MOVICI ARTICLE 4 — DATA AUDIT

Article:
Misallocation of Epistemic Credit in Socially Embedded Learning:
An Empirically Anchored Agent-Based Model of Information Search and Debate

Public data scope
- data/article/MOVICI_SMA_ARTICLE_DATA.csv: 96 pseudonymized participants × 82 variables.
- data/article/MOVICI_SMA_DYADS.csv: 212 directed evaluator→partner observations × 10 variables.

Purpose
These minimized files contain only variables needed to initialize, calibrate, or check the Article 4 agent-based model and the empirical expertise-allocation mechanism check.

Excluded from the public package
Names, email addresses, raw transcripts, free-text responses, raw URLs, workstation identifiers, exact dates/times, and other direct identifiers are not included.

Identifiers
ID_MOVICI, evaluator_ID, and evaluatee_ID are pseudonymous study identifiers. No mapping to direct identifiers is distributed in this package.

Model/results distinction
Simulation replications are not additional human observations. Monte Carlo quantiles describe stochastic variability across simulated worlds, not sampling uncertainty in a human population.

Release status
This audit accompanies the final V3.4 STANDARD reproducibility package prepared for OSF deposition and the JASSS manuscript.
