This repository contains the reproducible computational implementation used for the cross-domain evaluation described in the REPROPREP framework study. The notebook implements experimental workflows to assess preprocessing strategies across multiple datasets from the UCI repository, including systematic quality degradation, stratified cross-validation, and statistical comparison procedures with Benjamini–Hochberg false discovery rate correction. The experiments evaluate preprocessing effectiveness across diverse application domains and data quality conditions, providing empirical evidence supporting the REPROPREP methodology for preprocessing validation in business analytics. The notebook includes data preparation procedures, model training pipelines, evaluation metrics, and statistical testing routines necessary to reproduce the results reported in the associated research manuscript.