> Your performance is measured in a single week at the end of each year by statistically-controlled standardized tasks.
My main point was that if you're going to appraise teachers based on their students' standardized test scores, it would seem to make sense to use a statistical model that tries to measure how much of an effect the teacher has on the predicted test scores of their students. I'm completely open to the argument that we shouldn't be appraising teachers based on student test scores, but that does not seem to be the position the article is advocating.
My main point was that if you're going to appraise teachers based on their students' standardized test scores, it would seem to make sense to use a statistical model that tries to measure how much of an effect the teacher has on the predicted test scores of their students. I'm completely open to the argument that we shouldn't be appraising teachers based on student test scores, but that does not seem to be the position the article is advocating.