The (Education) Heisenberg Uncertainty Principle
As the debate continues to rage about the place of standardized testing in American culture (and other cultures — this piece in Time on South Korea is quite interesting), I feel as though a few things about transparency and data need to be laid out.
The number one most important principle for the use of assessment and data:
First do no harm.
I think this is what Diane Ravitch, among many others (I’m going to refer to her and them as “Ravitchians” for simplicity’s sake), has a great point about. The stated goals of NCLB were good: closing the achievement gap, promoting equality of education, and improving teacher quality/distribution among underserved student populations (minorities, low SES students, ELL students, special ed. students, etc.). All great. Ravitch’s point is that, unfortunately, the cure has been worse than the disease. Specifying minimum goals for reading and math, and then attaching high-stakes tests to those goals, has meant that other crucial subjects such as science, history, art, and music, have been crowded out. This is becoming more and more of a concern, as people begin to ask whether killing ourselves to excel on multiple choice tests is sacrificing the innovative/creative ability of future generations, which is generally considered to be America’s greatest strength.
In light of the obvious negative effects of the NCLB standards/high-stakes testing regime, the Heisenberg Uncertainty Principle is what comes to mind for me. In an (oversimplified) nutshell, the principle states that you cannot accurately know both the position and the velocity of a particle at the same time, because the very act of measurement of one changes your ability to measure/effect the other.
That’s what the Ravitchians (and myself), have become worried about: We’re testing so much that the testing itself is causing far too many negative consequences to be worth it.
But there’s an important difference between saying that standardized testing, as we do it now, is bad, and all testing/assessment is bad. Here’s where I think anti-Ravitchian folks get it wrong: I don’t think that critics of the NCLB and standardized testing regimes don’t want to know how students and teachers are doing, and I don’t necessarily believe they think it’s bad to hold teachers at least somewhat accountable for student achievement. They’re not anti-assessment. Of course it’s crucial to know how well a student is learning and grasping concepts. It’s just as important to know how a teacher is doing — where a teacher can improve and where a teacher is excelling.
It’s just a matter of figuring out how we do those latter things. The point of the Ravitchians is that assessment matters, but how we do assessment is just as crucial. To take a (fairly tired) example, it’s not that teachers and administrators in Finland don’t know how the students are doing, but it’s that they’ve figured out how to assess progress (and teacher quality/need for improvement) in a way that doesn’t create a drill-and-kill, standardized-test-cheating infused culture.
Let me say it again: Data is crucial, and should not be used in a punitive manner. We know how to assess students without using multiple choice tests. Half of what teachers do is assess, often in informal ways. Tests are just one, highly formalized, way of finding out what students have learned. Classroom polls, portfolios, exit slips, one-on-ones, tracking journals, etc. are all ways of keeping track of what students learn.
For those (including myself) who say that we must tie student achievement to teacher performance, it is no answer to say that we must use standardized testing as our standard for “student achievement.” Given the problems with using value-added data down at the individual teacher level, it’s at least worth discussing what measures of student achievement we want to use when evaluating teachers. Standardized, multiple choice tests are not our only option, and they are probably (certainly?) not the best option.