It is common in the laboratory to measure microbiological changes in dental plaque in subjects (animals) over periods of time. It is usually assumed that the repeated measurements from each subject are correlated rather than independent observations. The statistical procedures when correlation is present in the repeated measurements are different from those used under the independence model. There are three different statistical approaches that may be used in this situation. The simplest method is to use the univariate analysis of variance approach. The second approach is based on the multivariate analysis of variance technique. A third approach is the growth curve analysis. The growth curve analysis is based on a separate polynomial growth curve of time that is fit to each group of data. Data on microbiological changes in dental plaque from three groups of rats measured repeatedly over time were utilized to illustrate three types of analyses of repeated measurements. This study focuses on how the growth curve method can be successfully applied to analysis of microbial growth data in dental plaque since this approach has not been used by dental researchers.
