Email
Print
FeedScrubbed Out
State auditor Dave Yost breaks down his office’s investigation into school attendance record tampering, and explains what’s put Columbus City Schools into a class of its own.
Illustration by Michaela Schuett
In a series of stories first published in June, Columbus Dispatch reporters uncovered evidence that Columbus City Schools employees manipulated attendance data, purposely withdrawing and reenrolling students with poor attendance or test scores, possibly to boost state report card standings.
It’s a simple equation of subtraction and addition that can yield big results, because the test scores of students not continuously enrolled in the district from October through testing in the spring don’t count toward the district’s passing rate. Students are omitted to erase test scores that could drag down districts’ scores.
Columbus school officials have stridently disputed the accusation, saying no students were specifically targeted for removal just to better the district’s grade. But in late December, the district’s internal auditor released a report detailing student attendance data had been changed, resulting in a bump to the district’s report card.
An inquiry by state auditor Dave Yost into school districts across the state found more evidence of data-scrubbing in Columbus schools than in any other district. But when that investigation wraps up later this month, questions about Columbus schools will remain. In November, Yost’s team broke Columbus City Schools off into its own investigation after finding evidence that data other than attendance records had been altered. Investigators are now looking into evidence that failing grades were changed to passing grades.
“Criminal charges have always been a possibility,” Yost says, referring to Columbus school officials, “and it’s become increasingly evident we will have to make a criminal referral to the county prosecutor.”
In two interviews with Columbus Monthly, Yost spoke at length about the investigation and outlined recommendations his office is making to the Ohio Department of Education as a direct result of his office’s findings.
Q. The investigation was divided into three phases. Why investigate schools in waves?
|
A. With more than 3,400 school buildings in the state, Yost wanted to establish a guide that would narrow the pool of schools to audit. This way, investigators could spend more time looking into schools with a higher risk of scrubbing, Yost says. The first 100 schools were selected because they had the highest percentage of students whose test scores did not count toward the school’s report card rating. In October, phase two examined 81 schools with levies on the fast-approaching Nov. 6 ballot. Results combined from the first two phases gave investigators the data they needed to confidently predict likely offenders, Yost says. Statistical correlations applied to every school in the state pointed to 174 schools with a high risk of scrubbing (70 of which had already been identified in the first part of the investigation). The third phase set out to take a look at the remaining 104 schools. |
|
Q. How many Columbus schools showed evidence of scrubbing? |
|
A. Of the 174 schools identified as having a higher risk of scrubbing—whether there was intent to skew the districts’ state grades or not—50 are Columbus City Schools. “[Columbus is] not the only offender, but it’s the biggest,” Yost says. |
|
Q. Five correlations were used to identify schools with a likelihood of scrubbing. What was one you used? |
|
A. “One of the high correlations is what we call the double asterisks in place of a withdrawal code,” Yost explains. Its use can be legitimate—as in marking a student who has withdrawn from one district and enrolled in another, leaving a paper trail investigators can follow. “On the other hand,” Yost says, “if we see a withdrawal and reentry, which is going to give us a double asterisks code, and we can link up the school building number, we know they were withdrawn from the school and then reenrolled in the same school.” That revolving-door effect can be justified, Yost continues, but the coding gives investigators good reason to take a closer look. |
|
Q. Your office made several recommendations to ODE in the first portion of the probe. What is one flaw you see in the state education system? |
|
A. Count Week, Yost says, referring to the one week in October when school attendance is taken and used to determine funding for the entire year. “Instead of counting them one time in October and then giving people the same money for a whole year, let’s base disbursements weekly or monthly based on attendance as it goes through the year,” Yost suggests. “And at that point you are giving schools incentive not to just get kids there on Count Week.” The auditor’s office, in its initial report to the state education department, encouraged creation of a system of independent checks on attendance data. Yost also says administrators should not be able to see projected report card ratings before submitting final data (the projection is visible to administrators now via ODE’s secure data center). “That’s amazing,” Yost says. “I mean, it gives you a roadmap on how many kids you have to throw out of your averages to get from a D to a C. Don’t show them that. That’s like the teacher passing out the final exam and attaching her grading key.” |
|
Q. Are there incentives in place to deter scrubbing? |
|
A. “You have to remember most of the schools we looked at, in spite of the incentives to do things wrong, still did them right,” Yost says. “That being said, you still want to eliminate the incentives to act badly. Punishing people who do wrong provides incentive not to do wrong. And aligning the dollars so that they follow actual attendance rather than sample attendance are all things we should do to get rid of the incentives to act badly.” |
|
Q. Columbus has since been pulled out of the main investigation. Will results be released separately from the third phase? |
|
A. “Columbus is going to be dealt with to the side. What we are seeing there seems to be a different kind and quantity,” Yost says, explaining his office will likely refer criminal charges to the county prosecutor. The 40 Columbus schools yet to be investigated for scrubbing in the third phase have been separated out, leaving 64 school buildings statewide to be tested for scrubbing. |
|
Q. Will Columbus be the only district where charges are recommended? |
|
A. While it’s not accurate to say Columbus is the only Ohio district that scrubbed data with the intent to boost standings with the state, it is the only one where such evidence has been found, Yost says. “In Columbus, you’ve got a few people who have researched, ‘How do we get some of these low scores out of our system?’ They know which kids and they are specifically communicating, take this one, take that one. You are actually picking and choosing. That’s easier to prove.” |
|
Q. Columbus City Schools has claimed that if they had scrubbed students, more schools would have moved up the ranks. Will the final data show schools earned higher ODE rankings? |
|
A. “We are going to refer a list to ODE saying we found evidence of improper dis-enrollments and you need to recalculate the grades on these schools,” Yost explains. “The problem with the argument [Columbus City Schools] is making is they are basically saying because I didn’t do a very good job cheating, I must not have cheated. The fact that it didn’t change any grades doesn’t mean that they didn’t cheat. It just means they weren’t good at it.” |
How it’s Unraveled
June 2012: The Columbus Dispatch reports widespread instances of altering attendance records in Columbus City Schools.
June 15: The state auditor’s office launches an investigation into Columbus City Schools.
Late July: The investigation is expanded statewide after the auditor’s office finds what appear to be systemic attendance alterations.
August: The Columbus school board hires law firm Porter Wright Morris & Arthur to launch an investigation into the attendance probe parallel to that of the state auditor.
Sept. 20: Columbus Superintendent Gene Harris announces she will retire in July 2013.
Oct. 4: The first part of the state auditor’s investigation is released, finding evidence of scrubbing at all 10 Columbus schools it had looked into thus far. Questionable practices were also found in four other districts throughout the state.
Oct. 23: The second phase of Yost’s investigation, focusing on districts with proposed November levies, clears an additional 47 districts of scrubbing suspicions. Investigators now have the control data necessary to identify 104 schools for the final phase of the query.
November: Evidence surfaces alleging student grades were changed by administrators. Columbus schools are broken out into their own investigation, and criminal charges could result.
Dec. 20: A report from Columbus City Schools’ internal auditor shows students who had never left the district had been withdrawn and reenrolled, improving the district’s report card standings.
