Changing block order separated congruency and block coefficients in 39 of 78 ds005265 adults under constant person effects

We inspected 623 event files from OpenNeuro ds005265 version 1.0.0. The files contain 14,952 image trials from 78 adults. Each run placed the two congruency conditions in separate blocks. Thirty-nine people kept one order. Thirty-nine changed order across runs. Exact ranks and parameter changes showed that only the changing-order design separated personal congruency and block coefficients under constant person effects. A separate block effect in every run left congruency unconstrained in all people. In the changing-order group, the estimated congruency cost was 53.8 ms. Its 95% person bootstrap interval was 38.4 to 69.1 ms. The second-block estimate was 31.1 ms, with an interval of 18.4 to 43.7 ms. An exploratory model with linear changes in both effects gave a congruency estimate of 52.4 ms. The package preserves all used source files, hashes, retrieval code, plans, proofs, analysis code, estimates, bootstrap samples, and checks. All 16 result files matched on a repeat run. These are conditional behavioral estimates. They do not establish a neural cause or a clinical trait.

approach-avoid task · congruency · block order · design identifiability · behavioral response time · OpenNeuro · ds005265