Sunday, May 26, 2013

Even This Brookings Institute Shill Says APPR Won't Work

Given that this was in the Daily News, I had to check twice to make sure this writer was saying what I thought he was saying - the NY State teacher evaluation system based upon student test scores is unworkable as currently designed.

The Obama administration was mad as a hatter to require states to promise to evaluate all their teachers in part based on student academic growth, and the city and the UFT have fallen down the rabbit hole into this fantasyland.

Why? Because, despite reform rhetoric to the contrary, student growth can be calculated reliably and linked to individual teachers only for English language arts and math in grades 4 and 5 — a small fraction of those in the system.

To start, the universe of teachers taking the state test that allows for “value-added” calculations encompasses English and math teachers in grades 3 to 8. Kindergarten and first- and second-grade teachers are out. So are high school teachers, who cannot be judged on the Regents exam, because that’s an end-of-year test that cannot be used to measure student growth.

But even in that narrow subset, we start ruling even more teachers out. Third-grade instructors can’t fairly be included because their students don’t have the prior-year test scores that are necessary for growth to be calculated.

In fourth and fifth grade, teachers generally have their students for the full day, so they can legitimately be held accountable for results. But when students move to middle school (usually in sixth grade), they have different teachers for different subjects. The result is that their gains on the state English and math tests aren’t easily attributable to a single teacher.

Taking into account all these factors, as well as the churn of teachers as they move between schools and grades, the portion of the teacher workforce eligible for evaluation based on student growth is easily below 20%.

(And, it should be pointed out, even when a teacher can legitimately be measured using relative gains on student test scores, that portion of their evaluation adds up to just 20% of the overall evaluation, with 60% being principal observations and 20% being other locally determined measures.)

The only way around this would be to implement standardized testing of course-specific knowledge at the beginning and end of every class in every subject and in every grade. That’s a massively complex and expensive undertaking that New York does not propose to undertake.

What all this means: The union and the city have locked horns in a fight that neither can win, because evaluating all teachers fairly and validly based on student growth on standardized tests is, in the near term, a pipe dream.

The Brookings Institute shill thinks we should devise a system where all teachers can be evaluated using test scores but until that time, classroom observations and other criteria ought to be used to evaluate teachers.

He never gets into the unreliability of VAM or the wide swings in stability the system is known for when used on teachers.

But even this guy says that APPR will not work because it is not designed well and the test scores cannot be "calculated reliably and appropriately attributed to individual teachers."
 
Let the lawsuits begin!

No comments:

Post a Comment