Mean city response time before mountain errors was 38 ms lower within stored codes in 23 people from ds007216 v1.1.0 under fixed assignments

We separated the city response-time difference before mountain errors into a term for code shares and a term for speed within codes. We used all 43 GradCPT event tables in OpenNeuro ds007216 v1.1.0. The main comparison retained 1,449 pairs in 110 cells from 23 people. Each cell contained both mountain outcomes and both city response codes. The mean error-minus-withholding difference was -103.66 ms. The code-mix term was -65.36 ms. The within-code term was -38.30 ms, with an approximate family 95% interval of -56.03 to -21.15 ms. We sampled whole people 50,000 times. A broader run-only check retained 2,560 pairs and gave a within-code term of -43.83 ms. A stricter count check retained 16 people, and its interval included zero. This is a follow-up study of a known effect in the same people. It assumes fixed stored trial assignments. The terms do not identify causal effects. The package preserves the plan, code, source files, row decisions, results, and verification logs. Offline and fresh-download runs reproduced all 26 numeric files byte for byte.

sustained attention · response time · commission errors · response codes · decomposition · behavioral data · OpenNeuro