Teacher Effects: Past, Present, and Future

• In Event: Assessing Teacher Quality: Understanding Teacher Effects on Instruction and Achievement

Abstract

The effects of teachers on student achievement have been studied extensively the last 40 years. The findings in the teacher effects literature have consistently indicated that teachers differ considerably in their effectiveness to promote student achievement. By and large the evidence from experimental or observational data has suggested that there are substantial differences among teachers in the ability to produce achievement gains in their students. However, the teacher characteristics that are typically observed such as education, experience, licensure, or salary, have not been consistently related to student achievement and, as a result, the findings in the literature are mixed and inconclusive. In addition, the observed teacher characteristics do not seem to explain much of the variance in teacher effectiveness. For example, findings using Project STAR data suggest that teacher education and experience explain only about one percent of teacher effects (measured as value-added variation in classroom achievement). It appears that the most important teacher characteristics remain unobserved perhaps because they are difficult to measure reliably.
In order to become more successful in estimating teacher effects, both in terms of long-term research studies of teacher effects as well as applied programs and policies related to teacher effectiveness, several things need to be achieved in the near future. First, it is important to identify the factors that are likely to be accountable for most of the variation in teacher effectiveness. This can be accomplished by systematically observing classrooms and by interviewing teachers that are identified as effective as well as their colleagues. For example, teachers can speak to the desirable aspects of their instruction and what they think makes a difference in the classroom. The idea is to construct a synthetic profile of an effective teacher and then try to measure the characteristics of an effective teacher as reliably as possible. It is also important to identify effective instructional practices that take place in the classroom and measure them. This process will include defining explicitly what teacher effectiveness is, constructing reliable measures of the factors that constitute teacher effectiveness, and finally collecting high quality classroom data that will allow researchers to examine empirically what factors promote student achievement the most.
Second, more careful thinking has to be incorporated in designing studies to examine teacher effects. These designs could be carefully planned experiments that will eliminate pre-existing differences in students and teachers, or well-designed quasi-experiments or observational studies. Matching techniques could also be incorporated in these designs to facilitate the creation of similar classrooms and avoid confounding effects.
Third, data analysis is also an important part of teacher effects research. To estimate teacher effects quantitatively careful analysis is necessary. Although there have been advances in value-added methodology the last 10 years, more work is needed especially for longitudinal designs. In addition, analysts should be more careful when specifying their models and should strive to reduce bias in teacher effects estimates as much as possible. More rigorous analysis will result in more valid estimates and findings that will inform research as well as policy-making.

Author

• Spyros Konstantopoulos, Michigan State University