Study looks ‘under the hood’ of new teacher-evaluation systems

More and more states are adopting new teacher-evaluation systems in response to a growing consensus that improved teacher quality can spell improved student achievement. The idea is that measuring how teachers perform in the classroom will help schools take the first steps toward helping them get better. But so far, there’s little consensus on the best ways to make those measurements.

Many states are moving quickly to launch new systems—in some cases, without much thought or work to ensure that the methods are sound. One teacher evaluation expert, Charlotte Danielson, has warned of a wave of lawsuits in places that don’t proceed with extreme caution.

A new report out today looks at the early adopters of new teacher-evaluation systems—and how they differ—as a way to help states and districts consider both innovations and potential missteps. The report, “Measuring Teacher Effectiveness: A Look ‘Under the Hood’ of Teacher Evaluation in 10 Sites,” was commissioned by ConnCAN, 50CAN and Public Impact, all education advocacy groups that support overhauling how teachers have been evaluated historically.

While most states and districts have agreed to use multiple ways to rate teachers—including, usually, a combination of classroom observations and student test-scores—from there they often diverge.

The use of standardized test-scores to evaluate teachers has received a great deal of attention, but one of the bigger challenges that states and districts face is how to rate teachers whose students don’t take such tests. In most school systems, the vast majority of teachers—up to 88 percent, for instance, in the District of Columbia—do not teach subjects or grade-levels covered by standardized tests. The study found that the early adopters have come up with different ways to measure how the students of these teachers are progressing. In some instances, districts are simply adding more standardized tests. Elsewhere, teachers will be graded based on portfolios or teacher-created assessments.

In several cases, more than one measure looking at how a teacher has affected students is being taken into account. “The rationale is that no single measure is perfect, but combining multiple measures diminishes the weaknesses of any particular measure,” the report’s authors say.

Daniela Doyle and Jiye Grace Han of Public Impact, who co-wrote the report, also found differences in how student “growth” on standardized tests is measured; some of the more esoteric details of the calculations have been hotly debated among researchers.

When it comes to classroom observations, “there was a surprising amount of consensus,” according to Doyle and Han. But leaders of the various sites studied in their report made different decisions about who conducts the observations and how often they occur. (In our own reporting on efforts to overhaul teacher evaluations, we have found nuances among different systems that do seem to matter.)

“Measuring Teacher Effectiveness” doesn’t endorse any particular method: “None of these systems claims to have cracked the code for teacher evaluation,” the authors write. Instead, they focus on in-depth comparisons. What might seem like so many insignificant details to outside observers will loom large for the teachers and students they’re meant to help.

Comments & Trackbacks (7) | Post a Comment

Lisa Guernsey

Sarah, thanks for the information about this new study, which will be important for early education policy makers. The link to the report is broken — can you repost? Thanks.

Janet Atkins

The difficulty in teacher evaluation arises from the fact that teaching is an art. It is NOT strictly a science with a method that if followed correctly will make each and every student an A+ learner and successful in all of life. That’s the panacea that so many Americans desire, and that’s why we have so many packaged curriculums on the market that promise to make successful learners in each and every classroom. That’s why we test our children to their wit’s ends. Until the American public and our politicians realize that teaching is an art, and that art cannot be quantitatively evaluated—but rather must be evaluated qualitatively—then we will miss the boat on teacher evaluation. Good teachers know good teaching. If you check with the majority of administrators and district personnel—those who do evaluations—you’ll find many of them left the classroom because they just couldn’t stand teaching—or at best were mediocre in the classroom. Those of us who have stayed, and succeeded for many, many years to turn kids on to learning and to foster relationships that have made success story after success story, know that we practice an art. One simply cannot measure art.

robert e. soctt

One area worth considering when it comes to quality teaching practices would be long term studies, much like the framingham medical study.
Think it, if three parents show up at their local public school for the first day of first grade for their children and the principal tells parents their children are goiing to be a part of a study. Teachers have received extra training during the summer, we will be comparing our traditional program to a language based intensive program and one with a greater emphasis on the more affective aspects of learning.

Many parents would be very upset. Most will not their child being used like a lab rat. Therefore, there is little research that clarifies excellent and vital teaching practices you would see are more universal in people considered good teachers.

To create such research would be a major undertaking but one I think would be very instructive.

George Peterman

Keep in mind that ConnCAN was just given a national Bunkum Award (“If Bernie Madoff Worked in School Finance Award”) from The National Education Policy Center (NEPC), housed at the University of Colorado Boulder, for their gross misuse of data in their propaganda efforts to kill unions and divert public money to their privately run schools. ConnCAN is the lobbying group begun by the same folks who run Achievement First, a private educational company, founded with the help of hedge fund managers, wealthy pharmaceutical directors (a few of who have been fined majorally large sums for bad practices with vicodin), and the current Connecticut ed commissioner Stephan Pryor.

Patrick Crabtree
Sarah Garland

Interesting points. Charlotte Danielson’s teacher observation model is being used in a lot of places because it has some research backing backing its ideas about what teaching practices are best. (It’s also interesting to note in the context of the new evaluations that Danielson doesn’t support the use of student test scores to rate teachers.) So is it an art, or is there a way to do experimental research to find out which teaching methods help boost student success, and which don’t? Can we assume these methods would be universal, or how much does it depend on the classroom and school context?

Your email is never published nor shared.