Neurobehavior Inference Lab: Working-memory load, reaction time and trial-level behavioral statistics

01 · Context

Research question

How do working-memory load and delay relate to speed and accuracy within repeated behavioral task trials?

02 · Methods

Approach

• Read BIDS behavioral events and phenotype metadata, not MRI volumes.
• Validate response correctness and prespecified reaction-time windows.
• Fit participant-intercept Gaussian/log-RT and binomial mixed-effects models.
• Quantify participant heterogeneity and rerun with alternative outlier cutoffs.

03 · Evaluation

Reported results

• R mixed-effects models were fitted to 80 OpenNeuro ds000030 participants' public SCAP event files (3,840 event rows; 3,740 trials with coded accuracy; 3,241 correct trials). Both model fits were nonsingular, with no reported optimizer convergence messages. Coefficient interpretation remains exploratory.
• The observed execution artifacts, input hashes, package versions and restrictions are documented in the research evidence ledger. Full reproducibility and sensitivity gates are still outstanding.
• Interpretation remains exploratory and no causal, diagnostic or peer-reviewed conclusion is asserted.
• Re-fitting the 80-participant SCAP task model under three correct-trial RT windows gave log-RT Load coefficients +0.04270 (0.15–3.0 s), +0.04138 (0.2–2.5 s), and +0.03919 (0.3–2.0 s). Exploratory within-task associations are not clinical or imaging effects.