Monday, March 28, 2011

"Value-Added" Teacher Evaluations

This is how it's done in Houston:
"It is too unreliable when you're talking about messing with someone's career," said Gayle Fallon, president of the Houston Federation of Teachers.

She said many teachers don't understand the calculations. The general formula for the "linear mixed model" used in her district is a string of symbols and letters more than 80 characters long:

y = Xβ + Zv + ε where β is a p-by-1 vector of fixed effects; X is an n-by-p matrix; v is a q-by-1 vector of random effects; Z is an n-by-q matrix; E(v) = 0, Var(v) = G; E(ε) = 0, Var(ε) = R; Cov(v,ε) = 0. V = Var(y) = Var(y - Xβ) = Var(Zv + ε) = ZGZT + R.

"It's doctorate-level math," Fallon said.

Does Ms. Fallon have a doctorate in math? If not, how does she know the method is unreliable?

That isn't really the point. The point is that the teachers themselves can't really tell if the method is reliable or not, and hence cannot have any trust in the system. Trust is the real issue.


Mr. W said...

While I don't have a problem with evaluating teachers or even state tests, I don't think the whole "value-added" method works in certain areas. Take math for instance. Can you really compare a Geometry score with an Algebra 2 score? There is only one thing from Geometry that carries over into Algebra 2 and that is writing equations of circles. So how can you use that score to tell me if I am a bad/good teacher?

I won't beat the dead horse that always says something to the extent of "and there is no buy in or reason for the kids to try, so why judge me on something that doesn't matter to them"...but I won't go there.

mazenko said...

The inane idea of treating schools according to some business or scientific model is at the root of all the problems plaguing reform these days. And much of it is because left-brained business leaders like Bill Gates are guiding much of the reform discussion.

Steve USMA '85 said...

For what it is worth, it is actually Master's level math not doctoral. Very common type regression model. But, still, the model can be highly influenced by outliers and choice of variables used/not used.

Mr. W said...

I gotta agree with Mazenko here. You can't run a school like a business. I have said this before, but when I was a supervisor at UPS I could fire bad employees (late or didn't work hard). I am stuck with these students. If they don't work...I can't kick them out of school. Heck, they can't even get suspended for not caring. And yet, they expect us to turn every student into an advanced student.

I think all politicians should have special ed or "basic-level" people become their secretaries or something and tell them that they need to make these people best secretaries around. How long would it be before they are fired?