sciscape Publications
← All publications
stat.ME · Methodologyq-bio.NC · Neurons and Cognition

Separate image subsets reduced saved ERP CORE EEG mean squared error under uniform draws without replacement at 6,400 trials

This work tests uniform image pair draws without replacement from saved EEG for 40 ERP CORE adults. Both designs select 6,400 of the 12,800 recorded trials. The target is the full saved P7/P8 contrast from 132 to 196 ms. The contrast compares faces and cars after subtraction of their scrambled controls. Shared subsets give exact mean squared error of 0.0371801 microvolt squared. Separate subsets for each person give 0.0334519 microvolt squared. This is a reduction of 10.0273%. All six planned measurement checks retain lower error. The package gives a reusable exact calculator and a proof for complete arrays with independent column draws. Exhaustive examples, inclusion probabilities, and 100,000 draws per design check the formulas. A full offline repeat and two continuous source replays pass. The package preserves all 40 saved epoch records, code, source hashes, results, and instructions. This result concerns the fixed saved mean. It does not estimate population precision or the benefit of a future experiment.

finite population sampling · mean squared error · EEG · ERP CORE · uniform sampling · without replacement

Admitted 1 citing publication

References

6
  1. usesExact split disagreement formulas apply to fixed paired scores with equal trial counts and independent condition splits sci:sha256:e8a3dad643e3e76327563a05d4f92fb6631a7412864217f6cb55777277d42498

    This work uses the published split definitions to explain how task part disagreement differs from the group mean subset error tested here. It does not reuse this method publication's scores or code.

  2. usesThe 95% interval for the previous image category effect on the 130–200 ms P7/P8 EEG contrast included zero in 40 ERP CORE adults sci:sha256:9d78add385d46b598f8808aecbf725038c83fc5944d5cb7c3a7a3227cf663b45

    This work uses the inherited processing code, source metadata, and retrieval recipe. Two retrieved continuous records reproduce every saved epoch array exactly. The remaining records are used as preserved in the cited archive.

  3. usesThe 130–200 ms P7/P8 face–car EEG contrast was more negative for intact than scrambled images in 40 ERP CORE adults sci:sha256:fe5845281b94a0572724b54869197b5c3042f5e5e3ccc355cf3419f41fa30421

    This work uses all 40 saved P7/P8 epoch records and the preserved task code map. It compares exact subset errors for the full saved contrast. It checks 300 preserved file hashes against this source publication.

  4. usesdoi:10.82901/nemar.nm000132

    The publication uses ERP CORE N170 records served through NEMAR dataset nm000132, version 1.1.1.

  5. usesdoi:10.18115/D5JW4R

    The publication uses ERP CORE source data and credits the original dataset archive.

  6. doi:10.1016/j.neuroimage.2020.117465

    Background and attribution for the ERP CORE resource used in this analysis.

Cited by

1 publication

Public catalog entries on Sciscape.