Results of systematic observation of teacher-student interaction in 35 classrooms indicate that students exert influence over classroom events and that they accurately perceive how much control they have. In a subsample of classes, students' perceived influence but not their observed influence is positively and significantly related to their academic achievement. Thus, the unidirectional model of teachor-tostudent influence reflected in previous classroom research is found inadequate to characterize patterns of interaction.Research on classroom interaction has typically proceeded from the assumption that the teacher's behavior influences or determines the behavior of the students. Teachers are perceived as playing the active role in establishing and maintaining classroom interaction patterns, while students play mainly a reactive role. This conceptualization, dating at least from Anderson's (Anderson & H. M. Brewer, 1945;Anderson, Brewer, & Reed, 1946) pioneering studies, is also cmbodied in Flanders's (1965) Interaction Analysis, perhaps the most widely known instrument for scoring observed classroom interaction. Flanders's system is designed to permit characterization of teachers' influence behavior as predominantly direct (lecturing, giving instructions) or indirect (encouraging, using a student's idea); to this end it contains seven categories for coding teachers' verbal behavior and two categories for pupils' verbal behavior (student talkresponse and student talk-initiation).

The seven major observational techniques generally used in experimental or naturalistic settings are discussed within the context of research on dominance relationships in young children. In particular, findings related to sex differences and developmental changes in children's social interactions are presented. The observational techniques used in these studies are described in terms of the type of data collected (frequency counts versus measures of duration of the behaviors), and the research questions answerable through use of each technique.Illustrations of other types of data (e.g., from questionnaires) are presented to show how the results relate to the findings from observational data. The use of more than one technique in many of the dominance relationship studies shows how different techniques can be used to examine various facets of complex social interactions.The strengths and weaknesses of each method are discussed within the framework of the illustrative studies. Problems associated with the use of One-zero or Time-Sampling techniques are discussed in detail.

The study reports relationships between teachers' belief systems (Harvey), a teacher-pupil interaction measure of classroom influence, pupils' perception of the dimate of the dassroom, and pupils' academic achievement. Results demonstrate that more complex teachers encourage more pupil influence, and are perceived as having a more origin climate. In classrooms of high pupil influence and perceived origin climate, the pupils have higher academic achievement. These results are discussed in terms of the theory of personal causation and are cited as evidence for the value of the study of action sequences as related to verbal-utterances of the actors.Human actions can be conceived as a series of discrete behaviors produced by physiological and physical changes that follow natural laws. Or behaviors may be seen as reactions to stimuli in the environment of an organism based on principles of reinforcement. Or the actions of a human being may be studied from the point of view of what they mean to the actor and the other human beings in interaction with him. The latter point of view, the study of meaningful action sequences, has recently been suggested by philosophically oriented writers (Mischel, 1969) not as a substitute for the study of 'This paper is the result of two doctoral dissertations (Koenigs, Note 1; Spring, Note 2) that were conducted as part of a larger research project sponsored by NIMH (MH 2 2 127-01) on which the last author is principal investigator. (Subsequent to the dissertation, Martha Spring resumed using her maiden name for professional purposes only). Each dissertation is a major effort in its own right, deserving of a separate publication, in our view. Since, however, the two dissertations plus other evidence from the project represent a conceptual package, and in view of the publication explosion, it was felt that it should appear in one place. Koenigs and Fiedler should, as a result, both be considered senior authors; deCharms merely served to bring it together.

No abstract