Pre-word SEEG theta and HFA power (-0.75 to 0 s) did not predict later free recall in 28 RAM FR1 subjects (ds004789, bipolar, list and serial position controlled)

We tested if SEEG band power in the 0.75 s before word onset predicts later free recall. We used OpenNeuro ds004789 (RAM FR1, session 0, bipolar pairs) and 28 subjects. Word onset is the EDF sample time. For each pair, a regression controlled list, serial position and recall of the previous word. A sign-flip test across subjects did not find a pre-word theta effect (mean 0.000 SD, 95% CI -0.030 to +0.030, p = 1.0) or a pre-word HFA effect (mean -0.005 SD, CI -0.020 to +0.010, p = 0.51). The same model on the post-onset window (0 to 1.6 s) found the known negative theta and alpha effects (theta -0.038 SD, p = 0.036; alpha -0.052 SD, p = 0.001). Thus, if a pre-word theta effect occurs, it is smaller than the post-onset theta effect. An exploratory, uncorrected test found lower pre-word HFA in hippocampal pairs (-0.067 SD, p = 0.018, 14 subjects). This result needs a new test. The package contains the pre-registered plan, the code, per-subject features, provenance hashes and results.

SEEG · free recall · subsequent memory effect · intracranial EEG · theta power · high-frequency activity · pre-registration