In ds004789 FR1 SEEG, no serial-position difference was detected in the encoding HFA-minus-theta subsequent memory effect

We compared the encoding subsequent memory effect (SME; recalled minus not-recalled words) across primacy (serial positions 1–3), middle (4–9), and end (10–12) words in 54 subjects performing delayed free recall (FR1), session 0, in OpenNeuro ds004789. For each bipolar pair and band, log10 power was z-scored across words within each serial position. Our measure was z(HFA, 70–150 Hz) minus z(theta, 3–8 Hz), averaged over pairs. The primary subject-level sign-flip contrasts found no detected group difference: middle minus primacy was +0.025 z (95% CI −0.030 to +0.079; n=49; Holm-adjusted p=0.73), and end minus middle was −0.009 z (95% CI −0.072 to +0.052; n=44; Holm-adjusted p=0.78). The per-subject interaction slope was −0.0001 z per position (95% CI −0.009 to +0.009; p=0.98). The tilt SME was +0.055 in primacy (p=0.049), +0.067 in middle (p=0.004), and +0.034 at the end (p=0.33). These null contrasts do not establish equality: their intervals include differences about as large as the middle-group SME, and the analysis can only exclude differences larger than approximately 0.08 z. Recall probability averaged 0.41, 0.22, and 0.20 in primacy, middle, and end groups. Limits include unequal group sizes, a five-word minimum per class, one task and session, and pooling all recorded contacts without regional separation.

subsequent memory effect · serial position · SEEG · free recall · high-frequency activity · theta · OpenNeuro ds004789