It's Not the Test That Made Them Cheat

It's ridiculous to say that teachers and administrators in Georgia and elsewhere committed crimes only because they had to give high-stakes tests to students.

The Georgia indictments are, in fact, just the tip of the iceberg of White Chalk Crime in schools.

The indicted educators tried to get rewards that they had no right to, and to take them away from those who deserved them. This happens all the time in schools. It's the reason the education system is failing. The wrong people are in charge of most districts, and too many of the wrong people are teaching kids.

If we had a system for correctly evaluating teachers, like this one, we wouldn't need to test kids so much, and we would save billion dollars a year. We could trust the teachers to do their job, and part of that job is to know exactly what each child knows and how each child learns best. Standardized testing could go back to its original purpose: to make an official determination of which kids are most advanced academically.

It's Not the Test That Made Them Cheat
By Michael J. Feuer
Edweek.org.
April 9, 2013

News came down, or up, earlier this month about the indictment of the former Atlanta schools chief Beverly Hall and 35 other current and former officials for their alleged roles in a massive cheating scandal that has rocked the city for the past three years.

The best coverage of this story is by Atlanta Journal-Constitution reporter Heather Vogell and her colleagues, whose fine journalism uncovered the muck.

There is nothing good to say about cheating on tests, which, in this extraordinary case, involves allegations of tampering with student answers, racketeering, theft, influencing witnesses, conspiracy, and making false statements. It’s wrong, period, and if Ms. Hall et al. are found guilty, they will hopefully use their time in jail to think about the damage they have caused to the kids, to the system, and to the public’s trust in schools and in the measures we use to gauge their quality.

Still, some of the reactions to the scandal have been surprising, if not scandalous in their own right. The most troubling response comes from people opposed to standardized testing generally and to current federal policy specifically. They somewhat gleefully use this sorry episode as the ultimate smoking gun, the perfect we-told-you-so case that clinches their claims about the evils of testing, and, by extension, the entire reform movement. It’s a big nail, they hope, in the coffin of test-based accountability.

"Shall we excuse individual or group criminality because certain social institutions create pressures for greed and misconduct?"

Among the more remarkable statements is one posted by William C. Ayers on Valerie Strauss’ Washington Post blog. For Bill Ayers, an education professor emeritus from the University of Illinois at Chicago, the Atlanta story proves that “teaching toward a simple standardized measure and relentlessly applying state-administered (but privately developed and quite profitable) tests to determine the ‘outcomes’ both incentivizes cheating and is a worthless proxy for learning.”

Mr. Ayers goes further. Not only does he attribute the alleged cheating to the testing policy, thereby essentially absolving Ms. Hall and her colleagues of their own ethical and professional lapses, but he uses the example to issue a sprawling condemnation of the U.S. Department of Education, Secretary of Education Arne Duncan, and even the president. As he puts it, “the road to the massive cheating scandal in Atlanta runs right through the White House.”

I have four problems with this logic (echoed in other commentaries, such as Jason Stanford’s bold assertion in the Huffington Post that “high-stakes testing makes cheating inevitable”; and FairTest’s pronouncement in an Atlanta Journal-Constitution op-ed last year that “[t]hese scandals are the predictable result of overreliance on test scores”).

First, shifting the blame for egregious mischief away from the perpetrators and onto the system strikes me as morally and politically bankrupt. Here’s an analogy to consider: Do we react to the worst instances of tax evasion by condemning the concept of taxation rather than by prosecuting the evaders? I assume that Mr. Ayers would not call for abolition of the graduated income tax as a way to finance public goods and redistribute wealth just because the system has its imperfections and because some people lie on their tax returns. Shall we excuse individual or group criminality because certain social institutions create pressures for greed and misconduct? Banking executives accused of fraud will be delighted.

Second, even if one could make an evidence-informed case that testing “inevitably” leads to illegal behavior—as if high-stakes testing overwhelms the human capacity for moral choice—there is the added problem of guilt by association. Pinning the responsibility for the Atlanta disaster on the White House is an extravagant example of misdirected blame. Maybe current federal policies lead to unwanted outcomes, such as narrowing of the curriculum and teaching to the test, but that’s a far cry from the outright fraud of the sort listed in the Atlanta indictment. Nothing in the No Child Left Behind law requires states or districts to use test scores to fire teachers and principals or to protect and reward those who achieve targets by tampering with answer sheets. In any case, there’s no evidence that federal policy causes cheating, or that “cheating is inevitable.”

Third, indicting testing, rather than cheating, undermines the possibility for reform in the design and uses of tests. The compelling logic in Campbell’s Law—“The more any quantitative social indicator is used for social decisionmaking, the more subject it will be to corruption pressures ...” —is supported by abundant empirical evidence on the effects of overreliance on tests for accountability.

But what’s often ignored in the popular frenzy against testing, especially in the wake of cheating scandals, is the benefits side of the argument: Tests can help gauge individual learning, give teachers additional information about their students’ progress, provide objective indicators of student achievement, and expose inequalities in the allocation of educational resources.

We may never be able to completely “overturn” Campbell’s Law, but what’s needed is a sensible approach to assessing the ratio of benefits to costs and to the design of mechanisms meant to keep the ratio strongly positive.

Fourth, it turns out that in Atlanta there were schools, and kids, that actually did improve during Ms. Hall’s tenure, according to data from the National Assessment of Educational Progress. As U.S. Commissioner of Education Statistics Sean P. “Jack” Buckley noted, “There were obviously rampant violations of testing integrity going on there, but there were also schools there that were legitimately improving.”

We shouldn’t allow score gains inflated from cheating to be misconstrued as evidence that any measured improvement in student learning—especially among poor and minority children—must always be the result of cheating or other mischief. This kind of smoking-gun logic saps the morale of educators, parents, and policymakers working in behalf of our most disadvantaged students and provides free ammunition to those who believe investments in public education are essentially futile. Just when we education researchers and social scientists are facing increasingly mean-spirited political challenges to our profession, Bill Ayers’ and others’ evidence-free diatribes further erode public confidence in the credibility of our work. One can only hope that the temptations of guilt by association won’t prevail, and that the research community as a whole won’t be blamed for the shoddy logic of some of its members.

Michael J. Feuer is a professor of education policy, the dean of the graduate school of education and human development at the George Washington University, and president-elect of the National Academy of Education.