Florian Friedrich

In Augmented Reality (AR) applications, virtual objects are often anchored to real-world objects like furniture or bulletin boards. Little is known about how users mentally represent this spatial relationship between virtual and real objects - and how spatial memory is affected if the reference objects in the real world are moved. This study investigates how changes in the position of physical reference objects influence users’ memory for the locations of virtual objects. Users view and rate 18 AI-generated images presented as virtual objects via AR (Meta Quest 3) in a fully furnished room while incidentally memorising the objects’ locations. After a delay of 2-3 days, participants return to the room where some of the physical furniture has been moved and they have to place the virtual objects at the remembered locations. We hypothesise that participants mentally encode and represent the locations of the virtual objects in relation to nearby physical reference objects. Therefore, when the physical reference objects are moved, they should place the virtual objects at locations that are moved in the same way as the underlying furniture. This research lays the groundwork for follow-up studies that examine the use and influence of properties of the physical space in tasks where users create their own spatial layouts of virtual objects. Additionally, it will inform the design of AR applications that rely on connections between the real and virtual world under real-world variability.

The “doorway effect” or location updating effect is the decrease in performance for a simple memory task after passing through a doorway. Walking through doorways, according to the Event Horizon Model, constitutes the passing of an event boundary and switching between two mental event models - thus creating an interference effect on memory items. While previous research has consistently demonstrated this effect using a simple memory task, the present study aimed to take a closer look at the robustness of this phenomenon by applying a new, more complex object layout memory task, using a highly immersive virtual reality. No convincing evidence for the presence of a location updating effect could be found, but this study still reveals potential future research opportunities. In particular, using measures of subjective confidence as well as higher working memory loads seems to be a promising direction to examine the generalizability of the “doorway effect” to different memory tasks.

Studies using the priming paradigm often infer that unconscious processes have more veridical access to the world than conscious processes. These interpretations are based on a standard reasoning that erroneously infers good sensitivity of indirect measures from a clear priming effect. To correct for this fallacy, researchers should explicitly compute the sensitivities from indirect measures and compare them against the sensitivities of direct measures. Recent results suggest that indirect behavioral measures are not more sensitive than direct measures and challenge interpretations about veridical unconscious processing. We add to these behavioral results by focusing on neurophysiological indirect measures. In two EEG experiments, we investigate whether event related potentials (ERPs) are more sensitive to different visual stimuli than direct measures. The results show the opposite effect: higher sensitivities for direct than indirect measures. Therefore—contrasting commonly held belief—we find no evidence for more veridical unconscious than conscious processes in ERP measures.