The corporate reformers are watching their grandiose plans fall apart, and I suspect they are getting worried; their deep-pocketed backers wanted to see results by the end of the year, and it increasingly looks like that simply isn't going to happen. So they're shifting the goalposts and toning down their rhetoric. The preferred outcomes are now more modest, and the new tone is reasonable and conciliatory.
Witness today's op-ed from B4K's Derrell Bradford, featuring his new, wonky sales pitch:
The state Department of Education currently is piloting an overhaul of teacher evaluation in New Jersey consistent with the guidance of President Barack Obama and U.S. Secretary of Education Arne Duncan. The pilot makes student learning, demonstrated in a variety of objective manners, a key factor in how teachers are evaluated. My organization, B4K, supports the pilot program and using student learning as a significant factor to determine if teachers are awarded tenure.As I pointed out at the time, this "pilot" didn't even allow time for a real analysis of its outcomes before the system was imposed by the state on districts next year. ACTING Commissioner Cerf seems to have backed off on the timetable, because there's still plenty of evidence the state is no where near ready to implement a system that tracks this huge set of data. Maybe if they had started out with an advisory group that included actual teachers, instead of unqualified lobbyists like Bradford, they would have come up with something that actually had a chance of working.
The Department of Education’s evaluation pilot program makes objective measures of student performance constitute 50 percent of a teacher’s evaluation, with the remaining 50 percent being grounded in more traditional teacher observation against a high-quality rubric approved by the department. More importantly, both halves of this framework are flexible, with the objective 50 percent allowing districts to make as little as 35 percent of a teacher’s total evaluation linked to statewide standardized assessments (the rest can be district-created assessments or school-based measures).Oh, please: do you really think cash-strapped districts are going to spend money - let alone instructional time - on assessments other than the state tests they have to give anyway? And I'd like to know if Bradford agrees with what's happening down in Tennessee, where PE teachers pick which school-wide results on state tests (like writing or math) they want included in their evaluations. Because that's exactly what could - and probably will - happen under his proposal.
Moreover, those measures assess student achievement based on growth (how much did students learn over time), not as a single data point. The remaining, subjective 50 percent allows for competition among evaluation rubrics, as districts work to find the right fit. It also changes the rating system to four categories from two.Which means students taking MORE standardized tests, as the teachers can only be held to account for what happens in their classrooms during the school year - not the year before and not during the summer. I'll bet the testing industry is licking their chops at that. I can see it now: teachers and principals deliberately trying to get kids to score as low as possible in the fall the exaggerate gains in the spring.
Bradford's use of the word "growth," by the way, is quite deliberate: the NJDOE is pushing for Student Growth Percentages (SGPs) to be used in teacher evaluations. So there will be no effort made to try to correct for student characteristics and home environment, even though that has a far larger effect on test scores than anything that happens in school. And no accounting for the fact that students are not assigned randomly to teachers, so principals can reward favored employees and punish others. And no accounting for the biased and unreliable nature of the tests themselves. Or the fact that no matter the percentage of the evaluation, an inflexible number becomes all of the actual decision.
Hey, but no worries! Turns out these tests tell us the same thing as observations!
The University of Chicago, in its study “Rethinking Teacher Evaluation in Chicago,” found there “is a strong relationship between classroom observation ratings and value added measures, and the relationship holds for math and reading test scores. In the classrooms of highly rated teachers, students showed the most growth; in the classrooms of teachers with low observation ratings, students showed the least growth.”So, the tests confirm what the observations already tell you. Then why have the tests at all? And why try to classify teachers into several levels based on those tests, when the tests - which Bradford himself admits have high error rates - are clearly incapable of that level of precision?
Bradford also ventures into tenure reform in his piece, but curiously leaves out B4K's previous demands that tenure appeals be contained within the system. Maybe he finally understands that taxpayers do not want their schools turned into a bunch of little patronage machines, and that a tenure hearing without an appeal outside the district isn't tenure at all.
And he's pretty much given up on merit pay, which doesn't work. He's going to take established programs like National Board Certification, tweak them a little, and try to claim them as his own. Whatever.
But notice what's missing here: no charter schools. No vouchers. No NJEA bashing.
Bradford is making a last stand on incorporating test scores into teacher evaluations, a practice that is already the status quo in much of the country. Where's the bold new thinking? Where's the courage to stand up for the children? What we have here are the first pieces of evidence of retreat. They made their case, and it has been found wanting.
This isn't over. But the momentum is shifting.