By Joshua Benton
Staff Writer
Page 1A
All 699 schools suspected of cheating on the TAKS test will face a state investigation, the Texas Education Agency announced Monday.
Sort of. The word “investigation” can have many meanings.
For some schools, investigations could consist of little more than an exchange of letters. It remains to be seen how thorough investigations of 699 schools would be possible, given constraints of time and staffing.
And state officials still have no plans to seek the additional test data that would make a detailed investigation possible. For example, the state still does not know which students have the most suspicious test answer sheets.
“The task force believes strongly that test integrity is really important, so everybody needs to be investigated,” said Olga Garza, coordinator of the commissioner’s task force on test security.
The schools are the subject of state scrutiny because their 2005 scores on the Texas Assessment of Knowledge and Skills were flagged by a test-security firm, Caveon. The company was hired by the TEA last year to look for schools where teachers or students might be cheating on the TAKS.
The agency has set itself an ambitious timetable. It hopes to select which schools will be in “Phase 1” of its inquiry this week, with investigators heading to their campuses shortly after Labor Day. It hopes to have the meat of those investigations concluded by early October, which will be the next time the test-security task force meets.
Agency officials have not decided how many schools would be in Phase 1, or how many phases there would be, agency spokesman Debbie Graves Ratcliffe said. She said the schools tackled first would probably be the ones that were flagged the most times in Caveon’s analysis – for example, schools that had suspicious scores in multiple subjects and multiple grades.
Caveon looked for schools with suspicious scores. That could mean that some students had unexpectedly large score gains, or that a group of students in particular classrooms had identical answer sheets.
State Education Commissioner Shirley Neeley appointed the five-person test-security task force this month to determine how the state should proceed. The group had its first meeting Monday.
Ms. Garza said all 699 schools on Caveon’s list would be investigated. That’s much larger than many agency employees and observers had expected.
But Ms. Ratcliffe said that, for some, the investigation could be limited to a “test security audit.” That would not involve any on-site visits. Instead, the state would check to see whether the school was aware of state test administration policy and had had any violations in recent years.
In some cases, Ms. Garza said, districts might have a good explanation for why a school made the Caveon list. “Maybe some of the districts will be cleared, if there’s a statistical aberration that’s easily explained,” she said. In that case, it might take only a letter from district officials.
Full inquiry for some
Other schools will get a full-scale, on-site investigation. But if their number is large, it is difficult to imagine how the TEA would pull off operations on such a scale.
The agency is hiring more investigators, which would raise their number from five to 15. But even that number could be quickly tied up by just a few significant investigations. Traditionally, on-site investigations have involved sending two or three investigators to a school to conduct interviews with school staff and students.
“We’ll do what we can with what we have,” Ms. Ratcliffe said, adding that the TEA would consider subcontracting some of the work of investigating out to nonagency employees if necessary.
And those investigations take time. The largest in recent memory was in Wilmer-Hutchins, the Dallas County school district that has since closed. It took four months to complete and occupied most of the state’s test-security staff during that time.
Students move on
In addition, because of the 18-month delay in investigating, many students who might have cheated have already graduated and could be difficult to find.
At the October meeting, Ms. Garza said, the task force hopes to have a series of recommendations on how to improve test security – in time to administer portions of the TAKS that month.
Perhaps most important, the state still has not obtained more detailed information from Caveon about what the firm found unusual about each school’s test scores. At this point, almost a year after drafts of Caveon’s report reached the agency, the TEA only knows which broad category of suspicious behavior the company found in a given school or classroom.
Caveon has much more data available. It would detail, for example, which students in a given classroom were suspected of copying answers off of which of their peers. It would outline how many students in a given classroom had suspiciously large score gains, and which ones had unusually high numbers of wrong answers erased and replaced with correct ones.
Caveon officials have said they could make the information available to the TEA, which would seem important to any thorough investigation. But the agency still has not sought it. “We’re using the information we have now,” Ms. Ratcliffe said.