Lady Mohawk Preview
Christmas in Springboro
Santa arrives at the Dayton mall
Dayton Daily News education reporter Scott Elliott writes about schools, kids, teaching and learning.
Home > Blogs > Get on the Bus > Archives > 2006 > September > 04 > Entry
Ohio, testing and cheating
Last week when I wrote about Texas’ plan for a huge expansion of its program to monitor cheating on state tests, a couple friends were quick to E-mail me with this reminder — it never would have happened without the Dallas Morning News.
The News’ education team wrote relentlessly in 2004 and 2005 about inconsistencies in the state’s test results that strongly suggested some schools were cheating. Unfortunately, the Morning News’ stories are no longer available online.
But what they found in analyzing school-by-school results were examples of schools that, for instance, were among the best in the state on reading but among the worst on math. Or schools where fourth graders ranked among the worst but the next year in fifth grade the same kids at the same school ranked among the best. This led them to teachers, students and administrators who had amazing tales of teachers who coached their classes through the exams and other cheating horrors.
So the question I asked last week was a simple one — if this sort of thing happens in Texas, what are the chances it’s also happening here in Ohio, another big state with a high profile testing program?
Last week, I got the chance to ask Mitchell Chester, who heads the state’s testing program for the Ohio Department of Education.
Chester said the state is adding new accountability checks, but he also acknowledged that in recent years the only questionable incidents the state has handled were all self-reported by school districts or the schools themselves. In other words, the state has not independently caught anyone cheating lately.
But, according to Chester, Ohio has asked its testing contractors to perform analyses of score patterns that may reveal the sorts of near-impossible trends that could indicate cheating. For instance, computers can raise flags when several students in the same class mark the same answers in large chunks. There’s a discussion of this method in the book Freakonomics looking at how it caught teachers giving kids answers in Chicago.
Ohio has also asked for an “erasure study.” This is a fascinating approach in which computers that can read erasure marks count how often answers were changed from wrong to right. This is another way cheating schools can be caught — sometimes teachers or administrators take score sheets and erase chunks of answers to replace them with the correct answers. Logic would tell you the odds are not good of a class full of students all erasing a the same group of answers and replacing them with the correct responses.
So it should be interesting to see if these new approaches lead to any bombshells that expose cheaters in our state.
Permalink | Comments (3) | Categories: Testing
Comments
By DPS PARENT
September 8, 2006 7:38 PM | Link to this
Why did Joe Lacey run for the board again??? Oh Yeah to help his political career. Joe YOU WON, YOU ARE A BOARD MEMBER NOW!!! ACT LIKE one.. investigate it follow up with Dr. Mack Do soething other than pick petty battles with Littlejohn and others.By Joe Lacey
September 5, 2006 12:32 PM | Link to this
A DPS teacher told me about witnessing someone changing answers on an Ohio Achievement Test but that he/she would not go on record as the whistleblower. I emailed the ODE just a general summary of what the teacher had told me and someone from ODE called me. After a brief discussion of what the teacher told me he/she witnessed, the person from ODE said that they wouldn’tinvestigate any further unless the teacher was willing to go on record. They also said that the district would be informed who the whistle-blower was if the teacher decided to talk to the ODE. In general the ODE didn’t seem very anxious to investigate the matter.By Rick
September 5, 2006 7:21 AM | Link to this
It is good that this is getting addressed. The testing system is not perfect and constant improvement is required.