Use of analysis of covariance in periodontal clinical trials.

This article informs that in clinical trials of periodontal disease the initial state of the tissues, and/or some pre-test measure of the relative incidence of disease have special importance. If the purpose of the experiment is to observe the effectiveness of an agent in reducing a condition, such as inflammation, calculus deposits, pockets, etc., which is present to a varying degree among the subjects, those having the condition to the highest degree have the greatest potential for improvement. Increased efficiency can be judged approximately by comparing the error sums of squares for the covariance analysis with sums of squares for the analysis of increments.