“A Maximum Score Test” by Walter J. Mayer Economics Department University of Mississippi University, MS 38677 662-915-5980 wmayer@olemiss.edu and Chen Wu College of Business and Technology Black Hills State University Spearfish, SD 57799-9570 605-642-6417 ChenWu@bhsu.edu January 12, 2010 Abstract We propose a new hypothesis test for maximum score estimation that can be applied to a wide range of hypotheses. Potential applications include tests of regression coefficients and distributional assumptions. The test statistic is easy to compute, asymptotically normal under the null, and converges to infinity under the alternative. We also investigate the size and power properties of the test through Monte Carlo simulations. The results suggest that the test performs well in finite samples.