With a correct linear mean and equal independent errors, 366 STReEF SEEG order contrasts had a median variance factor of 4.32

We tested how pulse time affects recorded contact order contrasts in public SEEG data. The source was STReEF, OpenNeuro ds005448, version 1.0.0. Eight patients supplied 11 runs and 372 matched conditions. Of 366 conditions with both orders, 362 had one order switch. With a linear time term, the median contrast variance factor was 4.32 relative to the difference of means. This calculation assumes a correct additive mean and equal, independent pulse errors. It does not estimate measured noise variance. A fixed voltage sample contained 96 conditions and 6,803 readout instances. Linear time adjustment changed the 20 to 500 ms contrast by 1.649 times its original root mean square value. This is the mean of eight patient medians. The before pulse control also changed, by 1.890. These changes do not identify a causal stimulation effect or measured drift. We also tested all 252 balanced label orders at the actual times of 360 standard conditions. The median best quadratic model variance reduction was 76.79%. This is a mathematical reference under unchanged responses and pulse errors, not an observed brain benefit. The package includes three figures, an event audit, exact weights, proofs, source pins, code, and replay instructions. Fresh retrieval matched all 147 metadata files and 80 raw ranges. Direct waveform checks passed for one condition in each patient. All ten numerical result files matched a separate replay. Remaining voltage conditions use verified inherited products. The publication cites that prior work.

stereoelectroencephalography · SEEG · stimulation · contact order · pulse timing · contrast variance · time adjustment · experimental design