Dozens of Mich. Schools Under Suspicion For Cheating
An announcement by state officials that as many as 71 Michigan elementary and middle schools might have cheated on state tests last winter has burgeoned into competing scandals.
Some observers were struck by the scope of the trouble, noting that the number of schools cited for test-result "irregularities" in Michigan is more than double the number of New York City schools accused of cheating in a widely publicized 1999 investigation.
On the other hand, many local educators said the real scandal lies in how state officials handled the evidence of possible cheating. They angrily criticized the officials for releasing information on June 7 that implicated 21 districts without first hearing explanations.
The outcry prompted legislative leaders to call a joint meeting of the education committees last week. The committees, in turn, summoned state Treasurer Douglas B. Roberts, who oversees the test, for questioning. Even Gov. John Engler expressed dismay that an apparent news leak at the Michigan Department of the Treasury had truncated the state's usual process for dealing with suspected testing irregularities. Still, no one was ready last week to dismiss the evidence of possible cheating, least of all state officials.
The problems were found in the written answers students composed to questions on social science, science, and writing tests. Numerous answers did not seem like independent work because they matched too closely, officials said. It was unclear whether the cause of the similarities was teacher help, communication among students, or something else, they noted.
In some cases, clusters of students from the same class appeared to have handed in answers that were virtually identical.
"It's the first time we've had something of this magnitude," said Bridget Medina, a spokeswoman for the state treasury department, which administers the tests through its merit-awards office. At the same time, she said, "we recognize there may be an explanation for some of these similar answers."
The state has given districts until the end of the week to respond to the evidence.
Local and state investigations could mean that scores for the sets of papers deemed to reflect cheating fall to zero, thereby lowering overall scores for the schools and their districts.
In Michigan, as in other states, such tests are growing in prominence, are widely publicized, and in some cases carry significant consequences for students and schools. For example, high scores on one of the tests involved in the investigation, 8th grade science, can earn a middle schooler $500 toward college.
No deadline has yet been set for the state to decide which schools will be penalized, Ms. Medina said.
Of the 70 elementary and middle schools under suspicion for cheating—the total dipped last week because officials discovered that one school had been included by mistake—44 are in Detroit, Michigan's largest district, and the others are scattered throughout the state.
The Michigan Educational Assessment Program, or MEAP, tests in five subjects were given to 4th, 5th, 7th, and 8th graders in January and February. The questionable answers surfaced mostly in the 5th and 8th grade social studies tests, with a smaller number in 5th and 8th grade science and 5th grade writing.
Test graders hired by a North Carolina contractor flagged answers with close similarities. Those answers went first to the state's testing experts for review and then were examined by a nine-member panel of outside educators.
Treasury officials had intended to allow the districts with suspicious answers to review them and respond before making the list public. Instead, they alerted superintendents to the problems before a hastily called press conference on June 7, after learning that a radio news report about the suspected cheating was airing that day.
"As it was, it was a choice between them reading about it in the newspaper or getting a call from us," said Ms. Medina. "It's unfortunate, but I don't think it should overshadow the focus on the test irregularities."
But superintendents called the change in plan more than unfortunate.
"Talk about unethical procedures," said Alan M. Dobrovolec, the superintendent of the 11,000-student Taylor district in suburban Detroit, whose J. Edgar Hoover Junior High School was on the list for its 8th grade social studies test results. "It took us about an hour to figure out what the problem was," the superintendent said. "But the fallout is our reputation has been damaged, whether we are guilty or not."
He explained that the test question at issue in his district asked for a definition of NATO, the North Atlantic Treaty Organization. The school's three 8th grade social studies teachers had guessed that the test would include a question on NATO, and using approved MEAP coaching materials, taught the children a definition that included the phrases "alliance of country" and "prevent the spread of communism." It was those phrases that were flagged by the test graders.
"It's not a case of cheating," Mr. Dobrovolec argued. "It's a case of teaching and learning."
Preparing Too Well?
Similar stories about lessons well learned were repeated last week in Michigan newspapers, as a half-dozen superintendents around the state responded to what they viewed as unfounded accusations.
A spokesman for the 163,000-student Detroit district, Stan Childress, said late last week that the district's investigation was still going on, but that some Detroit schools may have also worked too well with preparation materials.
"Our curriculum experts believe many of these problems are going to be explainable by many of the patterned-response strategies in materials provided by the state," he said.
Daniel M. Koretz, a testing expert with the RAND Corp., a Santa Monica, Calif.-based research organization, called the problems in Michigan "surprising only in scope," and said the controversy could point to a problem that goes beyond cheating in the strict sense.
In preparing students for tests, he said, "there's this whole range of shortcuts that people take that are not frankly cheating, but they are not what we want to see people doing in response to accountability."
Vol. 20, Issue 41, Pages 18,30