A Comparison of Three Sources of Classroom Data: Teachers, Students, and Systematic Observation.

A study was conducted to observe the varying degree of consistency between three sources of data with respect to the same classroom phenomenon. Data was gathered in a public high school near Detroit in 1969. Using the basic 10-category Flanders interaction analysis system, systematic observations of controversial issues discussions were made in the social studies classrooms of 14 teachers during 28 different discussion sessions (averaging 70 minutes per teacher). Student talk/teacher talk ratios (as indices of the extent to which students rather than teachers talk during the discussions) and student-initiated/teacher-initiated talk ratios (as indices of the tendency of students to offer their own ideas without being induced to do so by the teacher) were derived from the data. The teachers were questioned by means of a paper and pencil questionnaire about students' freedom to express their opinions during these discussions, and a sample of their students responded to a similar questionnaire. An analysis of the consistency of these sources of data was then made, using product moment correlations as indicators of between-source agreement. Apparent distortions in the teacher-reported data seem to be related to the importance which teachers assign to student opinion expression as a teaching goal. The teacher data were found to disagree with the other two sources, and were judged to be an unsatisfactory source of data for this classroom phenomenon. (Author/JES)