Indy Democrat has a post about the recent teacher evaluation reports showing that 87% of Indiana teachers were in the top two categories. The post indicates that the low fail rate has caused skepticism by Rep. Behning and State Board of Education board member Gordon Hendry.
“I find it hard to believe that a system of evaluation where only a handful of people are said to need improvement is accurate or effective,” at-large board member Gordon Hendry said. “Clearly, the system failed.”
“We didn’t think it was possible for a D or F school to say all teachers are effective or highly effective,” State House Education Committee Chairman Bob Behning said. “We thought [the school ratings] would keep schools somewhat a little more honest.”
There are a number of possibilities. Maybe the teachers aren’t, by and large, the problem in schools where kids are failing. Maybe chaotic home lives and a social setting that doesn’t value education, combined with a tattered social safety net contribute more to struggling schools than the merits or lack of merit of particular teachers. No offense to the fine teachers and administrators in the West Lafayette school system, but I’m of the opinion that you could likely put some fairly bad teachers in our system, and the students in our area would, by and large, emerge well educated. Teachers are just one variable in a very large system, and the metrics we use to evaluate teachers and schools have never been very well thought out in terms of controlling for all of the variables that are well outside the control of the school system.
But, we have a rush to measure things; often in hopes of confirming preexisting biases and sometimes in the more venal hope of transferring gobs of public money from a constituency that donates and votes for the other guy to friends and well-wishers of one’s own political campaigns.
I tend to agree that a metric which lumps almost the entire population being measured into one or two categories probably isn’t calibrated in a way that provides much utility. But, first we need to clearly articulate what the hell we’re trying to measure. Then we need to test the tests to see if they actually measure that thing.