During the 1982-83 school year, public school districts in North Carolina were given the option of using clinical supervision as part of a state-wide performance appraisal process. This option resulted in considerable variation in the implementation and institutionalization of clinical supervision in schools and provided the opportunity to study variables associated with the institutionalization of change in school systems. Berman’s (1981) implementation paradigm was used to identify and categorize predictors of institutionalization. A set of five variables was selected as having the best potential for accounting for the variation in institutionalization of clinical supervision in the public schools of North Carolina. 1. Principal’s perception of the amount of time required to perform one clinical supervision cycle. 2. Amount of internal support for clinical supervision. 3. Amount of training in clinical supervision. 4. Principal’s belief in the effectiveness of clinical supervision. 5. Type of school administered. A <u>Principal’s Survey</u> was developed and mailed to a random sample of 450 public school principals in North Carolina. Information was received from 288 principals (64%). A follow-up survey of nonrespondents verified the representativeness of the original respondents. Principals responding to the follow-up survey were added to the original respondents for a total sample of 300 principals (67%). Multiple regression analysis was applied to the data with institutionalization as the dependent variable. The multiple R was .30 and R² was .09 (F = 4.03, p < .00). Internal support was the only significant predictor of institutionalization (b = .20, t = 3.62, p < .05). Two demographic variables, age and sex, were added to the multiple regression as a side analysis. With these 4 variables added, an R² of .10 was obtained (F = 3.37, p < .001). Age was determined to be a statistically significant predictor of institutionalization (b = -.06, t = -2.48, p < .05). / Ed. D. / incomplete_metadata