NY Times on teacher ratings: "Margin of error" is "35 percentiles in math and 53 in English, the city said." Garbage in, garbage out.
And Yoav Gonen from the Post suggests it's even worse:
Maximum margin of error for a teacher's percentile ranking on#TDRs is 75 in math and 87 in ELA
In either case, as Ravitch pointed out, your odds are worse than a coin toss. That is some pretty alarming data about the VAM junk science that has landed teachers on pages with labels like "City's Worst Teachers." A problem with VAM is not only that it is junk science, but that no form of it has ever been established not to be. Arguments that a new system will contain less junk than this one are impossible to verify, and preposterous in any case since there is no planned pilot period. Even more ridiculous, as revealed by Leo Casey on Edwize, they anticipate system will improve, though they have no basis whatsoever for that assumption, and wish to raise 20 percentage points of junk to 25. (This assumes locally negotiated 20, or consequent 15% will not consist of more junk, an assumption I'm not yet willing to make.)
Casey claims only extremely low scores will jeopardize teacher jobs, but Michael Winerip shows it's entirely possible for good teachers to get very low scores:
According to the formula, Ms. Isaacson ranks in the 7th percentile among her teaching peers — meaning 93 per cent are better...What you would think this means is that Ms. Isaacson’s students averaged 3.57 on the test the year before; they were predicted to average 3.69 this year; they actually averaged 3.63, giving her a value added of 0.06 below zero.
A new system may indeed be better. Or it may be worse. Or it may be better sometimes and worse other times. These are the drawbacks of using junk rather than science to make conclusions, and the fact is, under the proposed system, teacher jobs can and will depend on that.
Let me return to FOIL. This is the ability of people or entities to request information, just as the papers did so they could release the crap data reports that have humiliated so many working teachers for no good reason. While your S or U could not be FOILed, numerical data can. VAM formulas and proposed 1-100 grades assigned to teachers are numerical data, and as such, subject to FOIL. This clearly suggests the public shaming of teachers based on junk science will be an annual event.
When you are told it will not be, remember we were also told the grades that were released would not be. While I don't think the agreement was ever a good idea, I'm not faulting UFT leadership for the release of grades. There was, in fact, an explicit agreement they would not be. But Michael Bloomberg's morally bankrupt Department of Education recognizes no rules, and after having worked to break the agreement, simply lied about it.
Here's what our unions cannot promise--that the DOE will keep its promises, even written ones. This is particularly true because the DOE cannot make promises on behalf of the Daily News, the NY Post, or even the faux liberal New York Times. In fact the DOE can ask these papers to FOIL info help them out of inconvenient promises, as we've seen this week. If these and other media outlets see fit to FOIL numerical teacher data, as included in new reports, there will be nothing we can do about it. And our First Annual Shaming is a precedent, highly likely to be repeated.
It's time to back the New York Principals and stop this in the legislature. It's time to stop appeasing anti-teacher zealots. It's time to look this thing right in the face and see it for what it is--a disaster waiting to happen.
Update: In Rupert Murdoch's rag, the witch hunts have begun.
Posts
