Conditional data value and decision loss in credit screening
Abstract
Replication materials accompanying “The marginal value of alternative data in credit screening: Evidence from Chinese digital lending” by Yuan Chen and Jiawei Xu. The package contains instructions for obtaining the source data, environment specifications, data processing and modelling scripts, parameter settings, random seeds, source-file checksums, aggregate results, and scripts for rebuilding the manuscript tables and figures. The full analysis was rerun in an isolated Python environment. All 41 aggregate reference CSV files matched the rerun outputs. The Table S4 implementation was reconstructed from the documented specification and numerically verified against the saved results. Convergence warnings affecting some logistic regression candidates are documented in the package. Raw borrower-level data and individual predictions are not included. Users must obtain the input data from the original providers and comply with their access and reuse conditions. Data acquisition and execution instructions are provided in the README. Funding: This research was funded by the Humanity and Social Science Research Project of Anhui Educational Committee (Youth Project), grant number 2025AHGXSK40238.