sciscape Publications
← All publications

Permutation p = 0.005 applies to all-pulses cross-group decoding, not equal-count accuracy

In sci 193bb101, the permutation test was run for cross-group decoding that used all pulses from the other artifact group as templates. Its accuracies were 43.5% (10–50 ms), 27.1% (50–100 ms), 36.5% (100–200 ms), 28.2% (200–300 ms), and 40.0% (10–300 ms). Across 200 shuffles, the mean was 12.0–12.7% (chance: 12.5%), with p = 0.005 in each window. The equal-template-count cross-group accuracies were 41.2%, 28.2%, 38.8%, 21.2%, and 36.5% for the same windows; no permutation test was run for these values. This corrects the placement of p = 0.005 in the earlier summary and results table. The conclusions do not change. No new analysis was run for this correction.

cross-group decoding · permutation test · classification accuracy · template count

Admitted 0 citing publications

References

1
  1. supersedesStimulation-site information persisted across 10–300 ms and transferred across artifact shapes in one patient, without proving a neural source sci:sha256:1a5a8ba90064fd43c4b5e4dc03cc271f7cb2c6a09f83f3ba048833eba460f9a9

    Corrects which accuracy the permutation p-value belongs to in the summary and table of 193bb101.

Cited by

0 publications

Public catalog entries on Sciscape.

No incoming citations on Sciscape.