NSBA’s Center for Public Education has released an analysis of the push to change teacher evaluation systems, especially the inclusion of statistical measures of teachers’ effect on student learning.
That report, “Building a Better Evaluation System,” finds that the current teacher evaluation system is lackingand most evaluation systems now fail to identify truly effective and ineffective teachers.
“Current evaluation systems identify less than 1 percent of teachers as unsatisfactory even though by all accounts there are many more teachers that fall into this category,” said Jim Hull, the Center’s senior research analyst and author of the report. “This also means that our very effective teachers are lumped in with average and below average teachers and not recognized for their exceptional performance.”
Many states are looking for ways to tie teacher performance to their pay and to use teacher evaluations as a factor in determining layoffs. And the number of states that are linking teacher evaluations to student performance data is growing: In 2005, 13 states were able to do so; currently, 35 states are able to do so and the number is expected to grow.
“The push to change teacher evaluation systems, and especially to include statistical measures of teachers’ effect on student learning, is here,” Hull said. “Clearly, this is a fast-moving train that will likely affect many, if not most, school districts eventually.”
The Center has prepared a summary to help school board members prepare:
1. The current system is lacking. Current evaluation systems fail to identify the true variation in teacher effectiveness by rating all but a few teachers as “satisfactory.” One study of teacher evaluation systems nationwide found that only 1 percent of teachers are evaluated as “unsatisfactory.” Other research proves that there’s huge variability among teachers, even within schools, but it’s hidden by inadequate evaluation tools. A recent quote by Randi Weingarten, head of the American Federation of Teachers, sums it up: “As important as evaluation is to assessing teacher performance, what passes for teacher evaluation in many districts frankly isn’t up to this important task.”
2. Improving teacher effectiveness can dramatically impact student learning. Research has shown that teachers have the single greatest impact on students’ performance, more than family background, socioeconomic status, or school. By improving teacher effectiveness, districts could improve student achievement and save money at the same time, because they would be able to identify ineffective teachers early and provide them with appropriate support, rather than having to replace struggling teachers who leave the profession because of a lack of assistance.
3. Value-added models have flaws, but are much better than the system we have now. The fairest way to identify strong teaching is through a system that looks at student gains. Value-added models, which work to isolate the impact a teacher has on his or her students’ achievement from other factors, are the latest refinement of such a system. However, value-added models have come under intense scrutiny and criticism because the scores can be imprecise, and this criticism needs to be considered. However, while imprecision is a concern, the variation in scores should be considered against the current evaluation system, which almost certainly misidentifies many ineffective teachers as “satisfactory.” As long as they are used in concert with other methods of evaluation, value-added scores provide a useful insight into teachers’ impact.
4. Statistical measures are used to evaluate people in other industries effectively. Using imprecise statistical measures in evaluations is a generally accepted practice in fields outside of teaching. Major League Baseball, for instance, bases its million-dollar salary decisions largely on a player’s statistics, which can vary from year to year about as much as teachers’ do in value-added models. Other professions evaluated on similarly imprecise year-to-year measures include realtors; investors’ rate of return; utility company repairmen; and others. Value-added models should not be compared to a criterion of perfection, but whether including value-added models as part of a comprehensive teacher evaluation system would be an improvement over what is in place now.
5. There are ways to improve value-added models. The more years of data are used, the more precise value-added models become. For instance, the chance of misidentification drops by 10 percent when three years of data are used instead of one. Better state assessments, and aligning the assessments to what is taught, could also improve value-added models.
6. Multiple measures are the way to go. Virtually all researchers advocate using value-added data as one of multiple measures when making decisions about teachers. Using traditional measures, such as classroom observation, along with value-added data will present a fuller, more accurate picture of a teacher’s true effectiveness. In current formulas that use value-added models, the value-added score generally accounts for 25 to 50 percent of the total rating. Which measures to use and how much weight to put on each are decisions best made locally based on data, resources available, and the district’s goals for the teacher evaluation system.