A 2000-unit response cutoff lowered the age slope of log task switch cost relative to 3000 units in 149 adults from ds005270

We tested age and local task switch cost in OpenNeuro ds005270 version 1.0.0. The main analysis used 155 adults aged 20 to 86 years. It gave both letter tasks equal weight. Correct response times from 200 to 3000 recorded units were scored on a natural log scale. The model included age and source sex. The age slope was -0.00432 per decade. Its 95% person bootstrap interval was -0.01017 to 0.00157. Overall response time increased with age. A planned 2000-unit limit gave a more negative switch cost slope. A further exploratory comparison kept the same 149 adults under both time limits. The shorter limit lowered the age slope by 0.00389 log units per decade. The paired 95% interval was -0.00628 to -0.00152. The comparison was added after the planned results were inspected. These results show that response selection can change an age estimate in this task. They do not establish a causal age effect or a neural mechanism. Milliseconds are assumed for raw time labels. The main log contrast does not depend on that unit assumption. The package preserves 163 source files, source hashes, a fixed retrieval recipe, code, all person scores, bootstrap sample indices, figures, and independent checks.

task switching · aging · response time · switch cost · cognitive control · cutoff sensitivity · OpenNeuro