performed a spatial memory switching task that included trials in both egocentric and allocentric reference frames. Participants learned a large virtual neighborhood by navigating to various target stores. They then completed runs of the memory task, with approximately equal numbers of egocentric and allocentric trials. Critically, we manipulated whether the same reference frame was repeated or switched on a trial-to-trial basis to determine whether there is a cost associated with changing reference frames. Results showed no significant differences in spatial memory accuracy across conditions but significantly slower RTs for switching trials. Importantly, this reference frame switch cost was only observed for egocentric-to-allocentric switching (not vice versa) relative to all other conditions. These findings suggest that the cognitive process of reference frame switching is asymmetric, which went against our predictions from neural models. Future neuroimaging studies to test predictions generated by our results will provide crucial insight into human spatial memory function.