The 95 percent bootstrap interval included zero for ink-color Stroop prediction gain in 62 TCP test adults, assuming independent people

We tested whether first-run ink-color profiles improve prediction of second-run Stroop cell means. We used OpenNeuro ds005237 version 1.1.3. We selected 183 adults with at least two correct trials in each cell of both runs. A fixed model was fit in 121 adults and tested in 62 separate adults. The base model used mean speed and congruency. The added model also used each person's ink-color profile. Test MSE fell from 16,157.73 to 16,001.10 ms². The observed gain was 156.63 ms², or 0.97 percent. A bootstrap with 5,000 draws refit the model and resampled test people. The 95 percent gain interval was -520.56 to 709.23 ms². It included zero. This test did not establish an added prediction gain. The interval assumes independent people. It applies to selected people, one person partition, and two blocks in one session. The package preserves source files, hashes, code, trial records, predictions, bootstrap indices, results, and reproduction instructions. Independent calculations and a fresh-source rerun checked the results.

Stroop task · response time · individual differences · prediction · ink color · bootstrap interval · Transdiagnostic Connectome Project