Some dimensions of college teaching : An exploratory study using critical incidents and factor analysis of student rating

The first phase of this two-phase exploratory study of college teaching employed a modified critical incident technique. Approximately 200 students at the University of Houston reported critical teaching behaviors, some of which referred to large or small class situations. A behavior classification system was developed and frequencies of behaviors within each behavior category were compared across the graduate vs. undergraduate and large class vs. small class dimensions. Primary Phase I results were: 1. Six major 'Classes' of behaviors were designated (1) Presentation - Content Structure and Scope, (2) Presentation - Student Participation, (3) Presentation - Instructor's Style, (4) Teacher-Student Rapport and Class Interaction, (5) Evaluation of Students and (6) Requirements of Students. 2. Illustratively, the undergraduates in the sample seemed relatively more concerned with (1) the instructor's organization and clarity of presentation and (2) the instructor's straying to irrelevancies. Graduates seemed relatively more concerned with (1) their own participation in presenting materials and (2) inadequate speech characteristics of the instructor. 3. Small class size was significantly associated with a greater number of positive reports on (1) rapport, (2) encouragement of class participation and (3) 'team' assignments. In the second phase three questionnaire forms concerning college teaching were responded to by a total of 798 University of Houston students regarding their sixteen instructors of Introductory Psychology. The first form contained forty-four statements which had similarly been used in a series of University of Michigan studies. The second form consisted of items taken directly from the classification system which had been developed in Phase I. The third form combined the items from the other two forms. Factor analyses of the scores were interpreted as follows: 1. Factors from the first questionnaire showed notable similarities to five of the six Michigan factors: (1) Skill, (2) Overload, (3) Structure, (4) Interaction and (5) Feedback. Five additional clusters were tentatively named (1) Communication Sensitivity, (2) Guided Rapport, (3) Overcritical Appraisal, (4) Permissive Goal-Setting and (5) Unguided Rapport. 2. Factor analysis of the second form showed an unexpected division of the typical 'general teaching skill' factor. Nine factors were labeled (1) Presentation Vitality, (2) Presentation Clarity, (3) Classroom Interaction, (4) Clear Requirements, (5) Personal Interest and Help, (6) Presentation Overcomplexity, (7) Unreasonable Requirements, (8) Dictatorial and Derogatory Behaviors and (9) Structure. Based on the overall results the following general conclusions were reached: Phase I seemed to show that the critical incident approach holds promise for (a) identifying and describing variables of instructor behavior, (b) use with other populations that are important in college instructor evaluation, (c) studying relationships between other variables and differing patterns of perceived instructor behavior and (d) aiding in the development of forms which may increase the objectivity of teacher appraisal instruments. Phase II factor analyses of (a) the items from the critical incident classification system suggested that the combined approach may be useful in the possible identification of new factors, (b) the Michigan items supported the earlier Michigan results and, while suggesting tentative extension of the factors, demonstrated inter-school similarities in student perceptions of teaching. Possibilities for further investigation emerged. For example, there may be merit in pursuing the notion that critically effective behaviors are of a different kind than critically ineffective behaviors and, further, that an instructor's lack of effective behavior patterns does not necessarily make for 'ineffectiveness' and vice versa. In sum, the overall approach used in this study appeared to have promise as a model for producing more definitive descriptions of classroom processes, for indicating relationships between certain kinds of classroom variables, for describing characteristic patterns of instructor behavior and for generating hypotheses for future investigation.Psychology, Department o