Validity and power investigation of a test for serial correlation in univariate repeated measures analysis.

Abstract

A common design in biostatistics and psychometrics is the repeated measures design. A popular example is a clinical trial with one grouping factor, e.g. treatment or drug, and one repeated measures factor, e.g. time or visit. For selected 'real world' parameter sets, we performed a validity and power study on a test for serial correlation in repeated measures analysis and also conducted a similar study of a test for sphericity. Simulation showed that the serial correlation test identified the serial correlation structure, whereas the sphericity test, for some parameter sets, failed to detect this non-spherical pattern as often as 85 per cent of the time.