Abstract
Combining virtual reality (VR) with EEG and eye-tracking offers new possibilities for the investigation of cognitive processes. It is unclear to what extent established findings from studies using conventional monitors translate to setups using VR head-mounted displays. Here, we assessed whether EEG markers of visual short-term memory load, the evoked-response amplitude of the contralateral delay activity (CDA) and the power of induced alpha oscillations, can be replicated with VR glasses and whether they depend on the horizontal eccentricity of the stimulus in the visual field. We tested observers’ visual memory in a delayed match-to-sample task with bilateral stimulus arrays of either two (low memory load) or four (high load) coloured circles. Moreover, we varied the horizontal eccentricity of the stimulus arrays (either 4, 9 or 14 degrees of visual angle). At the beginning of each trial, we displayed a cue informing observers which array (left or right from fixation) would be probed in the later memory test. We presented the stimulus arrays for 200ms, followed by a retention interval of two seconds. Simultaneous eye-tracking allowed us to include only trials without saccadic eye movements in the final analyses. Observers’ memory performance was better for low as compared to high memory load—irrespective of stimulus eccentricity. Both, CDA amplitude and alpha power during retention, increased significantly with memory load across all eccentricities. We further corroborated these findings by fitting time-resolved spatial filters to decode memory load from voltage as well as time-frequency data. Classification accuracy during the retention interval was above chance level for both, the evoked potential and induced alpha power, and did not vary significantly across eccentricities. We conclude that memory load effects on CDA amplitude and alpha power can be shown using commercial VR glasses and for stimuli with horizontal eccentricities of up to 14 degrees of visual angle.