Signs of trouble in Maryland's school testing program arrived with the 1995 scores.
Six months after the students had hunkered down for the problem-solving exercises, the test results for a few schools soared while most others improved modestly or not at all.
Officials who had praised the dramatic scores now acknowledge they were too good to be true.
An investigation of the 1995 Maryland School Performance Assessment Program (MSPAP) test has revealed cheating incidents, lapses in monitoring and misunderstood directions.
In the aftermath:
Teachers in Baltimore City and Baltimore County have received reprimands or suspensions without pay.
Union officials and testing specialists have raised concerns about the test and test directions.
Educators are questioning whether discrepancies seen as evidence of cheating are actually the result of old-fashioned teaching methods.
Although officials in the State Department of Education have refused to comment, it appears that the MSPAP is suffering growing pains.
What happened during the 1995 test is not entirely clear because officials refuse to explain how the violations occurred, how the investigation was conducted, or whether the state's testing program has been improved as a result.
Pieced together from public documents and interviews with sources who are familiar with the investigation, this much is known:
When the scores from the May 1995 MSPAP test came out that December, officials in Baltimore and Maryland heralded signs that the still-young statecurriculum and testing reforms had taken root. It was only the third year that the MSPAP was officially used to judge school quality.
"They should bottle what's going on and spread it around the city," Mark Moody, the state's top test score analyst, said of gains made by Baltimore's Tench Tilghman Elementary School. Last week, The Sun reported that scores at eight city schools, including Tilghman, had been lowered by the state as a result of the cheating investigation.
Soon after the scores were released in December 1995, officials received anonymous tips alleging security violations in some high-performing schools around the state.
About the same time, testing officials began to doubt the sudden rise in scores: It appeared unlikely that schools could record giant leaps when incremental growth was more the norm.
It also seemed unlikely that some groups within a school would outperform others by great strides.
By January 1996, the State Department of Education decided to examine results from 104 schools in 17 school systems that had high-performing student groups, according to a state bulletin published last year.
Officials would have to look at the test booklets completed by the children to understand why they scored so well.
In time, they found answers in the test booklets that were identical or nearly identical.
Letter of reprimand
Clearly, students copied each others' work in some cases, officials say.
One such incident occurred at Sandalwood Elementary School in Baltimore County.
Twice, a third-grader apparently copied parts of answers from a classmate -- a total of about four words each time, said Paul Mazza, director of the county schools' Office of Research and Accountability.
The children had completed a group activity. One of the hallmarks of the testing program is group problem-solving: Each student takes notes as his group solves a problem or conducts an experiment, then the student on his own writes answers to the test questions.
County school officials placed a letter of reprimand in the personnel file of the teacher who had administered the test. The letter states that she should be more careful about monitoring test takers and should clearly explain that children need to work on their own when group activity ends.
Teachers countywide were advised during 1996 MSPAP training sessions to be more careful.
County schools spokesman Donald I. Mohler said he thought the reprimand was too strong.
"You try your best as a teacher to make sure students don't have wandering eyes," Mohler said. "I felt bad for the teacher. Certainly there was never any suggestion that she was trying by design to help the kids."
In Baltimore City, there were many similar cases, confirmed Marcia Brown, president of the Baltimore Teachers Union.
Teachers told her that instructions for teachers and test takers did not distinguish clearly when children were to work together and when they were to stop. Teachers also reported that by 1996, the state apparently recognized this problem -- and revised some of the test materials. The state would not comment on this.
In addition to the cases where the state alleged lax monitoring, there was outright cheating on the part of teachers, it said. Officials will not say where this took place.
A state bulletin distributed in March says: "The instances of apparent teacher interference included cases where testing examiners gave answers to students or otherwise guided students' responses to assessment tasks."
