participation in the Florida Middle School Competency Program. Using Campbell and Stanley's Institutional Cycle Design, sixty teachers who volunteered to use the module were assigned to one of three groups. The second group was then subdivided by random selection to conform to the design.
Two original instruments were designed to obtain high and low inference measures of change in teacher behavior. The Pupil Questionnaire, used to measure change in teacher behavior as perceived by students, was administered by the teachers as a pre- and posttest measure of the module. The Student Interest Observation Instrument, designed to determine change in teachers' behavior following mastery of the skills of the module, was administered by trained observers following the procedures indicated by the Institutional Cycle Design.
Findings
Hypothesis 1, "There is no significant difference between the class means on the pre- and posttest of the Pupil Questionnaire," was tested with the T Test for Related Measures. Twenty-five classes with a total of 504 children showed an increased mean score of 2.1680. This resulted in a t ratio of
3.145 which was significant at the .01 level. The null hypothesis was rejected.
Hypothesis 2, "There is no significant difference between the total even scores of the experimental groups and control groups on the observation instrument," was tested with the T viii