Virtual reality for the study of true and false memory Gaën Plancher, Sylvain Haupert, Valérie Gyselinck, Hubert Tardieu, Serge Nicolas and Pascale Piolino* Laboratoire Cognition et Comportement, CNRS FRE-2987, Université René Descartes-Paris 5 71 avenue Edouard Vaillant 94772 Boulogne-Billancourt Cedex France *corresponding author: pascale.piolino@univ-paris5.fr Abstract An experiment was conducted in a virtual town to investigate episodic memory and false memory assuming that an active condition would increase memory compared to a passive condition. Subjects were university undergraduate students; active participants drove a car through the city, and passive participants were only passengers. Subjects were instructed to behave as if they had just arrived in a new town, and had to pay attention to everything in the surroundings. Their episodic memory was tested with a subsequent free recall test of the diverse elements of the town, and spatial and temporal details, and with a recognition test. Preliminary results showed a positive effect (p=.05) of active condition on memory for the recall of the diverse elements of the town and for the spatial location of a specific event (a car accident). Inversely, participants on passive condition were better (p<.05) at localizing the town’s elements and they had higher a global episodic score than active participants. Concerning correct and false recognitions, no difference was found. The difference between active and passive conditions is discussed in terms of divided attention in memory performance. Despite the presence of an attentional effect, our study showed that subjects memorize what is relevant to the task they were performing. Keywords: Virtual reality (VR), episodic memory, false memory, attention, sensory-motor implication Introduction Virtual reality (VR) can create the illusion of being in an artificial world. During the past decade, the use of VR has become increasingly widespread in fields as diverse as medicine, engineering, design, architecture, learning and transfer to reality. Techniques for the use of VR in neuropsychology (Marié, Chemin, Lebreton & Klinger, 2005; Marié, Klinger, Chemin & Josset, 2003; Parslow et al., 2005) and cognitive psychology (Gaggioli, 2003) are under development. Indeed, the advent of virtual environment has given a new kind of tool for studying issues of cognition. Cognitive psychology is the science that studies the mental processes that underlie abilities such as reasoning, memory, attention, or perception. VR can be a tool for studying these processes, and an increasing number of experiments are using VR to provide a more immersive and less artificial environment than more traditional experimental tasks. Using VR instead of daily life reality may be advantageous because in a computer-generated environment it is possible to manipulate variables in a fully controlled experimental situation. Moreover, VR permits experiments within a rich and personally relevant context close to daily activities. Therefore, VR appears to be a compromise between classical experimental situations and more ecological ones. A few studies on memory operations in virtual environments have been published in the past decade (Aguirre & Desposito, 1997; Andrews, Rose, Leadbetter, Attree & Painter, 1995; Brooks et al., 2002; Brooks & Rose, 2003; Brooks, Rose, Potter, Jayawardena & Morling, 2004; Maguire, Frith, Burgess, Donnett & O'Keefe, 1998; Morris et al., 2002; Wilson, Foreman, & Tlauka, 1996). Until now, VR has been essentially used to investigate spatial memory and navigation (Aguirre & Desposito, 1997; Maguire et al., 1998; Morris et al., 2002). However, VR can provide fully controlled experimental situations within a rich and personally relevant context close to daily activities which would be particularly interesting for the study of episodic memory. According to Tulving (2002), episodic memory allows conscious recollection of personal events, together with their phenomenological and spatiotemporal encoding context. It involves mental “time travel” accompanied by the personal subjective experience of remembering (i.e. autonoetic consciousness). Very simple paradigms are usually used to investigate episodic memory, such as learning lists of isolated words or pictures of objects. In these situations, the context in which the learning items are presented is very poor. It is thus difficult to generalize the results obtained to daily life, in which the context is often much richer. VR can provide events with a rich context. However, very few studies have thus for now investigated episodic memory in VR. Episodic memory has been studied using virtual environments essentially by only one group of authors (Burgess, 2002; Burgess, Maguire & O’Keefe, 2002; Burgess, Maguire, Spiers & O'Keefe, 2001; King, Hartley, Spiers, Maguire & Burgess, 2005). These authors used VR to simulate