In eliminating test scores to evaluate teachers, New York State has taken a step in the right direction (“New York Lawmakers Pass Bill to Drop Student Test Scores From Teacher Reviews,” The Wall Street Journal, Jan. 24). Although the action will be attacked for lowering the quality of education, it reflects reality.
So much of instructional effectiveness is dependent on the students that teachers happen to inherit. Teachers who are assigned a group of Talmudic scholars will shine in spite of, not because of, their expertise. Conversely, teachers who are assigned a group of future felons will fail in spite of, not because of, their expertise.
The only way to get around the issue is to randomly assign students. That eliminates the inherent advantage or disadvantage teachers have in producing results. But such a policy will never happen because principals want to reward certain teachers and punish others.
(To post a comment, click on the title of this blog.)
6 Replies to “Evaluating teachers fairly”
Not sure “random” assignment would work — assignment would have to be specifically designed to even out teacher burden rather than randomly. If pupil assignment were done randomly, then some teachers would get an above-average mix and some would get a below-average mix. Like rolling dice 25x and adding up the totals — the totals would vary from 25 to 150, although the highest points on the distribution curve would be in the 75 to 100 range.
As you note, some principals would not make the assignments in a good faith effort to equalize teacher burden (or even randomly).
There’s the additional problem that, for the good of the students, it is usually best if principals have discretion to assign more of the “problem” students to the teachers who are better at handling the “problem” students. Of course, it follows that principals, in writing the teacher evals, must take into account the differing teacher burdens that result.
During my years as a manager of attorneys in a govt law office, we routinely matched case assignments to attorney abilities — the more difficult and/or important cases usually went to the best analysts; the long but less complex cases usually went to the most dogged attorneys. And, of course, we took the differing case burdens into account in writing the attorney evals.
Labor Lawyer: Random assignment of students is not perfect, but overall it’s the fairest way of evaluating teachers. Few principals take into account the kind of students that teachers have. As a result, some teachers get far better ratings than they deserve and vice versa. Incorrigible students need to be placed in special classrooms taught by teachers with advanced training.
So the reward for becoming a better teacher is larger classes and more of the “incorrigible” (not able to be corrected, improved, or reformed) students? That’s not been working out too well for my health. Why the magical thinking about teachers? There are no easy answers, that’s for sure.
MJSB — I was not advocating that better teachers be assigned larger classes or more of the problem students. Rather, I was advocating that principals — like all managers — should have discretion to match teacher strengths to student needs. I had a lot of personal experience with this concept when I managed attorneys in a govt law office — the office’s overall work product was better, the attorneys were happier and the supervising attorneys were less hassled when the managers, in assigning cases to attorneys, considered the attorney’s strengths/weaknesses when assigning different types of cases. The analogy to teachers is not precise — the supervisors and managers could and did ultimately reward the attorneys who were the best analysts/best writers with promotions. However, under this system, the actual work load of each of the attorneys was roughly equal because the attorneys who were less-skilled at analysis/weaker writers tended to get cases that required more hours of relatively tedious effort.
Applying the approach to teachers, it seems that there will rarely be a universally “best” teacher. Some teachers will be better suited for teaching “problem” students; other teachers will be better suited for teaching high-achieving students; still other teachers will be better suited for teaching minority or low-income or ESL students. In any event, principals, if they do try to match teacher strengths to student needs in student assignment decisions, should do their best to ensure equal total teaching burdens among teachers.
I agree with EdHed that many — hopefully not most — principals will reward their teacher favorites with easier student assignments. In theory, requiring random student assignment would eliminate this problem. But, I think that the favorite-playing principals would just find either a way to rig the random assignments or another way to give their favored teachers an easier life (like better classrooms, better subject assignments, fewer preps). And, random assignment would, as I noted in my original comment, not necessarily produce equal teaching burdens — sometimes flipping a coin 10 times will result in 8 heads and 2 tails.
mrjscienceblog: Based on what I saw during the 28 years I taught in the same high school in the LAUSD, most principals reward their favorite teachers with the best students and punish those they don’t like with the worst. There are always exceptions of course. Incorrigible students need to be removed from classes where students want to learn and placed in separate classrooms supervised by teachers with special training and given combat pay.
Labor Lawyer: I’m glad you clarified the issue. There’s no question that some teachers are better than others in handling certain students. As long as the match is not done unfairly, I agree with you. Principals who engage in favoritism invariably lose the respect of their faculty. I taught under a principal who was a bully who harassed teachers he didn’t like for one reason or another. He destroyed morale but was never replaced.