COMMENT 7 — COMMON-SAMPLE MONTE CARLO AUDIT

This folder contains the diagnostic code and summary outputs used to
supplement the revised Monte Carlo study.

Run BLLU_Comment7_MoM_Quantile_Audit.R from this folder. The script
expects BLL_MC_raw_results.csv and BLL_MLE_diagnostics_by_replication.csv
in ../S4_Estimation_diagnostics/. Both replication-level input files are
included in this package. The S4 files contain the full Monte Carlo output
(300 replications per configuration) and full MLE replication diagnostics,
not only the earlier compact three-replication diagnostic subset.

The script regenerates the Monte Carlo samples in the original order
(seed 123), checks MoM optimizer convergence, applies the quartile-based
quantile estimator to Scenarios I–III, verifies alignment with stored
estimates, and recalculates summaries using cumulative prefixes
M = 100, 200, 250, and 300.

Undefined quartile-based estimates are counted as failures and are not
imputed. Bias, SD, and RMSE for the small-n diagnostic are conditional
on valid estimates; see Supplementary Note S3. Replication-prefix checks
use nested subsets of the same 300 replications, not independent Monte
Carlo experiments.
