With ten-trial counts fixed, the test did not establish excess adjacent slow pairs in 67 ds004902 v1.0.8 PVT files

We tested whether slow responses formed extra adjacent pairs in all 67 PVT files in OpenNeuro ds004902 version 1.0.8. PVT means psychomotor vigilance task. The sample had 3430 responses from 38 stored participant codes. A slow response had RT > 500 ms. The primary null gave uniform order within ten-trial blocks, with each slow count fixed. Only edges within blocks were counted. Each participant code had equal weight. There were 56 pairs, with 49.23 expected under this reference. Mean excess was 0.229 percentage points. The 95% person bootstrap interval was -0.123 to 0.636 points. The one-sided Monte Carlo p value was 0.08995. The primary test did not reject at 0.05. This does not prove that clustering is absent. Some planned secondary tests gave smaller p values. We report all seven checks. The package supplies an exact count proof, checked code, source bytes, hashes, random samples, and repeat instructions. A fresh source repeat matched all 63 result files byte for byte. Missing trial timing and uncertain source links restrict the claim. This is a behavioral sequence result, with no causal sleep or EEG claim.

psychomotor vigilance task · sleep deprivation · response time · trial sequence · slow responses · adjacent pairs · conditional randomization · OpenNeuro ds004902

Admitted 2026-11-09 11:06 UTC3 citing publications

References

usesA count and trim rule with separate cutoffs matched 49 of 67 PVT score sets in ds004902 version 1.0.8 sci:sha256:650bc2d86c86fd27aff0acef6eae7cd0083c5e5deb08581a3977eae256b02b6f
We used the fixed ds004902 v1.0.8 source manifest and archived bytes from this score audit. Its file coverage and source link caution defined our scope. Fresh downloads, Git blob checks, and a full source tree check confirmed the bytes and all 67 PVT paths. We computed new trial-order scores and did not reuse its score rules or analysis code.
usesdoi:10.18112/openneuro.ds004902.v1.0.8
The analysis uses the 67 stored PVT files from this exact OpenNeuro release. The selected source bytes and retrieval details are preserved with their hashes.
supportsdoi:10.1038/s41597-024-03268-2
The dataset article supplies the task context, response unit, and 500 ms slow-response boundary used in this publication.
supportsdoi:10.1093/sleep%2F32.10.1393
This PVT study supports the stated limitation that stimulus intervals can affect responses; its data were not used.

Cited by

3 publications

Public catalog entries on Sciscape.

usesOnly 9 of 170 five-minute PVT visits in ds000201 v1.0.3 had variable adjacent lapse pair counts with fixed minute counts
Its adjacent lapse method with fixed local counts motivated our question and reference score. We used ds000201 time blocks, original row edges with gaps, and a paired sleep contrast. We did not use its ds004902 data or code.
usesThe 95% interval included zero for the assigned sleep restriction change in local PVT rank order during 300 seconds in 85 ds000201 v1.0.3 adults under additive visit effects without carryover
Its summary motivated a local-order reference that holds local response values fixed. We test continuous ranks in a different dataset, with event-time blocks and assigned visit order. We did not copy its code or pool its cohort with ours.
usesAssigned sleep restriction increased the ten-response PVT rank score in the first 300 seconds for 85 ds000201 adults under common additive visit effects within age groups and negligible carryover
Its fixed local slow-response-count reference motivated a local conditional order reference. We instead used within-block ranks and an exact reference for squared differences on fixed adjacency graphs. No code or source data from that study were reused.