Figuring out how students are doing in school is a lot harder than most of us realize. That’s what I’m taking away after digging into a recent study by the American Enterprise Institute.
Think about it for a moment. How do you know if a school is good?
If you’re like most people, you’ll probably say a good school will have a lot of kids graduate and go to college. You might call out “good” teachers at the school and, if you’re really tuned in, you’ll point to data that show how many kids are up to speed on reading and math, what wonks refer to as “proficiency rates.” Or maybe you think a school is good because your friends like it or a lot of people are trying to get their kids into that school.
As for me, I’d take it a step further. Everything I mentioned above is definitely a sign that a school is doing well, but I’m more impressed if a school can help kids who weren’t achieving those things achieve them, or if schools help kids perform at higher levels when they leave than where they started.
I looked at most of those things when my wife and I chose a school for our own son. But we also considered how close it was and the school’s approach to teaching and discipline.
It turns out those measures may not be as reliable as we might think.
Testing Isn’t The End-All, Be-All
At least when it comes to school-choice programs (think charter schools, vouchers to attend private schools, magnet programs, etc.), researchers found that improving test scores doesn’t guarantee students will have better chances of achieving long-term measures of success like high school graduation, entrance to college or higher incomes. They also found that some school-choice programs made significant improvements with graduation and college entrance without moving the needle on test scores.
Those who are already skeptical about testing, or flat-out against it, are using this study to validate their belief that tests should be done away with or at least disconnected entirely from school accountability. The authors explicitly say that’s not their intent. They’re also careful to recognize that it’s not just testing that presents a challenge and acknowledge that even the measures they used to judge testing are potentially flawed.
“Now, perhaps these attainment measures do not matter as much as we might think,” they wrote. “If schools have watered down graduation requirements so that students who should not be receiving diplomas are receiving them or if unprepared students are matriculating into college only to fail, perhaps even these measures are imperfect looks into what we really care about.”
These caveats they mention aren’t just hypothetical. For years, we’ve seen reports of graduation inflation or fraud—D.C. Public Schools is just the most recent example. And far too many parents and students have firsthand experience with required remedial classes in college, suggesting that students are often unprepared despite getting accepted. Even parent demand can be a misguided measure if parents are basing their information on faulty or incomplete information.
We Still Need Tests
Tests may not give us perfect information, but they still give us useful information. They expose achievement gaps between low-income students and wealthier ones, or between students of color and White ones. They show teachers and administrators where they can focus more of their attention, or who needs more support.
All the things we look at to determine a school’s quality are ultimately just approximations for what we really care about: “Changing the trajectories of children’s lives.”
There isn’t a perfect single measure, short-term or long-term, that can tell us unequivocally how well schools are achieving that goal. But many measures, considered together, can give us a better sense of what’s happening than none at all, or even one alone.
This is why literally every person I know who advocates for education reform recommends that we look at more than just testing. It’s why our current federal education law asks states to define other measures of school quality in addition to information gleaned from test scores.
Unfortunately, it doesn’t always work like that in practice. In Florida, for example, school grades are determined almost entirely by test scores. And the authors of the American Enterprise Institute study point out that most school choice programs live and die based on test scores.
Where that’s the case, they’re right to suggest that “test scores should be put in context and should not automatically occupy a privileged place over parental demand and satisfaction as short-term measures of school choice success or failure.”
Deciding which schools are good is complicated. It’s messy. But so is just about everything else in our lives.
That doesn’t mean we stop trying, and it doesn’t mean we abandon every tool that gives us insight on the quality of our schools. That’s not what the authors of this study advocate, and it won’t get us any closer to changing the trajectories of children’s lives.