Reconstructed Executable Scripts
Description
This collection contains four Python scripts reconstructing the linguistic feature-extraction and statistical analysis procedures used in the study. The scripts reproduce readability measures, 18 linguistic features, and five statistical analyses, including group comparisons, agreement tests, mixed-effects models, principal component analysis, and regression. Validation against the archived dataset of 320 legal passages demonstrates exact or near-exact reproduction of most results, with documented exceptions for MATTR-50 and simulated parallel-analysis thresholds. The scripts are corpus-validated reconstructions, not necessarily the original historical code, and are provided to support transparency, reproducibility, and methodological verification.