Accountability: Assessment Problems and Possibilities

Abstract

Logically, there appear to be three major strategies contending for a role in the evaluation of teaching skills. The traditional and most widely used strategy to date has been an assessment of the quality of the program within which the teacher was trained (i). The aspects of that strategy have led to the movement for competency-based teacher education (CBTE). As a part of traditional evaluation in teacher preparation, measurement of the teacher's knowledge continues to be relevant.Two other strategies appear to be viable ones within the broad context of evaluation of teacher competence: measuring the growth of pupils taught by the teacher and measuring the teaching behavior of the teacher. Measurement of Pupil GrowthThis is an assessment strategy which is immediately appealing to many. Probably there are a number of reasons for this. Since the business of schools is to produce change in pupils, it seems reasonable to assess the success of the school by measuring the growth of pupils. In some instances, businesses pay workers in terms of production; why not pay teachers on the same basis? Such a solution is immediate and compelling, but examination of this possibility raises questions.The influence of the Classroom A major difficulty in evaluating the teacher is the amount of influence the classroom can have in relation to other influences on the pupil. A series of papers published by the Office of Education (2) concluded that the relative influence of the teacher or the school is not great. A documented example of a specific nonschool affect, the relations between attitudes and expectations of parents to intelligence and achievement of their children have been found to be strong. The relations hold even within a single socioeconomic group and have been demonstrated in a number of ethnic groups (3; 4; 5). Similarly, the peer group influence has been demonstrated.Presumably these are only a few effective non-school influences. If the teacher is only one of a number of influences on pupil growth, the correlation of growth for one pupil group with another the following year should not be high. This turns out to be the case. One study (6) showed a correlation of .08 for successive years of pupil growth in pooled achievement measures for a group of 55 teachers. Rosenshine (7) has summarized a series of studies indicating relations typically in the .30's for growth for successive years. Brophy (8) has reported successive year data which are highly variable, with correlations ranging from low negative to high positive, but with a median in the .3o's. As test-retest reliabilities, correlations like these would not be acceptable.To lay the pupil's growth, or lack of it, at the teacher's door, seems a major oversimplification considering the many other factors involved. Measurement-Statistical ProblemsThe solution of measuring pupil growth looks so simple-yet involves a series of problems. Specialists in educational and psychological measurement have labored with the difficulties for a generation or more, without final res...

Critical Elements of Classroom and Small‐Group Instruction Promote Reading Success in All Children

“…During the 1980s, effectiveness was defined in terms of interactions of classroom processes with student outcomes (see reviews by Brophy & Good, 1984;Rosenshine & Stevens, 1986). The strongest correlates were academic-engaged time, classroom management, and certain patterns of teacher-student interactions (Soar, 1973;Stallings, Robbins, & Presbrey, 1986). For disadvantaged students, the link between explicit instruction and achievement was notable, a finding supported in other classroom observational research (Brophy & Evertson, 1978;Good & Grouws, 1975).…”

Sources of invalidity in measuring classroom behavior

“…Unfortunately, the adjustment also depends on the characteristics of the pupils being tested, and this information is generally unavailable. 8 While residual gain scores and analysis of covariance are repeatedly discussed in the literature (Rosenshine, 1971;Soar, 1973) as "parallel" techniques , they, in fact, are not. These different computational procedures are not mathematically equivalent and, therefore, can, in any given research effort, lead to quite different results, introducing the distinct possibility that the researcher may reject the null hypothesis with one technique and fail to reject it with the other.…”

Professional competencies of home economics teachers in the middle school

“…The first approach involves the identification of teacher competencies which appear related to pupil performance (Soar, 1977) and the second approach involves expert judgment to identify teacher behaviors which presumably make a difference in pupil performance.…”

“…One of these questions is the unreliability of differences between pupil pre-and post-test achievement. Soar (1977) reports a nonlinear relationship between pupils initial score and the gain they show as low scorers gain little, moderate scorers gain greatly, and high scorers gain little. Other factors restricting the measurement of pupil gain are changes which take more than one school year to achieve in students, unsuitability of standardized tests for measuring all desired objectives, and the unstableness of teacher effects on pupils over time (Borich & Fenton, 1977 as determined by observations, show that teachers vary more from class to class than from teacher to teacher (Caputo, 1975;Shavelson & DempseyAtwood, 1976).…”

“…It is suggested that what the teacher does in a classroom is more pertinent to competence than pupil growth (Soar, 1977).…”