Participants first navigated a virtual reality environment in which they learned both the spatial layout of the city and the temporal order of deliveries of a passenger to different stores (Fig. 1A). Following encoding, we probed participants' contextual memory for their virtual reality experience (Fig. 1B) while undergoing fMRI. Questions consisted of a reference store followed by two probe stores; participants indicated which of the two probe stores was closer in either spatial or temporal distance from the reference store. Behavioral accuracy was 64.4 ± 9.7% for spatial and 61.5 ± 9.7% (mean ± s.d.) for temporal questions. Importantly, we found no difference in performance between conditions (paired t test, t(15) = 0.954, p = 0.355; for a complete description of the behavioral design and implementation, see Ekstrom et al., 2011). Analyzing the fMRI data, we first focused on two fundamental contrasts using functional connectivity and graph theory: 1) correct versus incorrect retrieval (i.e., trials on which the participants correctly judged the spatial or temporal distance compared to incorrect trials) and 2) spatial compared to temporal retrieval trials.
