Savin, N. E. and A. H. Würtz, Power of Tests in Binary Response Models. Econometrica, vol. 67, 1999.

Abstract

Most hypotheses tested in binary response models are composite, that is, the null hypothesis restricts only a subset of the parameters. The remaining parameters are referred to as nuisance parameters. The null hypothesis is usually that one or more of the slope coefficients are zero. Typically, the sequence of alternatives of interest is one in which the slope coefficients are increasing in absolute value. In this paper, we prove for any fixed sample size that the power goes to zero for this sequence of alternatives in cases which often occur in practice.