The indefatigable Paul Thomas of Furman University wrote an opinion piece for a major newspaper in South Carolina, warning against the error of evaluating teachers by test scores.
This, he says, is a bad idea that won’t die. In other words, it is a zombie policy.
Thomas carefully reviews the research on this idea for a lay audience, explaining what they need to know about the inaccuracy and bias inherent in this method.
He draws on the excellent research synopsis of Edward Haertel, the distinguished psychometrician of Stanford University.
Thomas writes:
For example, Edward H. Haertel’s Reliability and validity of inferences about teachers based on student test scores (ETS, 2013) offers yet another analysis that details how value-added methods fail as a credible policy initiative.
Haertel refutes the popular and misguided perception that teacher quality is a primary influence on student test scores. As many researchers have detailed, teachers account for about 10 percent to 15 percent of student test scores. While teacher quality matters, access to experienced and certified teachers as well as addressing out-of-school factors dwarf narrow measurements of teacher quality.
He also concludes that standardized tests create a “bias against those teachers working with the lowest performing or the highest performing classes,” which makes it hard to justify using student test scores as anything more than a modest factor in teacher-evaluation systems.
Instead, Haertel calls for teacher evaluations grounded in three evidence-based “common features”:
“First, they attend to what teachers actually do — someone with training looks directly at classroom practice or at records of classroom practice such as teaching portfolios. Second, they are grounded in the substantial research literature, refined over decades of research, that specifies effective teaching practices…. Third, because sound teacher evaluation systems examine what teachers actually do in the light of best practices, they provide constructive feedback to enable improvement.”
Haertel concedes that value-added methods may have a “modest” place in teacher evaluation. That’s no ringing endorsement, and it certainly refutes the primary — and expensive — role that they play in proposals to reform teacher evaluation in South Carolina and across the country.
Haertel concludes that test scores should not be a set percentage of any teacher’s evaluation, but South Carolina wants it to be 50%. A huge mistake.
Diane Ravitch's blog 