Atlanta teacher convictions: Do standardized testing pressures foster cheating?

A jury convicted 11 educators of racketeering Wednesday for their role in the Atlanta cheating scandal. But nationally, there’s a strong split between those who see their actions as an aberration and those who would convict right alongside them the accountability systems that have attached increasingly high stakes to standardized tests in recent decades.

The teachers and administrators face potentially harsh sentences for a conspiracy to manipulate test scores – which investigators said involved more than 44 schools and about 180 educators. Eleven out of 12 who went to trial were convicted, and they were sent immediately to jail to await sentencing (with the exception of one who is pregnant).

For opponents of such high-stakes testing, there’s likely to be more sympathy for the educators because of undue pressures being placed on teachers by an overemphasis on test scores. But for proponents of accountability, it’s just as easy to hold up these educators as an example of why strong objective systems are needed to oversee and measure educators’ performance.

The pressured atmosphere doesn’t justify cheating, but it’s one indication of a much larger problem, say critics of how testing has been used.

Especially as the federal government has pushed states to tie teacher evaluation policies to standardized-test gains, the testing regimen “creates a climate in school where you have to boost scores by hook or by crook,” says Robert Schaeffer, a spokesman for the National Center for Fair & Open Testing (FairTest).

Atlanta offered up extreme examples such as test-cheating “parties.” But “Atlanta is the tip of the test-cheating iceberg,” Mr. Schaeffer says, with other cases surfacing in about 39 states, including a dozen or more that showed widespread cheating.

The El Paso, Texas, superintendent went to prison in 2012 for fraud for manipulating federal accountability measures, and nearly a dozen others were held accountable for their role by the state education department.

An Arizona State University study surveyed Arizona educators in 2010 and found that 39 percent knew of situations in which colleagues encouraged students to redo test problems, while 10 percent knew of colleagues who did something they considered more outright cheating.

“We have a system in which people are told all the time that all that really matters is raising test scores,” says Daniel Koretz, a professor at the Harvard Graduate School of Education. 

Some examples of the “shortcuts” teachers are encouraged to take? Teachers are often shown “power standards” – the types of items most commonly tested – by administrators, and sometimes are taught to skip chapters of textbooks that don’t fall in that category, Professor Koretz says. And states now routinely offer teachers old test items to use for test prep, a practice frowned upon in the 1980s.

“Clearly cheating is unethical, but at what point does this other stuff become unethical?” he says.

Despite the conviction of the Atlanta test-cheaters, which may make people and systems more cautious to guard against outright cheating, Koretz says he’s skeptical that it will have much impact on this broader problem of shortcuts that shortchange students of quality teaching.

But some argue that cheating scandals shouldn’t be leveraged in the debate about accountability systems.

“There are plenty of reasons for teachers to take issue with some of the teacher evaluation [policies] that have been rolled out across the country.... But I’m a little bit troubled when folks say, ‘Oh, and it’s driving teachers to cheat,’ ” says Michael McShane, an education policy research fellow at the American Enterprise Institute in Washington.

People opposed to standardized testing might use the Atlanta cheating scandal “as a story of educators trapped in an unjust system ... like it’s almost like some form of civil disobedience, which is just not the case.... [The cheaters were] eroding trust in public institutions and setting horrific examples for young people in Atlanta,” Mr. McShane says.

There are many variations of state systems that tie test scores to evaluations, and most give 50 percent or less weight to such scores. The Obama administration’s Department of Education encouraged such policies through its Race to the Top grants and its waivers to states from certain portions of the No Child Left Behind accountability law.

But as opposition from teachers’ unions and parents has grown, particularly amid the challenging logistics of rolling out new tests tied to the Common Core State Standards Initiative, there’s “a real inclination among many people, even those who ultimately support the use of testing in evaluations, to hit the pause button,” says Patrick McGuinn, a political science and education professor at Drew University in Madison, New Jersey.

Supporters of test-based accountability say it’s important to keep something in place to check across schools, districts, and even states, on how students are doing, and to be able to analyze data by race, gender, and other categories to ensure that disadvantaged students are well-served. Historically teachers were not held accountable to such standards, and reformers “fear that this short-term pause is a ruse to put if off indefinitely,” Professor McGuinn says.

But research has long shown that state standardized testing can be gamed, and usually does little to actually improve the amount of learning taking place, Koretz says. For instance, students often make big gains on state tests while at the same time showing little progress on the National Assessment of Education Progress (NEAP), a national snapshot of reading and math skills at certain grade levels.  

Koretz says true accountability would include many unstandardized measures of student and teacher performance, everything from portfolios to observations, and that a limited amount of standardized testing then could be part of the oversight system to make sure teachers were applying appropriate standards. Instead, the way testing has been used, he says, has “taken an extremely complicated accountability problem and reduced it to something that’s ludicrously simplistic and just hasn’t worked.”