With every bubble of knowledge that students darken using their #2 pencils, our nation increases its infatuation with measuring teachers' performance through students' standardized test scores. Despite the intuitive appeal of evaluating teachers based on student learning, heavy reliance on these indicators narrows our view of effective teaching in potentially damaging ways. Recently, the sports world has improved its assessment of athletic performance using advanced metrics and other accountability techniques. Learning from these advances will help us avoid firing good teachers and retaining lousy ones.
First, consider NBA legend Kevin Garnett. If he were a teacher, he would have been rehired this year? Maybe, maybe not. Last year he averaged less than 15 points a game. Decent, but because winning games is ultimately about scoring points, many other players seem more appealing.
However, in the same way that scoring points is merely one aspect of basketball, effective teaching encompasses much more than students' test scores. The NBA has increasingly supplemented traditional statistics like points, rebounds, and assists with more comprehensive, informative measures like"plus-minus" (whether your team outscores the opponent while a particular player is in the game).
Thanks to "plus-minus" we know that 36-year old Kevin Garnett remains among the NBA's elite despite the drop-off in scoring from his younger days. No sane general manager would let him go. Garnett has classroom equivalents. Think of the 7th grade math teachers with modest student test scores whose extra work ensures that their students excel in algebra two years later. Teachers play multiple roles which affect a variety of student outcomes and our measures need to reflect these multiple roles.
Second, we might ask whether educating students is fundamentally a team or an individual sport. Although it is easy to conjure up images of teachers isolated in their classrooms, almost nobody can teach Spanish, art, and AP physics with equal facility. Teams of teachers, counselors, and specialists are required to educate all students.
NFL teams understand this. When studying game film, NFL players may see the same ESPN highlights as we do, but they focus on very different parts. We remain transfixed by the highly paid receiver sprinting into the end zone, but they focus on all six blocks that allowed the receiver to get open. The sole goal of these film sessions is to make the team better and hold each teammate accountable for his role. By contrast, many school districts' incentive pay systems reward teachers for out-performing, rather than helping, their teammates. Clearly, new measures need to account for team contributions.
Third, we need to contemplate how much data are needed before we can evaluate performance. In MLB, firing pitchers or signing sluggers to big contracts based on a couple of home runs in one game would be lunacy. Managers have the majority of a 162 game season to assess players' performances and make trades before finalizing their rosters.
All too frequently, the bulk of teachers' evaluations are based on a single set of students' test scores. Like most baseball statistics, students' test performances depend on multiple factors (not just their teachers): How well did last year's teacher prepare the students? Did the school counselor assign the class clown to Jones or Smith? Do students think their test even counts? By measuring student learning on multiple occasions and in multiple ways, educators can separate the signal from the noise.
The sports world has embraced these ideas of measuring performance comprehensively, balancing individual efforts with team contributions, and measuring more often. So how might educators learn from these not-so-dumb jocks? Unfortunately, desired educational outcomes are more complex than mere wins and losses. Thus, reliance on a few simple, elegant statistics like "plus-minus" is not feasible.
Instead, we need a constellation of measures to accurately reflect teachers' multi-faceted roles of teaching, affecting student learning, bolstering student motivation, supporting colleagues and providing service to the school. To assess this broad array, an evaluation system might leverage classroom observations, tests and school records, as well as surveys of students, other teachers, and administrators. Many districts already have these measurement tools -- they just need to be applied more creatively.
To better assess teaching, we can supplement classroom observations with periodic student surveys (at least for older students). After all students "observe" teachers every day.
Student learning should be assessed more frequently -- shorter, regular computer-based tests could test students' skills, knowledge, and thinking and provide immediate feedback. We can capture teachers' effect on student learning through new measures of students' future academic success. For example, how well do students who had Teacher X for 7th grade math perform once they get to algebra?
We can easily assess student motivation through the aforementioned student surveys, but also through future course enrollment. (After all, how excited should we be about the chemistry teacher with great student test scores if the girls stop taking elective math and science courses after being in her class?)
To assess whether teachers are good teammates, why not have them report on which colleagues most help them improve their instruction, handle challenging students, etc.?
Finally, schools need teachers who will monitor hallways, have the challenging students in their class, chair departments and so on. A combination of school records and administrator reports can inform which teachers should be thanked and rewarded for these usually thankless tasks.
By measuring more comprehensively, by recognizing teachers as part of a team, and by measuring more frequently, we can better distill which teachers are most effective. Like good sports metrics, this approach would have the added benefit of pinpointing where improvement is needed. To be sure, this is more complicated than the current approach. Yet the effort seems worth it. After all, aren't we all on the same team when it comes to helping students succeed?
The point is, that it is whole systems that need to be overhauled to encourage teachers to want to be learners, rather than measuring teacher "effectiveness" against educationally-invalid indicators, such as test scores.
The difference between successful schools and "failing" schools rarely has much to do with the quality of the teachers.
Seriously though. I was a professional scientist in the private sector before becoming a teacher and I never received nearly the amount of observations and evaluations in my previous career as a scientist in the private sector as I do as a teacher. No private sector business would even consider the teacher evaluation process as it would put most of the private sector out of business due to the excessive time and costs.
Like every other profession, let the professionals (the teacher) do the evaluating and observations. They know who does and does not have the chops to be a teacher.
Furthermore there does not have to be a long check-off list to perform evaluations and observations like the author is proposing. It will be costly to taxpayers and waste the already limited time we have with students.