Observing and Delineating Technique of Managing Behavior in Classrooms
Jacob S. Kounin
If all observations have a locus in reality, records or reports of observations should be evaluated on two bases: (1) their objectivity, and (2) their completeness.
The degree of objectivity of an observational record is conventionally determined by some measure of the degree of agreement among independent observers. Do different people seeing the record agree on what happened? If a record, for example, reads that "John caused trouble in school today" and one observer "sees" John throwing spitballs while another observer "sees" John urging other boys to smoke in the washroom then this record lacks objectivity. An objective record must enable all who see it to see the same thing or event.
Completeness is another matter. Does the observation report everything that happened at the time and place? A record may be objective yet incomplete. Thus, sixty different children, a teacher, and a principal may all agree in reporting that "John was in school today." This record may be adequate for an attendance record but it is not a complete record of all that happened to John in school that day and of everything that John did while there. A good record, then, should not only tell the truth but should tell the whole truth.
This author and his colleagues spent five years studying desist techniques -- techniques used by teachers in attempts to stop a child's misbehavior. We accumulated many objective records but they were not complete records. The very selection of misbehavior-desist events as a unit of study and the neglect of all other events in classrooms constituted a lack of completeness. Apparent differences in findings about desist events ob-
This essay appeared in Journal of Research and Development in Education, vol. 4, no. 1, Fall 1970, of the College of Education, University of Georgia, Athens, Georgia.