Abstract
Prior functional magnetic resonance (fMRI) studies examining the neural correlates of retrieval success and precision have reported inconsistent results. Here, we examined the neural correlates of success and precision in a test of memory for spatial location. The present study extended prior findings by employing an experimental design that minimized temporal overlap between mnemonic and visuomotor processing. At study, participants viewed a series of object images, each placed at a randomly selected location on an imaginary circle. At test, studied images were intermixed with new images and presented to the participants. The requirement was to make a covert recognition memory judgment to each image and to attempt to recall its studied location, guessing if necessary. A cue signaling the requirement to make a location memory judgment was presented 4 seconds after image onset. Memory precision was quantified as the angular difference between the studied location and the location selected by the participant. In an analysis that combined the data from the present study and a closely similar prior study, we replicated prior reports that fMRI BOLD activity in the left angular gyrus (AG) and the hippocampus tracks memory precision on a trial-wise basis. Linear mixed effects modeling indicated that the activity in the two regions explained independent sources of variability in these judgments. In addition, multivoxel pattern similarity analysis revealed robust evidence for an item-level reinstatement effect (as indexed by encoding-retrieval overlap) in the left AG that was restricted to items associated with high precision judgments. These findings suggest that the hippocampus and the left AG play non-redundant roles in the retrieval and behavioral expression of high precision episodic memories.