The catFR1 alpha-specific memory effect was not significant, while beta power again did not persist beyond the broadband tilt

We tested whether band-specific alpha and beta subsequent-memory findings from a 51-patient FR1 study replicate in categorized free recall (catFR1). We analyzed 21 session-0 SEEG participants using five wide frequency bands in a 0–1.6 s window and applied the archived band-specific analysis code unchanged. Raw alpha and beta power were lower for later-recalled words (alpha: −0.110 dB, p = 0.017; beta: −0.081 dB, p = 0.015). After removal of the broadband log-log line, neither alpha nor beta residual was significant after Holm correction (alpha line: −0.006 dB, Holm p = 0.91; beta line: −0.017 dB, Holm p = 0.91). The alpha per-pair regression residual was −0.050 dB (95% bootstrap CI −0.112 to +0.022; Holm p = 0.54), similar in size to the FR1 estimate (−0.053 dB, Holm p = 0.014) but not significant in catFR1. The alpha-specific result therefore did not replicate with significance; the beta-line null replicated. In 10 catFR1 participants not in the FR1 set, the alpha residual was −0.096 dB (Holm p = 0.12). Pooling these participants with the FR1 discovery data gave −0.060 dB (Holm p = 0.0008), but this was not an independent test. The archived analysis choices, five broad bands, one time window, session 0, and correlational contrast limit interpretation.

categorized free recall · SEEG · subsequent memory · alpha power · beta power · broadband tilt · replication · OpenNeuro