Tuesday, May 15, 2012

Looking at the RI Student Growth Model

Elisabeth Harrison:

Here’s the Rhode Island Department of Education’s newly minted tool for comparing school performance.

The website graphs both proficiency rates on annual state tests of English and Mathematics and individual score changes from one year to the next. Education officials call it “the Rhode Island Growth Model Visualization Tool.”

Starting next year, the state will start keeping track of growth data for individual teachers and classrooms.

This tool is pretty uninteresting aside from its implication for future school and teacher evaluation. In particular putting growth on the X axis and overall achievement on the Y for all the views makes it hard for me, as a a parent or member of the general public looking at schools and districts, to draw any non-obvious conclusions.

The big flaming omission here is the inclusion of only one year's worth of data. If there is a practical reason for that, they should say so. Probably it is simply to obscure the instability of the scores year over year.

What you should check out, just to get a sense of where all this is going, is double clicking on some schools and get the breakdown by grade, which in some cases should narrow you down to the scores of just a handful of teachers. Then you can see, for example, that the fourth grade teachers at Vartan Gregorian might be on the hot seat soon. Or it is just random noise? There's a 30 point spread between the growth percentiles in grade 4 and 5 math at Reservoir, where I'm pretty sure it is just two teachers per grade level, but only a 6 point spread in reading. I'd love to see how these numbers jump around over the past five years. RIDE should have that data, and CPU cycles are cheap!

For more on student growth percentages in teacher and school evaluation, see Bruce Baker.

No comments: