NONPARAMETRIC ADAPTIVE DISTRIBUTION-FREE PROCEDURE FOR CROSSOVER DESIGN WITH REPEATED MEASURES

Abstract

We propose to apply adaptive nonparametric procedures (Hill, Padmanabhan, & Puri, 1988) on 2x2 crossover design with repeated measures. We will derive the test-statistics (based on function of ranks) and find their asymptotic distributions. These test-statistics will be used to test (a) equality of carryover effects; (b) equality of direct treatment effects; (c) equality of carryover effects over time (repeated measures); and (d) equality of direct treatment effects over time (repeated measures), as suggested by Johnson and Grender (Johnson & Grender, 1993). We will be testing these hypotheses using modified versions of the test statistics derived by Johnson and Grender (Johnson & Grender, 1993) and Brunner et al. (Brunner, Domhof, & Langer, 2002) tailored to the underlying distribution of the data. In addition, we provide examples to illustrate the new methods.

The methods proposed extend the methods developed by Sun (Sun, 1997) for c-sample problems.