Join the conversation

Value-added modeling is back in the news (although did it ever really go away?). Among its appearances this week, Stephen Sawchuk reported that Representatives Jared Polis (D-Colorado) and Susan Davis (D-California) recently introduced a bill that would require states to oversee new systems for evaluating teachers and principals. It would also require that student achievement growth as measured by state or local assessments—or value-added analysis where available—be the predominant factor in teacher evaluations.

While Sawchuk points out that this bill will likely not advance on its own, he says to watch it. Given both Polis and Davis are on the committee that will take the lead in shaping the House’s revision of ESEA, it could be an indicator of what is to come.

Language like that of this bill frustrates me to no end. While most, if not all, education stakeholders agree that current methods of teacher evaluation are flawed, why does it always come back to value-added models? After reviewing the research, which clearly shows the limitations of this strategy, how can one argue that more than 50% of a teacher’s evaluation be based on it?

And please do not say, “It’s better than doing nothing.” Over at School Finance 101 Bruce Baker shares an analogy in response to that argument:

If we were in a society that still walked pretty much everywhere, and some tech genius invented a new cool thing – called the automobile – but the automobile would burst into a superheated fireball on every fifth start, I think I’d keep walking until they worked out that little kink. If they never worked out that little kink, I’d probably still be walking.

He had previously written about how this rate relates to likely the errors in teacher dismissal (dismissing effective teachers as ineffective) that would occur when using typical value-added modeling approaches. (That post was based on the recent National Center for Education Statistics report Error Rates in Measuring Teacher and School Performance Based on Student Test Score Gains.)

And that piece only really looked at the statistical problems with value-added assessments. It did not consider whether the tests teacher evaluations would be based on measure student knowledge of things we actually care about in a valid way. Roxanna Elden’s piece in the latest Education Next gives me a reason to question that. It includes an experience with a practice FCAT reading question:

Which of the owls’ names is the most misleading?

[Elden] was stuck between (F) the screech owl, because its call rarely approximates a screech, and (I) the long-eared owl, because its real ears are behind its eyes and covered by feathers. The passage explains that owls hear through holes behind their eyes, so the term long-eared owl seemed misleading. Then again, a screech owl that rarely screeches? That is pretty misleading, too.

I personally don't want 50% of a teacher’s evaluation based on whether s/he was able to prepare students to answer a test full of questions like that.

But acknowledging problems with value-added modeling doesn’t mean we should accept the status quo. Instead, as Baker suggests, we could look beyond the false dichotomy (or “really stupid argument”) that there are only two choices in teacher evaluations: status quo or value-added. We could also consider peer reviews of teachers. We can consider other examples of student achievement, such as portfolios or writing samples. Et cetera. We owe it to our teachers—and ultimately, to our students—to ensure that those in the classroom are the ones who really should be.

Image from Maximaximax