THE PRINCIPAL'S OFFICE

State is wrong on teacher evaluations

Posted

I’m putting my current series on hold for an important message:

It’s official: The Regents have approved a new teacher evaluation plan. In short, the formula contains two variables: classroom observations (50 percent) and test scores (50 percent). This is wrong, wrong, wrong!

Previously, I’ve written and reported extensively in this column and during my reports on WCBS Newsradio 880 about my strong opposition to giving so much weight to the tests. Now it’s the “law of the land.”

Although I want to focus on test scores, I’ll comment briefly on observations. Obviously, they’re key variables in the equation. However, one caveat: They can be very subjective. Hence, they should be conducted by two, preferably more different supervisors. Generally, the minimum number is different for tenured and probationary teachers.

There are several “types.” First, there are unannounced versus announced (preceded by a pre-conference and followed by a post-conference). Some administrators differentiate the two types as “evaluation/rating” versus “supervision/improvement.” Second, there are brief “pop-in” visits as well as full-period, bell-to-bell write-ups.

So much for observations, at least for now. I want to enumerate the many reasons I’m so against making test scores count for 50 percent of the ratings. These numbers are based on so-called “growth” (improvement) scores from pre- to post-test. It sounds good in theory; to non-educators, this formula makes total sense. After all, they say, a teacher’s evaluation should be based on achievement. But it’s not that simple — and I’m going to poke holes in that reasoning.

To begin, is there consensus as to what the standards should be — and what’s being tested?

Second, have teachers been adequately prepared and given the necessary books and materials to deliver these lessons?

Third, who’ll construct the pre- and post-tests? Individual teachers? Colleagues? The state? A private firm?

Fourth, is there a guarantee that these tests will be reliable and valid (measuring what they purport to measure)?

Page 1 / 3