MTL high-frequency activity did not show a significant lag-dependent recall effect in a pre-registered spacing test

A pre-registered test examined whether medial temporal lobe (MTL) high-frequency activity (HFA) at a word’s second presentation predicted later recall more strongly at longer lags, as expected by the study-phase retrieval account of the spacing effect. We analyzed archived 70–150 Hz HFA from 0.10–0.75 s after word onset for 26 SEEG patients and 4,589 repeated-word items in session 0 of OpenNeuro ds005411 v1.0.1. The mean within-subject HFA-by-log-lag interaction was +0.0154 recall probability per HFA SD per log-lag unit (95% bootstrap CI −0.0056 to +0.0356; 16/26 subjects positive; one-sided sign-flip p = 0.081), so the primary hypothesis was not supported. Recall increased with lag (+0.0451 per log-lag unit, 95% CI +0.0092 to +0.0809, p = 0.024), showing a behavioral spacing effect in these items. The HFA main effect and change-score interaction were also not significant. This null neural result is only weak evidence against study-phase retrieval: the analysis used archived MTL features, sparse hippocampal coverage, one time window and frequency band, within-list lags, and 26 subjects.

spacing effect · study-phase retrieval · medial temporal lobe · high-frequency activity · SEEG · free recall · repetition lag