RUDOLFO CHAVEZ CHAVEZ HIGH-INFERENCE MEASURES/CLASSROOM CLIMATES
The Use of High-Inference Measures To Study
Classroom Climates:
A Review
Rudolfo Chávez Chávez
CaliforniaState College, Bakersfield
Review of Educational Research
Summer, 1984, Vol. 54, No. 2, Pp. 237-261
ABSTRACT
This paper reviews the impact high-reference measures had on studying classroom climates. Early literature of classroom climates from the late twenties through the early sixties is reviewed. It is noted that the study of classroom climates had two distinct, unrelated beginnings. As a result, the use of high-inference measures to study classroom climates was not prevalent in the literature until the mid-sixties, whereas the study of classroom climates using other measures was well established by the early sixties. The review then discusses several high-inference measures developed in the last 30 years. Also, pertinent classroom climate literature that incorporates specific high-inference measures has been reviewed.
After reviewing much of the current classroom research of the time, Withall and Lewis (1963) concluded that:
Much of early research of classroom learning was set in the stark framework of the given conditions in an educational situation . . . . It seems that the researchers of the 1920’s and the mid-1930’s sought to analyze the process of teaching in terms of the conditions brought to it by the teacher as a professional worker. She [ was presumed to be devoid of any personal needs, purposes, idiosyncrasies. The other actors in the classroom, i.e. the learners, were even more taken for granted. (p. 709, emphasis added)
More recently, John A. Goodlad (1979) stated that “too many researchers are preoccupied with research on single instructional variables that rarely account for more than 5% of the variance in student outcomes. Too few [researchers]study the complex phenomena of schooling in their natural environment, developing the needed new methodologies instead of seeking to adapt the old” (p. 347). Contrary to these remarks that learners were taken for granted and that the natural environment of schooling was not being studied, there has been a sustained research effort of studying learners’ interactions in day-to-day classroom activities in the natural environment. This review will discuss that research effort in two parts.
The first part is a review of the early literature of classroom climates from the late twenties through the early sixties. It should be noted that according to the literature reviewed, classroom climate research using high-inference measures had two distinct, unrelated beginnings. (Rosenshine & Furst, 1971, define a high inference measure as a rating system that requires an observer to make an inference from a series of classroom events using specific constructs, such as satisfaction,cohesiveness. etc.) The use of high-inference measures to study classroom climates was not prevalent in the literature until the mid-sixties, whereas the study of classroom climates with low-inference measures was well established by the early sixties. (Rosenshine & Furst, 1971, define a low -inference measure as a rating system that classifies specific, denotable, relatively objective classroom behavior and is recorded as frequency counts by an observer.) Nonetheless, the early classroom climate research using low-inference measures is relevant because it serves as a cornerstone for the development of a conceptual framework of classroom climate research. The second part of this review is a discussion of several high- inference measures that have been used in the last 30 years, with particular attention given to the My (lass Inventory (MCI) and the Learning Environment Inventory(LEI).
Early Literature of Classroom Climates
Researchers who took an early interest in classroom behavior were social psychologists. Their basic interest was in the interactions between student and student and between student and teacher (Medley & Mitzel, 1963). Research emphasizing these interactions seems to have begun with the work of Dorothy Thomas (1929). Thomas mentioned that the study of classroom behavior consisted largely of descriptive accounts (i.e.. case histories and diary records). She stated, however, that those accounts presented certain difficulties as material for scientific analysis:
The data obtained in such records are, at their best, objective, in the sense that they deal with certain verifiable facts, hut they are selective, inconsistent. and usually incomparable with other records. This is due to the tremendous complexity of any social behavior act and the consequent recording of different elements of these complex acts at different times. At their worst, these records are such an intermixture of fact and interpretation as to be utterly worthless from the scientific point of view. Even at their objective best, the selection and emphasis are more or less dependent on the recorder. The control of this sort of error in our social data is one of the first problems claiming our attention. In other words our data must become independent of our observers within a small and predictable range of error. (p. 3)
To accomplish this, she concluded that the study of social behavior could best be made by evolving indices of an individual’s overt actions involving other persons compared with overt actions involving objects (or abstractions) and the self (Thomas, 1929). Three general techniques were developed: observing a child for a given period in the nursery school and recording the child’s overt social behavior each time it recurred; recording specific social situations made by the child within the nursery school: and creating a psychological test situation with limited social and material stimuli instead of the nursery school for recording data. Because of the high standards set by Thomas (i.e., accuracy and objectivity), research many years later would have similar standards (cf. Medley & Mitzel, 1963).
An important influence on the study of classroom climates was the work of Lewin, Lippitt, and White (1939) and Lippitt (1940, cited in Withall, 1949). These studies examined the effect of three leadership roles and the concomitant group climates by observing the behavioral patterns of four “clubs” of five 10- or Il-year-old boys. Each club was examined under three leadership conditions: democratic, autocratic, and laissez-faire. In addition to matching the groups to control for individual differences, leaders were rotated to control for treatment variations. The data recorded were (a) the interaction within each group, (b) the interaction between the leader and the group, (c) the overt expression of aggression, and (d) the productivity in club projects.
The findings of both studies were similar. Lippitt (1940) concluded the following:
1.Different leadership styles produced different social climates and resulted in different group and individual behaviors.
2.Conversation categories differentiated leader behavior techniques more adequately than social behavior categories.
3.Autocratic leadership elicited either an aggressive rebelliousness toward the leader or an apathetic submission to the leader.
4.Leadership style was the primary factor in producing climatological differences and club personnel were of secondary importance.
For Lewin et al. (1939), the following conclusions stood out above all others:
1.Aggressive behavior was either very high or very low under authoritarian conditions, extremely high under laissez-faire conditions, and intermediate under democratic conditions.
2.Productive behavior was higher than or as high in authoritarian climates when the leader was present as in democratic climates but much lower when the leader was absent, moderately high and independent of the leader’s presence or absence in the democratic climates, and lowest in the laissez-faire climates.
Shortly after the Lewin et al. and Lippitt studies came investigations to determine the influence of teachers’ classroom personality on students’ behavior and students’ classroom behavior on each other (Anderson & Brewer, 1945; Anderson & Brewer, 1946) and the development of a method to observe students’ and teachers’ behaviors simultaneously (Anderson, Brewer, & Reed. 1946). These researchers developed 23 teacher behavior categories and 7 student behavior categories. More important, they divided the behavior into socially integrative or dominative behavior. According to the researchers, the former behavior promoted the interplay of differences and advanced “the psychological processes of differentiation, [that] facilitate[d] the emergence of originals” (Anderson & Brewer, 1945. p. 9). The latter behavior was “characterized by rigidity or inflexibility of purpose, by an inability or an unwillingness to admit the contribution of another’s experience, desires, purposes, or judgment (Anderson & Brewer. 1945. p.9). Socially integrative behavior (among teachers and students) was flexible, adaptive, objective, scientific, cooperative; dominative behavior, on the other hand, was seen as stifling someone else’s behavior coupled with a resistance to change and consistent with bigotry and autocracy (Anderson & Brewer, 1945).
The Anderson et al. (1946) study illustrated that students’ as well as teachers’ behaviors could be categorized and “it demonstrat[ed] further the validity of this method of studying teachers’ classroom personalities” (p. 153). However, because of the limited population (data were collected on four teachers and four students) it was impossible to estimate the reliability of any of the scores proposed for comparing different teachers, classes, or occasions in the same class. Moreover, itwas recognized that the validity of the measures of teacher behavior would have been higher if the proposed categories were internally consistent (Anderson & Brewer, 1945).
As the fifties approached, classroom climate research became theoretically and empirically oriented. The streams of thought captured for this orientation were Lewin’s (1936) field theory, Murray’s (1938) need-press model, Thelen’s (1950) educational dynamics model, and others. Hypotheses were derived from analysis of time-lapse pictures, recordings (stenographic and mechanical), and observations in the classroom by Sensitive and trained educators using newly developed measures, which were often compared with the results of standardized tests (cf. Medley & Mitzel, 1963; Withall & Lewis, 1963). The earliest of the several researchers to follow was Withall (1949). In his study, he renamed the interactions between students and students and students and teachers as the social emotional climate. This group phenomenon was defined as
a general emotional factor which appears to be present in interactions occurring between individuals in face to face groups. It seems to have some relationship to the degree of acceptance expressed by members of a group regarding each other’s needs or goals. Operationally defined, it is considered to influence: (I) the inner private world of each individual; (2) the esprit de corps of a group; (3) the sense of meaningfulness of group and individual goals and activities; (4) the objectivity with which a problem is attacked; and (5) the kind and extent of interpersonal interaction in a group. (pp. 348—349)
This definition illustrates phenomenological activities that are emotional and intellectual on the one hand, and individual and social on the other, where all activities are interactive within the classroom. However, Withal! (1949) did not seem to believe students’ interactions, as suggested by earlier studies, to be as important as the teacher’s interactions. He suggested that it should be possible to measure the socioemotional climate in terms of teacher behavior alone and developed the following seven categories to encompass all types of statements that teachers use in classrooms: “(a) learner-supportive, (b) acceptant and clarifying, (c) problem-structuring, (d) neutral, (e) directive or hortative, (f) reproving or deprecating, (g) teacher self-supporting” (p. 349). Withall perceived these categories to lie on a continuum, whereby a continuum from problem-centeredness to person centeredness could exist, or from objectivity to subjectivity, or from learner centeredness to teacher-centeredness. Therefore, Withall believed categories a, b, and c to be learner-centered and categories e, f, and g to be teacher-centered. The patterns of verbal behavior elicited by the teacher being observed would determine whether the teacher was learner- or teacher-centered or another of the seven mentioned categories. For example, if the ratio of category c outweighed the proportions of categories a and b combined or categories e. f, and g combined, then the teacher was said to be more problem-centered than learner- or teacher centered (Withall. 1949).
Withall (1949) found significant relationships among different measures of group process, pupil reactions, expert ratings, and styles of problem-solving activity. In addition, Withall and Thelen (1949) validated Withall’s Climate Index. The validation of the instrument involved (a) a comparison of the categories of the ClimateIndex with Anderson and Brewer’s (1946) Integrative-Dominance Ratio secured on the same data; (b) pupils’ pencil-and-paper assessment of their classroom situation on the basis of seven standardized questions; and (c) the use of an electrical graph for recording pupils’ positive and negative feelings expressed by their pushing one of two buttons while being exposed to experimentally varied “learner-centered” and “teacher-centered” social emotional climates. Moreover, Withall (1949) found that different teachers produced different climates with the same group of pupils.
Although the following two studies, also from the early fifties, are not directly related to classroom climate research, their results are a necessary link.1 Bovard (1951) argued that a group-centered classroom climate is more amenable to more students than a leader-centered classroom climate. He defines group-centered as having “student-to-student verbal interaction . .. fostered by a number of specific techniques, such as seating students in a circle, and deflection of teacher-directed questions back to the group” (p.215). He defines leader-centered as having “student- to-student conversations.., politely but firmly limited, and verbal interactionchanneled between teacher and individual student” (p. 215). For the study, two college psychology classes served as the experimental and control groups. Each class had the same teacher, course content, examinations, and text, and the same relative amount of time devoted to lectures, discussion, and role playing. Also, the two classes were matched on number of veterans, grade point averages of the previous semester, and scores on the Otis Self-Administering Test of Mental Ability (Form A) (Bovard, 195I). Bovard used two techniques to measure for differences:
(a)outside observation of class discussion, and (b) the administration of an “affect scale.” He found that 61% of student remarks were directed to other students in the group-centered class, whereas only 10% of student remarks were directed to other students in the leader-centered class. Also, the average level of affect was greater in the group-centered than in the leader-centered class.
The findings, however, suffer methodological error. To illustrate, Bovard mentioned in passing that students in the control and contrast classes knew about each other. One assumption may be made that students may have had the opportunity to converse about the class and its professor. Consequently, discovering that the professor was changing behavior from one class to another may have led the students to deduce that a prescribed role set was being fostered in the classroom.
This could influence the students (consciously or unconsciously) to act unnaturally or play up to the prescribed role, resulting in the guinea pig effect (see Isaac & Michael, 1978. pp. 62—63).
In spite of the above error, the implicit theme of Bovard’s study was to determine if Negro student veterans, Jewish and Catholic students, and students from a wide range of socioeconomic backgrounds would develop a cohesive group by allowing verbal interaction among them to occur. Bovard found this to be true. Bovard (195 I) assumed “from the present evidence that the amount of social interaction in the classroom will influence the individual student’s perception, feelings, and interpersonal relations, and perhaps even (the student’s) personality development” (p. 223). Moreover, because of the racial, denominational, and socioeconomic implications of this study and regardless of its methodological error, Bovard’s study is a pinnacle study which investigates classroom climate and the qualitative effects it may have on its student members.
Buswell’s (1953) study also had an indirect influence on classroom climates. Her study is important to this review because it considered social relationships in the classroom. Her purpose was to determine whether or not those students who were accepted by their peers differed in certain achievements from those who were rejected. She hypothesized that no relationship between the social structure of a classroom group and the students’ achievement in some of the basic elementary school subjects would exist.
Buswell (1953) studied a group of 286 kindergarten children and a group of 321 fifth grade children. Several standardized tests (Stanford-Binet, Iowa Test of Basic Skills, Ohio Social Acceptance Scale) and a sociometric test to determine the best liked peers were used. She found that the highly accepted group (according to the sociometric test) was significantly higher than the rejected group in mean achievements (i.e., the standardized test). Her primary conclusion “is that when we consider a classroom of boys and girls in either the lower or upper grades, it may be said that in general those who are succeeding in their school work will also be succeeding in their social relationships with their peers” (p. 51). Though Buswell’s work does not discuss classroom climates per Se, it does consider and discuss the students’ social interactions in the classroom and the resulting social system that may form a classroom climate (Getzels, 1969; Withall & Lewis, 1963).