Email Status
 

Volume 27, Number 6
November/December 2011

With Cheating on the Rise, Schools Respond

 

The cheating scandals that rocked school districts from Georgia to Connecticut have sullied school district reputations, ended education careers, and cast an unflattering light on the ugly underbelly of the accountability movement.

And the scandals may not be over yet. With test results increasingly tied to teacher evaluations, experts are urging school districts to remain vigilant because the incentives to cheat grow ever larger each year.

“When stakes are high, and teachers and principals can lose their jobs over low scores, they are more likely to cheat,” says David Berliner, a Regents Professor Emeritus of Education at Arizona State University, who has written about the unintended consequences of high-stakes testing. “If they believe it’s an unfair measure of their performance, the likelihood of cheating goes up.”

John Fremer, president of Caveon Consulting Services, a Utah company that has analyzed 20 million test results in 20 states, confirms that cheating is on the rise. Fremer, who has spent 50 years in the testing field, estimates that up to 2 percent of teachers and administrators cheat in some way on standardized tests.

The U.S. Department of Education is concerned. Spokesman Justin Hamilton says the department is working with states to ensure that states have a “robust response” to cheating scandals, with the department able to withhold or condition future federal funding if the response is lacking. Hamilton says the Department of Education is also considering requiring states to beef up security measures before receiving federal funds and requiring spot testing in districts with dramatic test improvement. “We’re looking into how else to best support state-led efforts,” he says.

Reports on testing irregularities have cropped up over the past three academic years from coast to coast, with confirmed cases of cheating by test administrators found in 30 states and the District of Columbia, according to the National Center for Fair and Open Testing, an organization that monitors the use of standardized tests to measure learning. The Baltimore City Public Schools found that cheating took place in 2009, 2010, and 2011, and Atlanta’s system was left reeling by a cheating scandal that left 178 professionals under investigation. Connecticut is investigating serious irregularities at a Waterbury elementary school that led to the suspension of its principal, vice principal, and every one of its teachers in grades 3 to 5. Florida is investigating questionable erasures on tests in 17 county school districts, and New Jersey has 34 schools under investigation for possible cheating. In one of the most recent announcements, the New York State Education Department found possible cheating in 62 schools, with six schools now under full investigation.

Investigators have discovered a broad range of irregularities—from erasures of wrong answers that were changed to right answers to principals who copied the tests before they were given and handed them out to teachers so they could tell their students the right answers.

Growing Pressures on Teachers
The pressures on teachers and administrators to boost test scores, especially in low-performing districts, are mounting. No Child Left Behind’s mandate that every child attain proficiency by 2014 has left districts scrambling to make the Adequate Yearly Progress (AYP) goals mandated by the law. The New York Times reported in August that 38,000 of the nation’s 100,000 public schools did not meet their targets in 2011, and U.S. Secretary of Education Arne Duncan has predicted that a vast majority of schools will not meet their targets in 2012 if the law isn’t changed. In September, Duncan announced that he would entertain waiver requests from states to avoid sanctions for not reaching 100 percent proficiency. But winning such a waiver would require states to adopt teacher evaluation systems that rely, in part, on student results on state tests.

Berliner predicts that the evolving system—which will focus on individual teacher and principal performance instead of schoolwide or districtwide results—will create greater incentives to cheat. “I think the cheating rate will actually go up,” he says.

Careers have been made by making AYP. But some teachers and administrators have cut corners to do so. The Georgia Office of Special Investigators’ report on the Atlanta scandal found that two conditions led to the widespread abuse: unreasonable test score targets set by school administrators and unreasonable pressure on teachers and principals to achieve those targets.

Adding to the pressures are teacher evaluation systems that incorporate student achievement into the formula used to rate teachers. Many of these systems were developed in 2009 and 2010 to obtain federal funds under the Department of Education’s Race to the Top grant competition. One of the winning states, Tennessee, pegged 50 percent of a teacher’s evaluation on student achievement and tightened up guidelines to make it more difficult for young teachers to obtain tenure.

In New York, up to 40 percent of a teacher’s evaluation will be tied to student achievement under regulations due to be released this fall. While he does not condone cheating, Rod Sherman, a high school math teacher at Plattsburgh High School in upstate New York, says there’s fear in the air, with educators worried that low test scores in their classrooms, or their schools, could cost them their source of income. “It’s about jobs—yours and the administrators’ as well,” says Sherman. “It’s going to be a gotcha system, and that’s not what the education community needs.”

Cracking Down on Cheaters
Administrators have responded by stressing enhanced test security and cracking down on teachers and principals involved in testing improprieties. This can be costly. In 2011, the Baltimore City Public Schools spent $400,000 to hire and train 150 monitors during the state’s testing week to make sure the testing materials were not tampered with before or after administration.

“You need to manage for it in terms of culture, which is about getting people to do the right thing, and you need to manage for it technically, by ensuring there are safeguards in place to minimize it,” says Andres Alonso, Baltimore’s chief executive officer. “I don’t think you can get rid of cheating in every single classroom. That’s simply too hard. But we can make sure that organized school-level cheating doesn’t happen.”
For Alonso, managing the culture includes taking a hard-line with those who violate the rules. That includes seeking the revocation of an educator’s license to teach in Maryland. A stern Alonso addressed school staff in 2011 in a video seen throughout the district and posted online in which he set down a zero-tolerance policy for cheaters: “You need to understand that your entire professional life is on the line. We are not talking about termination or a transfer. We are talking about you losing your professional license. There is no safety net out there.”

Fremer says that administrators concerned with cheating within their ranks should focus on three critical indicators: similar results, patterns in erasures, and unusual gains at particular schools (see sidebar “Cheating: What to Look For”). Cheating: What to Look For
John Fremer, a test security expert, says that administrators concerned with cheating within their ranks should focus on three critical indicators:

• Very high levels of similarity in specific test results of a pair or group of students. Fremer uses a rule of thumb of one chance in 100,000 as a standard when all the data is collected. He stresses that one needs to collect data from several tests so there is a high level of confidence that these are chance results. He warns that rare events do occur, but only rarely.
• Very high numbers of erasures, especially from wrong to right. He says students typically don’t make many erasures on standardized tests, so alarms should go off if there are large numbers of erasures from wrong to right. Typically, students have one or two erasures per test. Within a class of 30, there can be as few as 30 erasures. So when there are four or five times that many, it’s reason for concern. On average, Fremer says, students gain slightly when they change answers. But he’s worried teachers may advise students not to make changes lest they be flagged for irregularities. He believes that if there are strict controls on classroom proctoring and what’s done with answer sheets, this problem can be reduced.
• Very unusual gains from one year to the next. Such gains could be too good to be true, so Fremer urges special scrutiny when that occurs. He looks for gains that far exceed the typical pattern. When results are not just very good but unbelievable—going from the 20th percentile to the 80th percentile in one subject in one year—then educators should be skeptical and investigate further. If a high number of erasures and unusual similarities are also involved, there’s a good chance testing irregularities occurred.


That’s what occurred at Hopeville Elementary School in Waterbury, Conn., which experienced higher than expected increases on the Connecticut Mastery Test, which is administered to grades 3 to 8 each year. A testing coordinator at the school reviewed the tests and reported irregularities to the state, which, on its initial review, found what appeared to be “some sense of manipulation of the forms,” says Mark Linabury, spokesman for the state’s department of education. The Waterbury Republican-American reported that teachers were told by their principal to tell students to “check their work” when they noticed wrong answers while students were taking the test.

Linabury says the fact that the testing coordinator flagged the irregularities showed that the partnership on test security between the state and district was robust. The Hopeville school opened in late August, with its staff still on administrative leave while the state continued its investigation. “We’ve pulled in substitute teachers to handle the classes there,” says Waterbury Public Schools spokeswoman Nancy Vaughan. “We’ll just have to wait until the investigation is completed.”

Enhancing Test Security
In other cities and states, educators are tightening test security. In the past, teachers at New York schools would correct essay questions from students at the grade level they taught, and those students could have been their own. In 2012, New York teachers will no longer be able to correct the essay questions on English and social studies tests, since their evaluations will be tied to testing results. In September 2011, the state Board of Regents issued security guidelines that require state tests to now be administered on the same day statewide. Also, educators who proctor or correct state exams must now certify that they have been properly trained and have followed security protocols.

State officials in Georgia have responded to the scandal in the Atlanta Public Schools, which revealed a systemic disregard for the security protocols in place, by strengthening those protocols in 2011, with principals—in addition to district superintendents—now required to certify that the testing protocols were followed and, if security was breached, that a report was made.

Those protocols include making sure that tests—sealed in plastic shrink-wrap—are not opened until the morning of the test and are then secured immediately after the tests are completed. Under old regulations, teachers were allowed to review the answer documents to make sure there were no stray marks that could result in wrong answers when graded electronically. That, too, was changed following the recent scandal.

“We are walking a fine line because a majority of our educators are following the protocols,” says Melissa Fincher, the state’s associate superintendent for assessment and accountability. “But that language was taken way out of context and used inappropriately. So we just removed it completely.”

Test Prep: Missing the Point?
Erasing the wrong answers and inserting correct answers is perhaps the most egregious form of cheating by test administrators. But Daniel Koretz, Henry Lee Shattuck Professor of Education at Harvard’s Graduate School of Education, says that inappropriate test preparation is another way that teachers can game the system: boosting test scores without increasing learning among students.

Back in the 1980s, Koretz says, it was considered inappropriate for teachers to use a preceding year’s test as a teaching tool to prepare students for an upcoming exam. Today, it’s a central part of test preparation carried out in school districts across the nation. He says that it is also common practice for teachers to learn from test preparation companies what standards will be stressed during a certain year’s testing regimen.

Students also are taught tricks to master certain math concepts, such as the Pythagorean Theorem, which calculates the measurements for a right triangle and is characterized by the formula a2 + b2 = c2. Since there are very few sequences using round numbers that fit the sequence, students are taught that when such a question comes up, the sequence of 3, 4, 5, or some multiple of those numbers, will most likely provide the correct answer.

“That might get you the right answer on the test, but that’s not going to help you in the real world,” says Koretz. “If you hire a carpenter and he has to figure out an angle on your roof, you’re in trouble. But it’s everywhere. Everybody is cutting corners. We’ve gotten so hung up on raising test scores that we’ve lost the view that the whole point of the tests is to improve instruction.”

David McKay Wilson is a freelance journalist based in New York State.