Messages (7) Incoming (4) Overdue (4) My Assignments (0) Active (9) Archive (8)

All P-values Are Uninterpretable: The Case for Abandoning Significance Testing Jul 28, 2026 Rejected Overdue 100d Machine Learning Always Outperforms OLS: Evidence from 1,000 Real Datasets Aug 17, 2026 Rejected Overdue 80d Pre-registration Does Not Reduce False-Positive Rates: A Simulation Study Sep 6, 2026 Rejected Overdue 60d Bayesian Network Analysis of Social Media Polarization: A Cross-Platform Study Oct 1, 2026 Withdrawn Overdue 35d An Adversarial Collaboration on the Ego-Depletion Effect: Registered Report Mar 10, 2026 Published Overdue 240d A Unified Framework for Causal Mediation Analysis with Multiple Mediators Apr 19, 2026 Published Overdue 200d Structural Equation Modeling Best Practices for Confirmatory Factor Analysis May 9, 2026 Published Overdue 180d Randomization Inference for Small Samples: Power and Size Control Jun 8, 2026 Published Overdue 150d

Back to Articles

Machine Learning Always Outperforms OLS: Evidence from 1,000 Real Datasets

Closed submission Rejected Overdue 80d

Editor: Sofia Reyes
Submitted: Aug 17, 2026
Due: Aug 17, 2026
Reviews: 1 of 1
Status: Rejected

Author

Abstract

We compare 12 supervised learning algorithms against OLS regression across 1,000 publicly available datasets. In every dataset, at least one ML algorithm achieves lower out-of-sample MSE than OLS. We conclude that OLS should be abandoned in applied empirical work in favor of machine learning alternatives.