To obtain through the use of discriminant analysis a relatively bias-free sample of extraction and nonextraction, Class I, equally susceptible to both treatments' malocclusions and to analyze retrospectively the soft tissue changes between the two different treatment groups.