Normalized trial-wise SEBRs during memory recall in experiment 3 (A) and experiment 4 (B). LME analysis suggested higher SEBRs in CONA+ than in all CTX− conditions in experiment 3. Post-hoc paired t-test suggested the first subset (containing the first trial of each condition) and the average of the first two subsets (containing the first and the second trials of each condition) are the optimal trial numbers to index memory retention. These results were confirmed in experiment 4. Group data are shown as group mean ± SEM. “CTX−” is the average over all four CTX− conditions. The top of each plot shows the linear trial index of CTX− conditions, and the bottom shows the linear trial index of the CTX+ condition.
