Neural serial-position slopes did not predict primacy across 51 patients; both correlations opposed prediction

We tested whether patients with a steeper serial-position decline in encoding high-frequency activity (HFA, 70–150 Hz) or a stronger increase in theta power (3–8 Hz) show greater behavioural primacy. The analysis used archived per-word log10 band-power features (0–1.6 s after word onset) from 51 pure-SEEG patients performing free recall in OpenNeuro ds004789, FR1 session 0. The pre-registered measures were the recall-controlled mean per-pair neural slope and behavioural primacy B = P(recall at positions 1–3) − P(recall at positions 4–12). Neither prediction was supported: HFA slope and B had Spearman rho = +0.282 (95% bootstrap CI −0.027 to +0.557; one-sided p = 0.977), and theta slope and B had rho = −0.254 (CI −0.507 to +0.039; p = 0.965). Both estimates had signs opposite to the predictions. The measures had split-half Spearman–Brown reliabilities of 0.81 for B, 0.70 for HFA slope, and 0.58 for theta slope. None of eight secondary tests was significant after Holm correction; all 54 archived subjects gave the same primary pattern. The sample had 80% power only for |rho| of about 0.38. The analysis averages over electrode pairs, uses one session and one time window, and cannot establish causation. The findings do not support a simple account in which the size of the encoding-power change sets the size of primacy.

free recall · primacy effect · serial position · high-frequency activity · theta power · intracranial EEG · individual differences · reliability · SEEG