Wednesday, January 11, 2012

The Folly of Ranking Schools by Test Scores

This was written by Alfie Kohn for the Calgary Herald in February of 2005.

by Alfie Kohn

There’s bad news in today’s Outlook section. It’s not that the test scores listed there are low. It’s that the test scores are listed there at all.

As a rule, the less people know about education – and about exams like the Provincial Achievement Test -- the more willing they are to judge schools (and students) on the basis of such scores. It’s not that standardized test results don’t tell us anything. They’re very accurate measures of the size of the houses near a given school. As many researchers, including Todd Rogers at the University of Alberta, have shown, a school’s test scores can be predicted with reasonable accuracy if one knows the socioeconomic status of its students – and nothing about what happens in its classrooms.

But what if a school’s scores are rising over the years despite having roughly the same kind of students? In that case, the results aren’t meaningless -- they’re worrisome. Better results on the P.A.T. may actually be a bad sign because of what had to be sacrificed in order to make that happen.

Those numbers in the Outlook section do not reflect academic excellence. Indeed, teachers in Alberta and elsewhere often tell me that some of their most impressive students don’t perform well on these exams, whereas other kids get terrific scores just because they’re good at taking tests. The research confirms this: Three studies have found that high standardized test scores often go hand-in-hand with superficial thinking. What’s true of individuals is true of schools. Some of the best may not be at the top of those rankings precisely because they don’t stuff kids full of testable (and forgettable) facts. Rather than slavishly covering the provincial curriculum, teachers in great schools help children to discover ideas. Their students are learning what it means to think like a scientist or an historian, not merely memorizing dates and definitions.

Indeed, in visiting Saskatchewan and Manitoba, I’ve been impressed by the nearly complete absence of standardized tests there. They spend their time and resources on teaching rather than testing, and their schools are better off as a result.

Any attempt to reduce learning to numbers is misguided. Schools just can’t be rated like laundry detergents. As one U.S. educational researcher put it, “Measurable outcomes may be the least significant results of learning.” The damage is compounded when those numbers are test scores. Exams like the P.A.T. tend to measure what matters least.

Worst of all is using these scores not just to rate but to rank, such that the emphasis is on who’s beating whom. This encourages a toxic competition between schools. Educators are less inclined to work together to do a better job at educating all our children because they’re forced to worry about their relative standing.

There are only two explanations for the wretched ritual of publishing these scores. The first is simple ignorance. The second is some sort of political agenda – exactly the kind, come to think of it, pursued by the Fraser Institute, which wants to privatize everything in sight and privilege profit-oriented enterprises over democratic public institutions.

Frankly, if my goal were to make public schools look bad and pave the way for vouchers, I would do exactly what they’ve done: use the language of “accountability” to push for more emphasis on testing -- and set schools against one another.

But a right-wing think tank can’t accomplish this alone. It needs Alberta Learning notonly to force schools to give all those tests but to boil down the results into summary statistics that are ripe for publication. (The State of Connecticut refuses to do this in order to thwart those who would try to reduce schools to a single set of numbers.) Last week, Education Minister Gene Zwozdesky told a CBC interviewer that it’s “patently unfair to make comparisons between schools based on scores.” The question is why he is pursuing policies that facilitate just such comparisons.

The Fraserites also need newspapers like the Herald to do their dirty work for them by publishing these charts. The effects aren’t always easy to see or measure, but every year about this time, Calgary’s schools become a little worse because this paper leads parents and teachers to focus on test scores rather than on more meaningful indicators of quality.

Elsewhere, I’ve written at length about those indicators. Is the focus more on understanding ideas from the inside out (rather than on memorizing facts)? Do students experience their schools as caring communities? Do teachers create democratic classrooms so kids can participate in making decisions? Do children often learn together (rather than alone)? Do parents receive qualitative accounts of kids’ improvement (rather than traditional grades)?

By contrast, the most constructive thing you can do with those pages in today’s Outlook section is use them to wrap fish. The numbers printed there are not only meaningless but dangerously misleading.

One figure, though, is worth looking at: the percentage of “exams not written” at each school. When that number begins to climb, it will mean that Calgary-area parents are realizing that they’ve contributed to the problem by allowing their children to write the exams. When enough of them get organized and decide to boycott the P.A.T.s, Alberta Learning will ask nervously, “What if we gave a test and nobody came?”

That will mean the flourishing of grassroots democracy and a more sensible set of educational priorities. And that, finally, will be something worth reporting.

1 comment:

  1. Totally agree with you on this one Joe. Sadly CBC just posted the Fraser Institute's Top 20 most improved schools in Alberta. Radio piece kept stressing the importance of standardized tests. Why does anyone lend credence to this report?


There was an error in this gadget

Follow by Email