The Valid Use of Student Performance Measures for Teacher Evaluation

Abstract

Student achievement test scores appear promising as indicators of teacher performance, but their use carries significant risks. Inappropriate tests improperly used may encourage undesirable shifts in curricular focus or poor teaching practices, and may unfairly favor teachers of more able classes. It is often said that standardized achievement test batteries are unsuitable for teacher evaluation, but few systematic alternatives have been suggested. The purposes of this paper are to analyze some problems in using student test scores to evaluate teachers and to propose an achievement-based model for teacher evaluation that is effective, affordable, fair, legally defensible, and politically acceptable. The system is designed only for detecting and documenting poor teacher performance; rewarding excellence in teaching is viewed as a separate problem, and is not addressed in this paper. In addition to pretesting and statistical adjustments for student aptitude differences, the proposed system relies upon attendance data and portfolios of student work to distinguish alternative explanations for poor test scores. While no single set of procedures can eliminate all errors, the proposed system, if carefully implemented, could expose teaching to constructive scrutiny, organize objective information about teaching adequacy, and help to guide its improvement.

Teacher performance evaluation in the southeastern states: Forms and functions

“…The public wants teachers to be evaluated on the basis of students' achievement scores, even though this issue is surrounded with difficulty (see Haertel, 1986). While some studies have shown positive correlation between teachers' evaluation ratings and student performance (Brandt, 1990;Holdzkom, Kuligowski, & Stacey, 1990), such evidence of linkages is largely absent or unexplored in the evaluation systems described here.…”

Fifty reasons why student achievement gain does not mean teacher effectiveness

“…These factors can account for a sizable proportion of the gain that may be exhibited in student achievement. Many of the factors have been identified previously by Berk (1984c), Haertel (1986), Medley et al (1984), andMillman (1981). In addition, several reviews of research on input-output analyses of schools by Bridge, Judd, and Moock (1979), Centra and Potter (1980), Cohn and Millman (1975), and, especially, Glasman and Biniaminov (1981) provide valuable insights into the impact of numerous variables on achievement.…”

“…Establishing such a plan on the basis of achievement test gains is fraught with difficulty. The difficulties stem primarily from limitations in the testing technology, from factors that influence a teacher's effectiveness beyond his or her control, and from the infeasibility of executing rigorous experimental design procedures in the natural school setting (see Haertel, 1986).…”

“…Expenditures (variable 9) for administration, instruction, and extracurricular activities were positively correlated with reading and composite achievement (Benson et al, 1965;Cohn & Millman, 1975;Kiesling, 1969Kiesling, , 1970. Research on variables 10 through 12 was examined by Haertel (1986). He concluded that (1) quality of instructional materials may influence achievement (Wiley & Harnischfeger, 1974); (2) teachers can be more effective in schools with favorable learning climates (Bridge et al, 1979;Brookover, Beady, Flood, Schweitzer, & Wisenbaker, 1979); and (3) instructional support at the elementary and secondary levels can affect student performance.…”

A commentary on the role of student achievement data in the evaluation of teachers

“…This interest often focuses on the use of norm-referenced standardized achievement test scores as the index of student learning and therefore of the quality of teacher performance. However, both Haertel (1986) and Berk (1988) have spelled out in precise detail why this use of standardized test scores for a summative teacher evaluation is indefensible from both assessment and evaluation perspectives. The reasons rest principally on (1) the insensitivity of the assessment instruments to the effects of particular teachers, and (2) the long list of factors beyond the control of the teacher that influences the scores.…”