The use of analysis of covariance to increase precision wtih DMFS measures

For an experimental study with pre- and post-DMFS measures two methods of analysis were compared with respect to precision. It is shown that for large sample size, the analysis of covariance with the post-DMFS measure as criterion and the pre-DMFS measure as the covariate will be more precise than the analysis of variance of increment scores as the value of beta differs from 1. When beta is approximately 1, the two techniques will have essentially the same precision. The dental researcher interested in precision is advised, therefore, to use analysis of covariance with the pre-DMFS measure as the covariate and the post-measure as criterion rather than the traditional analysis of variance of increment scores. For large sample size, the precision of the former will tend to be at least that of the latter. The researcher is not encouraged to use analysis of covariance with the increment as the criterion; if the procedure increases precision over analysis of variance of the increments, it merely indicates that analysis of covariance (with the post-measure as criterion) should have been used originally.