> If you want to evaluate how good a teacher is, one of the things you need to measure is how well their students performed vs. how well those same students wou

> If you want to evaluate how good a teacher is, one of the things you need to measure is how well their students performed vs. how well those same students would have performed under the hypothetical average teacher. But underlying this assumption is that you have a good measure of student performance in the first place -- or that such a measure can easily be crafted. It's questionable if it even makes sense to quantify a student's abilities into a single catch-all number. But even if quantification of student performance makes sense, it becomes problematic when it becomes high-stakes (i.e. teachers' payment or school funding depends upon test scores). The problem with tests (or any other metric tied to incentives) is that they are gamed over time; SATs become increasingly meaningless with the cottage industry of test-prep; and so do standardized tests become increasingly meaningless as teachers are incentivized to teach directly to them. (In politics, crime rates are gamed, jobs rates are gamed, etc.). Were schools so terrible before standardized tests -- when administrators and other educational experts rated teachers? There were existing non-mechanical rating systems that of course predated standardized testing -- is it necessary to mechanize evaluations simply because we like the superficial illusion of objectivity, even if that objectivity bears no resemblance to the thing we are trying to measure? Or even acknowledging that perhaps existing non-mechanized teacher performance measures may have been flawed, have standardized tests somehow improved schools? No, probably not -- teachers and students both are more miserable and are increasingly tied down by the requirements of possibly meaningless tests.