The parahippocampal second-presentation memory effect was near zero in 14 later-session patients

We tested whether two session-0 findings in RAM RepFR1 (OpenNeuro ds005411 v1.0.1) occurred again in later sessions of the same patients. For each of 31 patients, we used the first session after session 0; 18 had medial temporal lobe (MTL) bipolar pairs. The primary result, an exploratory parahippocampal subsequent-memory effect at the second presentation of repeated words (+0.063 z in session 0), did not replicate: the later-session mean was -0.0002 z (95% CI -0.081 to +0.079; n=14; one-sided p=0.50). This failure to replicate is not proof of no effect: the confidence interval does not exclude an effect of the session-0 size. In secondary analyses, the raw MTL repetition decrease recurred at a similar size (-0.047 z; 95% CI -0.087 to -0.011; n=18), but did not remain significant after Holm correction (p=0.084). Matching for serial position gave -0.014 z (95% CI -0.046 to +0.017), as in session 0. HFA decreased with serial position for first presentations (-0.0061 z per position; 15 of 18 subjects; p=0.001). The all-MTL subsequent-memory effect for words presented once was also uncertain (n=9; mean -0.042 z; 95% CI -0.146 to +0.066; Holm p=0.80). Samples were small, and later sessions used the same patients and electrodes as session 0. The trials were new, but the subject sample was not independent. In 12 patients with parahippocampal values in both sessions, the descriptive cross-session correlation was r=-0.16.

subsequent memory effect · high-frequency activity · intracranial EEG · word recall · repetition suppression · serial position · replication · RAM RepFR1 · OpenNeuro ds005411