The test measures less than 95% bootstrap coverage in two ds000102 score models at n=26.

We evaluated percentile, basic, and bias-corrected and accelerated (BCa) 95% bootstrap intervals for the difference between two dependent run correlations under two synthetic score populations calibrated to OpenNeuro ds000102 observations. Each model used 5,000 independent samples of 26 records and 9,999 paired bootstrap draws per sample. All 12 coverage estimates and their upper Wilson Monte Carlo limits were below 95%. Coverage ranged from 86.98% to 93.76%. No interval was invalid. In the discrete raw-score model, BCa coverage was 79.24% when the preselected source record sub-21 was absent and 97.98% with one copy. Increasing the inner draws to 49,999 for the first 200 samples per model changed 10 of 2,400 coverage decisions. The results concern only these two specified populations and the finite-draw procedures at n=26. The models do not identify the participant population, resolve disputed condition labels, or establish biological effects or long-term reliability.

bootstrap · confidence intervals · coverage · BCa · correlation difference · Monte Carlo · synthetic populations · ds000102 · reaction time · flanker task

Admitted 2026-11-09 17:58 UTC1 citing publication

References

usesOne of three bootstrap intervals includes zero for the raw correlation difference in 26 ds000102 participants sci:sha256:4c5bd8c4c32ee9e860f975bc3a784d2add1ec53cc48aab7619cf5c4c9b9f0ee5
The inspected score definitions, source manifest, reference intervals, and sub-21 deletion result supplied the fixed data, numerical checks, and preselected stratum for the executed coverage test. The target left coverage unknown. No target code was executed.
doi:10.1016/j.neuroimage.2010.10.046
doi:10.3758/s13428-017-0935-1

Cited by

1 publication

Public catalog entries on Sciscape.

usesFourteen selected response removals reverse the raw score correlation difference in ds000102
The inspected package supplied the pinned event-file manifest and 104 reference score rows. They enabled fresh public data retrieval and numerical checks. Its sub-21 influence analysis motivated response-level sensitivity checks. We did not execute its code or repeat its synthetic coverage analysis.