by Rob Levine
Mar 18, 2015, 10:10 AM

Performing feats of legerdemain

There’s a big difference between a teacher’s rating on a flawed and invalid evaluation system and her “performance.”  Just don’t try and tell that to local media.

Three years ago the Washington, DC based education reform lobbying outfit called 50CAN, along with its wholly-owned Minnesota subsidiary called MinnCAN, came up with an ingenious new plan to pressurize and weaken teachers’ unions: They began a project to brand their movement’s attacks on public school teachers as the simple proposition of evaluating teachers on their “performance,” versus current practices of evaluating teachers by their seniority, educational achievement, and classroom observations.

The attack wasn’t grounded in research or logic; instead, the organizations used public polling on the subject, purporting to measure public sentiment on teacher retention decisions by pitting teacher seniority versus “performance.” It was a clever, if dishonest tactic, since polling can obscure on the ground realities. Just because the public believes something to be true doesn’t make it true.

Thus if a well-funded organization with deep political connections is actively trying to deceive the public on matters of public import a deceptive poll can be a very effective tool for changing public policy. The real trick of the matter is to convince media and opinion leaders to accept the framing of the issue, no matter how illogical and counter-intuitive the premise.

Normally one would expect those crucial audiences to see through such blatant attempts at propaganda. After all, journalists are social scientists who are supposed to be trained to recognize attempts at public opinion manipulation. Unfortunately media organizations of today have been denuded of experienced reporters, and the money thrown around by movements such as that of the education reformers have turned organizations and people who in other contexts might have been counted on to shed light on sophisticated propaganda campaigns into grifters – people whose careers are predicated on not seeing through the data-induced fog.

But today the fabulists at 50CAN must be giving themselves high-fives – and deservedly so – as media organizations, politicians and on-the-payroll activists have shamefully taken up the performance vs. seniority frame and uncritically run with it.

So what, exactly, is wrong with the formulation of performance vs. seniority as a polling question? For starters, the two answers are not mutually exclusive. One can quite obviously be for both propositions. In most job environments in the US, in fact, seniority is equated with performance, and for good reason. 

But there’s a deeper problem with the formulation: The question of whether or not to rate teachers by their “performance” is nearly tautological. That is, who in the world would not want employees rated by their “performance?” It’s kind of astonishing that MinnCAN could find anyone who thought rating teachers by performance wouldn’t be a good thing.

The real question is how the word “performance” is defined, and just whose “performance” is being rated – since there are no primary objective measures. In the past seniority, educational attainment, and classroom observations served as a proxy for performance. Despite all the huffing and puffing by education reformers over the supposed weakness of the relationship of those factors to teacher performance, there is no reliable proof that their proposals to replace them with student test scores on standardized tests are scientifically valid.

In any event, strictly speaking, proposed changes to teacher retention policy in the Minnesota legislature don’t tie retention to teacher “performance.” No, they tie it to a proposed new system of teachers’ ratings built on the new state teacher evaluation system. That’s a big difference.

The teacher evaluation law, passed in the 2011 legislative session and implemented in 2014, mandated that 35% of a teacher’s evaluation be based, essentially, on student test scores. For a movement that is all about data this is a serious bout of cognitive dissonance, as the best research on the subject shows that, at most, differences in teachers account for about 15% of the variance on student test scores. Teacher impact is in fact dwarfed by factors outside the schools, such as family environment and economic circumstances. The biggest factor limiting educational achievement is poverty.

Just last month more than 800 education researchers stuck a pin in the nation’s testing craze, urging federal lawmakers to abandon the blizzard of tests unleashed by the misnamed No Child Left Behind law, writing that they “…strongly urge departing from test-focused reforms that not only have been discredited for high-stakes decisions, but also have shown to widen, not close, gaps and inequities.”

There’s even more wrong with evaluating teachers by student test scores. Standardized testing in Minnesota and around the nation covers only three subjects: Math, English, and science. Just how are we to evaluate the scores of teachers who practice in subjects not covered by testing?  Some districts let teachers choose which non-taught subject they will be evaluated on. Many use so-called Value Added Measures (VAM), a highly disputed and discredited method. One social science teacher in Minneapolis recently received her VAM score, which was based on student test results in math. Amazingly, the cohort of math students who comprised her VAM rating was just one student!

And there’s yet another troublesome aspect of our new teacher evaluation system. The law mandates that a portion of a teacher’s evaluation be done by superior or peer observations. Since the legislature did not adequately fund the new requirements, districts are taking teachers out of their classrooms for the year to do peer evaluations, resulting in raising class sizes around the state. Whatever one might think about peer evaluations, social science research has shown unequivocally that lower class sizes is one of the sure fire ways to increase academic achievement.

Given the above, no one can argue with a straight face that Minnesota’s teacher evaluation system is any way a valid way to rank teachers, let alone that its implementation will lead to increased academic performance or educational equity. Worse, local media have equated teacher evaluations to individual teacher “performance” and “merit,” which is absurd and misleading. An exodus of experienced teachers with advanced degrees is underway in Minnesota. One has to wonder whether that was, in fact, the point of this whole evaluation exercise.

When the Minnesota House passed legislation last week tying teacher evaluations to retention, the Star Tribune misleadingly and erroneously headlined its story “House votes 70-63 to require school districts to consider performance, not just seniority.” Apart from issues of whether the state’s evaluation system actually measures “performance,” the headline is just plain wrong. The legislation effectively ties ratings on the evaluation system, not performance, to teacher retention.

A few days later the Star Tribune editorial page weighed in to reinforce the same mistake, headlining its editorial “Schools should include performance in teacher layoff decisions.” As you can see, the paper, in both its news and editorial departments, has bought the MinnCAN propaganda formulation positing that teacher “performance” is oppositional to rating teachers by seniority, educational attainment, and observational reviews. This conflation of “performance” with evaluations on an invalid evaluation system cannot be accidental, as education policy debates pervade the state’s discourse.

If the Star Tribune really cared about education in Minnesota it would apply some of its investigative weight into the farce that is the state’s teacher evaluation law before it’s allowed to further damage education in the state. Barring that, it would be nice if it stopped equating teacher “performance” or “merit” with evaluations on a convoluted and invalid teacher rating system. But given past “performance,” I’m not holding my breath.

*  *  *  *

I did try to get the Star Tribune to change its reporting on this topic.  I contacted the writer of the story linked to above, but he was quite insistent that his characterization of  “performance” was correct.  And to drive the point home used the exact same wording a few days later.  I then contacted Executive Editor Rene Sanchez with my complaint, but he never responded.  I then wrote this piece as an op-ed and submitted it, but it too was rejected. The newspaper is not eager to take its thumb off the scale.


Thanks for your feedback. If we like what you have to say, it may appear in a future post of reader reactions.