Twelve years ago, George Bush signed “No Child Left Behind” into law. Among other things, the law requires that by the end of this school year every student in America has to meet standard. That level of success will never happen, of course, not even in Finland, but no one has bothered to change that part of the law. Instead, the Obama Administration has used that law as leverage to advance their own educational agenda, which includes expanded school choice, adoption of the Common Core State Standards and tougher teacher evaluation laws. They’ve done this by granting waivers from the law's punitive aspects to states that adopt certain policies.
Washington State received one of those waivers, along with 31 other states. And for the most part, we’ve toed the line. We now allow charter schools, we’re transitioning to the CCSS, and we have a brand-new Teacher and Principal Evaluation Project. (TPEP)
But there’s a problem. As written, TPEP allows state assessment scores to be used for teacher evaluation. The feds want TPEP to require that they be used. The feds have recently notified our state, warning us that we risk losing our waiver unless TPEP is changed so that it mandates the use of state assessment data.
As a teacher, I can see no possible way in which state test scores can be used as a valid basis for my evaluation. I teach fourth grade; my students took a state test last year and they’ll take another one this year. But it’s not the same test. Last year they took a third grade test and this year they’ll take a fourth grade test. The smart kids in my class passed their test last year and they’ll probably pass their test this year. The kids who are struggling this year didn’t pass their test last year and they’ll have a tough time passing this year’s test.
That’s just common sense. But how is that supposed to translate into my evaluation? Help me with this. Am I supposed to get credit for having more kids pass this year than passed last year? That doesn’t work; this is a new class, with different kids. Am I supposed to get credit for having a higher “average score” for my class on their fourth grade test than their average was for their third grade test? That doesn’t work either. Again, it’s a different – supposedly harder – test. You cannot use a third grade summative test as a proxy for a fourth grade pretest. It isn’t statistically valid, since the two tests cover different skills and concepts, and they're normed to different populations.
It’s not that I’m against using student tests as a part of my evaluation; I just don’t see how state test scores can be used. We give a reading and math test at the beginning, middle and end of the year. These work perfectly for my evaluation. If I can’t get my students to improve on those two metrics, then I should be doing something else.
I think we all agree on that. At least those of us who teach or can understand logic agree on that. The feds, however, don’t teach and apparently don’t understand logic. They insist that we use state test scores for teacher evaluations. And they have the power to make that happen. They’ve made it clear that if we don’t make the change, we’ll lose the waiver.
It’s not entirely clear, however, what would happen should we lose that waiver. According to Senators McAuliffe and Rolfes, we could lose federal funding, or at least have to label nearly every school in the state as “failing.” According to the WEA, we wouldn’t. Personally, I don’t want to find out what would happen.
So now there’s a bill heading toward the state senate with new language that should save the waiver. It reads: “… for teachers who teach reading or language arts or mathematics in a grade in which the federally mandated statewide student assessments are administered, one of the multiple measures of student growth must be the student results on the relevant assessment.” Did you notice the word must? The only real difference between this bill and the bill passed last year is that one word. This bill replaced the word can with the word must.
It's only one word, but it's a big, big deal. We finally have a teacher evaluation model that everyone seems to support. Teachers like me are getting real, constructive feedback. I've discussed with my principal which student assessment data would be appropriate in regards to my evaluation, we've settled on a plan that should work and we're currently implementing that plan. As far as I'm concerned, TPEP looks great. Why on Earth should we abandon a great model for something that relies on statistically invalid data?
There are teachers who are willing to die on this hill. They're willing to tell the feds what they can do with their waiver and carry on with what we know is right. And although I do feel strongly about this, I’m not one of them. First of all, the stakes are too high. I do not want to send a note home to all my parents telling them my school is a failure. To me, that inaccuracy is far more egregious and ultimately more damaging than using a state test on my evaluation. Nor do I want to risk losing federal funding. I teach in a Title 1 school. Losing those funds would be devastating.
Besides that, I doubt there’s much support outside the teaching community for noncompliance. In fact, I think I would have a hard time holding a non-teacher’s attention long enough to explain the issue, much less getting that person to join some sort of movement to keep the use of state tests as an option and not a mandate. If you doubt me, it's worth noting that the primary sponsor of this bill is Rosemary McAuliffe, who has been WEA's strongest ally for two decades.
So I'm not in favor of standing up to the feds. But here’s what I do propose: Go ahead and change the word can to must, but then add five new words that effectively cancel out that very change.
We could simply take this sentence from the proposed bill: “… for teachers who teach reading or language arts or mathematics in a grade in which the federally mandated statewide student assessments are administered, one of the multiple measures of student growth must be the student results on the relevant assessment.”
And change it to this: “…for teachers who teach reading or language arts or mathematics in a grade in which the federally mandated statewide student assessments are administered, one of the multiple measures of student growth must be statistically valid data based on student results on the relevant assessment.”
The genius (If I do say so myself) in this wording is that no one – not even the feds – can argue with using “statistically valid data.” How could they? Yet you and I both know that there really is no statistically valid data based on state test scores that would be of any use in evaluating a teacher. If there was, we'd already be using it. Therefore, the issue would be effectively moot and we could carry on as before, with our waiver secure, using data that makes sense, as part of a functional and constructive teacher evaluation system. In other words, we'd be constructing a loophole that would have to satisfy the feds, yet would allow us keep using our current model.
So there it is. A loophole big enough to drive a waiver through. Let’s build it!