Posted on March 11, 2011

When Test Scores Don’t Add Up: 32 Metro Detroit Schools Show Improvements Too Good To Be True

Peggy Walsh-Sarnecki, Chastity Pratt Dawsey and Kristi Tanner-White, Detroit Free Press, March 6, 2011

Each year, millions of children in Michigan and across the nation take state standardized tests that impact everything from a school’s reputation to how teachers will be evaluated to whether schools will even survive.

The pressures to perform, experts say, tempt some school administrators and teachers to cheat.

The Free Press, as part of a” target=”_blank”>nationwide investigation with USA TODAY and other partners, analyzed millions of test score results and found that 34 schools across Michigan–32 of them in metro Detroit–showed test score gains over a one-year period that experts say are statistically improbable. More broadly, the analysis found 304 schools in six states and the District of Columbia that had test scores so improbable, they should be investigated. Besides Michigan, the states were Arizona, Colorado, California, Florida and Ohio.


Education reformers caution against jumping to the conclusion that a surge in scores reflects cheating. But experts in testing procedures say the increases exceed what can be reasonably expected.

The pressure to raise test scores is enormous, particularly among schools that fail to meet the requirements of the federal No Child Left Behind law. And with that pressure, the incentive to help students cheat also is growing, the experts say.


But high leaps and other indicators also signal that the scores might be too good to be true, he and other experts said.


“Student reports consistently showed a pattern of improper assistance by school staff in answering MEAP test items,” the state investigation concluded.

The Michigan Department of Education ordered the school to tighten its testing procedures and be subject to close monitoring by the Office of Educational Assessment and Accountability.


In its own report, state officials noted that Crofoot fourth-graders taking the math test went from being 39% proficient in 2003 to 87% proficient in 2004, and 100% proficient in 2005. Fourth-graders taking the reading test went from being 52% proficient in 2003 to 84% proficient in 2004 to 89% in 2005.

Such massive improvements in test scores–while not conclusive proof that test tampering has taken place–are statistically improbable and should be investigated, experts say.


Yet wild gains in test scores don’t trigger investigations of cheating in Michigan. The state opens up investigations only if a credible allegation is raised. Like most states, Michigan leaves the implementation of guidelines for administering the MEAP up to the school districts.

Fremer said the idea of investigating high scores seems ridiculous to most officials. The response often is: “You want me to investigate when scores are going up? What’s the matter with you?”


An analysis of Michigan scores from 2007 to 2009 found 83 schools statewide with standard deviations of three or more. When focused more narrowly on 2008 to 2009, it found 34 such schools–32 of them in metro Detroit.


In at least 10 states, teacher pay already is–or is about to be–linked to test scores. Several school districts in Michigan have tied teacher salaries to test scores, too. Under a 2009 state law, school districts have until September to develop plans to make student test scores a significant part of teacher evaluations. Michigan also plans to raise the bar for what is considered proficient on the MEAP, saying schools aren’t adequately preparing students for college.

The federal No Child Left Behind law penalizes schools that fail to make “adequate yearly progress,” often leading to staff being reassigned or schools being closed. Most of the 130 schools in Detroit Public Schools that have closed since 2005 were cited for having low test scores.


Across the nation, incidents of testing irregularities abound–cheating has been found in Delaware, Illinois, Maryland, Massachusetts, Nevada, New York and South Carolina, according to FairTest, a Massachusetts nonprofit that aims to prevent the misuse of standardized tests.

In Georgia last year, the state investigated 191 schools, including many in Atlanta, following an Atlanta Journal-Constitution report identifying schools with unusual gains. The state has subsequently begun collecting erasure data–the number of times an answer is erased on a test and the number of times it has been changed from wrong to right–to trigger investigations.

Failing to investigate questionable scores could mask students’ real gains and academic needs, said John Tanner, executive director of San Antonio-based Test Sense, an organization that helps schools make use of test scores.