AX-CPT time normalization gave no clear agreement gain across two baseline runs in 52 adults with at least three valid trials per cell

We compared two AX-CPT time scores across two baseline runs in one visit. Each run had six AY and six BX trials. The primary analysis retained 52 adults with at least three correct timed trials per cell in both runs. The normalized time ratio had absolute agreement ICC 0.331. The raw AY minus BX difference had ICC 0.313. Their paired gain was 0.018, with a 95% bootstrap interval from -0.059 to 0.089. Thus, the gain remains uncertain. These data do not prove equal agreement. Both scores used the same people and trials. Five planned checks also left the gain uncertain. Twenty-one people changed score sign between runs. Six event-to-log checks matched for all 7,920 records. A full rebuild reproduced all 14 result files byte for byte. The package preserves source inputs, provenance, code, scores, bootstrap draws, figures, and reproduction instructions. The study does not test stability across days, other groups, or neural causes.

AX-CPT · proactive behavioral index · test-retest agreement · intraclass correlation · cognitive control · response time · DMCC55B