fMRI pattern similarity analyses. (A) Schematic of pattern similarity analyses. For each subject and ROI, the similarity (Pearson correlation) between each retrieval trial pattern and each study trial pattern was computed and transformed to Fisher’s z. For each retrieved picture, category reactivation was operationalized as the difference between its similarity to other pictures from the same category (within-category similarity) and its similarity to pictures from other categories (between-category similarity). Item reactivation was operationalized as the difference between within-item similarity and within-category similarity. (B) Category reactivation (left panel) and item reactivation (right panel) for high- and low-vividness trials in the retrieval practice phase. (C) Category reactivation (left panel) and item reactivation (right panel) during retrieval practice as a function of subsequent false alarms (FA) and correct rejections (CR) during the final memory test. In both B and C, category and item reactivation were expressed as residuals after regressing out main effects of picture category from each of the raw reactivation measures. Significance was determined from paired samples t-tests for each ROI (orange = ANG, red = MPC, blue = VTC). Error bars = SEM across subjects, *P < 0.05, **P < 0.01, ***P < 0.001 (not corrected for multiple comparisons).