Get on the Bus

About Get on the Bus

Dayton Daily News education reporter Scott Elliott writes about schools, kids, teaching and learning.

Quick news updates by e-mail

Start your workday informed by signing up for our e-mail local news headlines and breaking news alerts. Sign up

August 2008

RSS Syndication

XML Get on the Bus
XML More 'DDN' feeds

More recent entries

• DPS attendance very low at some high schools
• School begins smoothly in Dayton
• Sewing champ helps police the prom
• Dayton teachers approve contract, raise
• Stanic: It's about learning and the levy this year
• Dayton schools name new deputy superintendent
• Here's your chance to give feedback
• Will they arm wrestle over math and science teachers?
• Dog? Check! Pony? Check!
• Strickland: Give me feed back on education reform ideas
• More entries...

Archives

• August 11, 2008
• August 9, 2008
• August 7, 2008
• August 6, 2008
• August 5, 2008
• August 4, 2008
• August 3, 2008
• More archives...

Home > Blogs > Get on the Bus > Archives > 2007 > February > 04 > Entry

City Day: Practice test raises questions

By selliott | Sunday, February 4, 2007, 12:04 AM

By Scott Elliott

Staff Writer

DAYTON — Last March, fifth graders at City Day Community School knew which unit of measurement Bob used to cover a floor with carpet.

Sixth graders knew that 6 — not 2, 3, or 12 — was the largest number that divided both 12 and 30 evenly.

And seventh graders knew that the square Coral drew was a rhombus because it had four sides.

The question is, how did they know?

In each case the students had seen these questions — or questions that were nearly identical — on practice tests they took the week before the school administered the Ohio Achievement Tests.

In all, a Dayton Daily News investigation found 44 questions on practice tests taken by City Day students that were identical or substantially the same as questions that appeared on the actual state exam they took just days later.

And when state report cards came out last year, huge gains in the percentage of students who passed the test helped propel the chronically underperforming school out of the state’s lowest rating category of “academic emergency.”

Ohio has specific procedures for the handling of state achievement tests, but trusts each school to maintain the security of the exams. Practice questions are available on the Ohio Department of Education Web site, but they aren’t drawn from the actual tests, and in many cases don’t closely resemble the real test questions.

City Day Superintendent Roseda Goff said a consultant hired by the school used traditional testing preparation materials to drill the students prior to the test, and that no state procedures were violated.

But Carl Robinson, who joined the charter school as principal last summer, said he questioned the scores, particularly after he had assessments done on some of the students who did well on the tests. Robinson said he was fired in November, one day after approaching Goff about his concerns.

The test results seemed too good to be true, Robinson said.

The sixth-grade class at City Day, 318 S. Main St., didn’t have a single student pass the math portion of the Ohio Achievement Test in 2004-05. A year later, 100 percent of its seventh graders passed the state’s seventh-grade math test.

Similarly, no one in the school’s 2004-05 fourth-grade class passed the state math test that year, but a year later 59 percent of the fifth-grade class passed the state math exam. It now appears that those too-good-to-be-true results were just that.

“I said wait a minute,” said Robinson. “There is no way in the world to yield that kind of jump by doing what she explained to me. I thought, ‘did she have somebody come in and perform magic?’ It didn’t make any sense.”

The school did nothing wrong, according to Goff. A consultant she hired helped drill the students using traditional test prep materials, she said.

But the consultant, Rachel Armour, told the Daily News that all the materials she used to create the practice tests were provided to her by Goff. The newspaper obtained copies of practice tests she made for City Day and compared them to the actual state exams. In some cases, only names in the word problems or small details are different from the actual questions.

Robinson said an Ohio Department of Education investigator has interviewed him about the testing procedures at City Day. A spokesman for the board would not confirm that an investigation is underway, but Greg Cizek, author of a book on standardized test cheating and member of Ohio’s testing advisory panel, said, “f you have identical questions from a test that was not to be disclosed on a practice test, that’s against every testing regulation the Ohio Department of Education has.”

Robinson, who joined the charter school as principal last September, said he reviewed the practice materials used last year in hopes of recreating the impressive gains in this year’s tests. But Robinson, who has a doctorate in education from Miami University and left a university teaching job in North Carolina to take the principal job at City Day, said test preparation couldn’t account for the improvement made by the students last March.

Consider seventh grade math. In 2005-06, City Day’s seventh grade had an average score on the math test that was about 22 points above the average score for all Ohio schools. The prior year, the school’s sixth graders had an average score that was 45 points below the state average.

“If it’s largely the same kids, that’s a huge change from 45 points below to 21 points above the mean,” Cizek said.

Of the 14 City Day seventh graders who took the state exam in 2005-06, about 28 percent scored accelerated or advanced, the two highest of five scoring levels. The prior year, none of the 13 sixth graders tested even passed the test.

The fifth graders also made sizable gains from the previous year’s scores.

Of the 24 fifth graders who took the state exam in 2005-06, almost 17 percent posted accelerated scores. The prior year, none of the school’s 22 fourth graders passed the test. In 2004-05, the fourth-grade class at City Day was 54 points below the state average in math. A year later, the fifth graders were just seven points below the average for all Ohio students.

The Ohio Achievement Test in math last year was taken by every Ohio public school student — which includes charter school kids — in grades three through eight.

Overall, City Day’s 2005-06 test success improved so much the school shook off two years in the state’s lowest category of “academic emergency,” jumping two spots on the rating scale to “continuous improvement.” City Day, a K-8 charter school that opened in 1998, had consistently ranked among the worst performing schools in Dayton for the percent of students passing the state tests.

After he was hired, Robinson said he became suspicious enough of the results to ask a few teachers to conduct assessments of what their kids knew. After that, he said, he had no doubts.

“There was a blatant discrepancy in terms of what they (the students) were able to do,” he said.

Robinson later obtained a compact disc containing four practice tests created by Armour. He said a week after he was fired an anonymous caller asked to meet him in a parking lot. A car pulled up and a woman handed him the CD, he said.

A state investigator has asked him for the CD, Robinson said.

Armour, a Dayton-based educational consultant, said she was brought in for eight weeks last year to help prepare City Day kids to take the state exams. She acknowledges creating practice tests that included questions strikingly similar to those on the state exams. But she said she merely reworked and retyped what she thought were practice questions the Ohio Department of Education provides on its Web site as test preparation.

“Mrs. Goff told me they were downloaded from the Internet,” Armour said. “It just looked like it was something printed from a digital laser printer. I’m baffled by the whole ordeal. No principal has ever given me any type of secure test materials.”

Goff denied giving Armour real test questions.

“The test I gave her was a practice test,” Goff said. “Our (state) tests are kept under lock and key. They are taped, sealed and locked up. I don’t know how anybody could get anything.”

Ohio has very specific procedures for the handling of state achievement tests, but in the end the education department depends on school personnel at each site to maintain the security of the exams. Sealed, numbered copies of the exams must be delivered to each school no more than seven days before testing is scheduled to begin.

The label information on the computer files that contain Armour’s practice tests show they were created beginning at 6:35 p.m. on Feb. 27, 2006, seven days prior to the scheduled test at City Day. Other files from the following morning show they were created at 3:02 a.m., 3:21 a.m. and 5:03 a.m. Those documents have a copyright warning identifying Armour as the author.

Armour acknowledged she worked through the night after she said Goff gave her the sample questions. She said Goff told her she wanted the students to begin practicing for the tests the next day.

“That’s just how I work,” she said. “Mrs. Goff put a lot of pressure on me. I’m just a diligent person.”

Goff insists the security of the state tests wasn’t compromised. The tests were sent to the school and addressed to Juanita Barnett, the school’s part-time psychologist who also serves as its testing coordinator, she said.

Goff said she didn’t recall if she or Barnett received the tests, but in either case they would have been secured in a locked room. Barnett then typically locks the tests in a cabinet, she said. Only Goff and Barnett have keys to the room, Goff said, and only Barnett has a key to the cabinet. Goff said at one time there were two keys to the cabinet but Goff lost her key prior to February 2006.

“There is no way I could have given (Armour) any test,” Goff said. “I had not looked at a test. And I don’t see how she could have gotten it.”

Barnett declined comment. Armour’s attorney, David Williams, said his client is willing to submit to a lie detector test to support the truthfulness of her story.

Cizek said any test scores of a student who saw questions ahead of taking the test would be meaningless.

“The Ohio Department of Education has strict guidelines on what is and is not appropriate to do,” he said. “It’s flat out illegal to take secure test questions and expose them to students before the test.”

Armour, a Chicago native and Wilberforce graduate, is a former participant in the prestigious Teach for America program. She also served as an intern at the Dayton Daily News in 2003. Armour said she developed her own writing instruction program that she has taught in Illinois and Michigan, as well as in several local schools.

Armour said part of her approach to teaching urban kids is to put problems into a language they can understand. She said she commonly writes problems and questions for kids using brand names they know, such as Bad Boy Records, and proper names that may be familiar to them from popular culture, such as actress Raven Simone or NBA star LeBron James.

“What I do is take a problem and alter and change it so it reflects something the kids are interested in,” she said.

Robinson said he struggled throughout the fall to create a curriculum for the school, where he said books and materials were in short supply. He said his dismissal resulted from the questions he raised about the school’s test preparation procedures.

“I wasn’t going to stop asking questions,” he said.

Goff declined to discuss Robinson’s departure.

Investigations into cheating allegations on state tests are extremely rare in Ohio. Mitch Chester, who heads Ohio’s state testing program, told the Daily News in September that the state’s only cases of testing irregularities in the past five years came after they were reported by the schools themselves. The state does have the power to revoke an educator’s teaching license for violating the security of the tests. The normal process for an investigation of testing irregularities is for the school to first conduct its own internal probe and report its findings to the state, which decides if more investigation is needed.

As a charter school, City Day is sponsored by Education Resource Consultants of Ohio, a Cincinnati-based charter school authorizer with primary responsibility for oversight of the school. In December, ERCO Assistant Director Aaron Kinnebrew said he was not aware of any problems with state test administration last year.

“They are allowed to give practice tests,” he said.

Kinnebrew did not return repeated phone calls seeking further comments.

J.C. Benton, education department spokesman, said the state put in place a policy this school year not to discuss test security investigations, or even acknowledge when an investigation is underway.

Cizek said states like Ohio could discourage cheating by claiming copyright on all test materials.

“It seems to me that at least one way that the state should proceed on this matter is to sue the individuals involved for infringement of copyright,” he said. “I’m no attorney, but it seems like this is a pretty clear-cut case.”

Permalink | Comments (10) | Categories: Charter Schools and School Choice, City Day Investigation, My Favorite DDN Stories, Testing