Methods for Repeated Measures Analysis

Abstract

Repeated measures of the same kind, like, for example, blood pressures obtained from a single subject at subsequent times, are different from single measures of separate subjects, because repeated measures in a single subject are generally more similar to one another than those obtained from entirely different subjects, and statistical analyses have to take this difference into account. For that purpose there are paired and unpaired t-tests, repeated and non-repeated measures analysis of variance (ANOVA) (Chaps. 1 and 2), paired and unpaired tests for binary data (Chap. 3), and paired and unpaired non-parametric tests (Chap. 2). Paired data or repeated measures means that multiple observations are performed in a single subject with the advantage that less subjects are required for answering a scientific question. A special type of repeated measures are longitudinal data including times series and survival data. They have been discussed in the Chaps. 16 and 43.