Standardized testing -- especially the high-stakes variety -- has earned a serious and growing backlash, and for good reason. The weight of the research evidence shows that it has not improved education, and that it undermines the kind of academic behaviors that support critical and divergent thinking. High-stakes testing has distorted the teaching and learning process, resulting in time taken away from actual instruction, cheating scandals, and more. Testing is also expensive, representing yet another way in which scarce education funds have been diverted away from student learning toward powerful private interests. And while they're inelegant at best in performing their intended function -- measuring student knowledge -- they're now being inappropriately used to close schools, evaluate (and shame) individual teachers, and more.
Clearly, this system is fraught with issues. Yet policymakers and some members of the public still argue that, despite their flaws, it's good to have an "objective" measure of student, teacher and school performance -- an argument I find completely maddening. I've said it before and I'll say it again: standardized tests are not objective.
Just because a process somehow results in a number, does not necessarily mean anything about that process is objective (unless you're prepared to argue that men standing on a street corner rating passing women on a scale of one to perfect 10 are engaging in an "innovative form of female evaluation," instead of borderline sexual harassment). At every step of the evaluation process, human beings are making subjective decisions and judgments about what constitutes mastery of a given subject area, and whether someone has demonstrated such mastery. Once again, for the cheap seats in the back: the evaluation of intelligent human behavior will always be a subjective process.
Given that, we are faced with a choice between 1) a subjective evaluation system that emphasizes the judgments of people far removed from the classroom, with dubious qualifications, scoring tests under ridiculous circumstances, or 2) a subjective evaluation system that capitalizes on the trained judgments of informed people involved with the teaching and learning being evaluated. (As a former student and teacher, I'd prefer that my merit be judged by qualified people who have actually seen what I can do, than by some random person I'll never meet who was hired on Craigslist. So here's at least one vote for system number two!)
In the spirit of not complaining without offering solutions, here is my overview of the characteristics of a better evaluation system. I'm speaking in terms of teacher evaluation, but note that these characteristics/this overall structure could easily be analogized to evaluating students, school leaders, and schools.
In my entirely subjective, but pretty intelligent, opinion, a sound evaluation system would emphasize:
1. Shared evaluative responsibility, among multiple constituent groups. All sorts of bad things can happen when one principal with an axe to grind (or way too much on his or her plate) holds the sole power to judge teacher performance. No one's fate should ever rest on the singular judgment of one individual. Instead, several informed individuals should contribute meaningfully to each person's evaluation, including instructional leaders, peer teachers, students, parents and the teacher her- or himself. Including multiple perspectives on each person's performance has several advantages, like partially offsetting any particular individual's bias, and empowering teachers to examine their performance through the eyes of those they work with and serve. This can help them spot systematic strengths and weaknesses (say, not attending enough to parent relationships) and improve their practice.
2. Ongoing, actionable feedback -- all substantiated by concrete evidence, and embedded in real work. Evaluation should, first and foremost, be about supporting people to do their best work, not punishing them. It should also be realistic, without adding significant amounts of time or work to the evaluator or evaluatee's current responsibilities. Therefore, it should be an ongoing process throughout each school year, not just a once-a-year (or once-every-few-years) experience. This helps teachers improve their practice in real time, benefiting their students immediately instead of at an unknown time in the future -- if at all.
Moreover, every rating should be evidence-based, especially if the information is going to be fairly used to make high-stakes decisions like dismissal or promotion. Evidence, in this case, would be things like artifacts from the classroom, videos and/or real-time observation, and true examples of student learning -- their actual work, instead of problematic test scores chewed up and spit out by an opaque algorithm.
3. Multiple, varied ways for teachers to demonstrate that they've met their professional standards. All teachers are different, just as their students are. So they need to be supported to develop their unique strengths, which collectively help the whole school to develop their students' unique strengths. For example, noting that one teacher excels in using humor to engage students, while another is exceptionally adept at creating hands-on experiences for students who think concretely, not only supports school leaders to identify teachers' different leadership capacities, but also supports those leaders and students' families to identify which teachers will work best with specific kinds of students. (This is infinitely more helpful to those looking to make optimal classroom assignment decisions than an all-but-random numeric ranking could ever be. It's impossible to credibly determine who is the best teacher in a building. But it is possible to determine who might be the best teacher for a specific student.)
If any of this sounds pie-in-the-sky, or like a revolutionary breakthrough, it shouldn't. This is what many of the best schools -- here and abroad -- already do, and it's one key reason why they get better results for their children. We owe it to our students, and our shared future, to marshal the resources and the public will necessary to create school systems that invest in their people instead of tests, and that support each and every person to be his or her most intelligent and talented self.