More Second Thoughts on Using Test Scores to Evaluate Teachers

In this important article in The Education Gadfly, Michael Petrilli, a leading conservative commentator, reviews these commonly-voiced concerns about test-score accountability for teachers:

• Are test scores valid measures of teacher performance, and what happens with teachers of non-tested subjects?
• Will tying students’ test scores to teacher evaluation warp instruction and curriculum by spurring teaching to the test?
• Can we “principal-proof” schools by using one-size-fits-all evaluation models?
• Is it wise to introduce high-stakes accountability at the same time that new Common Core curriculum standards are being introduced?

But Petrilli’s biggest worry is that the two arguments put forward by proponents of test-score accountability for teachers don’t stand up to critical scrutiny:

• Test data will make it possible to identify and remove the least-effective teachers. Petrilli doubts that principals will be willing to give less-than-effective teachers low marks, and even if they do, he believes due-process protections will make it next to impossible to fire tenured teachers. “If you’re a school principal,” he says, “why give a teacher a bad rating if you know you still can’t remove her from the classroom? All you’ve done is create an enemy – or set yourself up for a lawsuit. Smart principals know better and will do what they’ve always done, which is to find a way to play the ‘dance of the lemons,’ sending their bad teachers to another school.” 

The only way change this dynamic, says Petrilli, is for states to make radical changes in employment law: “Either policymakers need to combine evaluation systems with reforms that make it plausible to fire ineffective employees, or they shouldn’t bother with high stakes at all.” Indiana, Colorado, and Tennessee are trying to make significant employment-law changes, but he doesn’t think “blue” states will rein in tenure for veteran teachers. Unless a state is willing to go all the way, he says, “There are significant costs in terms of dampening teacher morale, provoking a parent backlash, and over-encouraging teaching to the test. Maybe it’s worth it if we can identify and remove the worst teachers. But if that’s not going to happen, it’s a loss, not a victory.” 

• Test-score accountability will improve teaching by providing critical feedback. Petrilli is highly skeptical. If that’s the goal, he says, “then let’s just do that. Don’t even call it ‘evaluation.’ Don’t attach any stakes. Just provide the data to teachers and principals – and continue to train the latter on how to conduct high-quality teacher observations – and call it a day. The stakes are high – not just for teachers, but for the reform movement. As, I suspect, history will show.”

“All or Nothing on Teacher Accountability” by Michael Petrilli in The Education Gadfly, Aug. 29, 2013, http://bit.ly/1a0spdE 

From the Marshall Memo #500