Now, here's a classic headline about testing misbehavior: "Cheating or misunderstandings?"
Nine Arizona school districts have invalidated portions of their spring standardized test scores because teachers gave students extra time to complete parts of the test or read sections of the test to students.
In another case, state education officials are trying to salvage test scores of hundreds of students in Phoenix's Creighton Elementary, Gilbert Unified and Yuma Elementary districts, where teachers allowed students to take two, even three days, to finish essays for the most recent AIMS test. Directions clearly state that the essay must be written in one sitting. The state has asked the three districts to pay for a study to determine if the extra days gave students an unfair advantage.
I would be very surprised if it didn't, although if the kids had truly poor writing skills, extra time might not have helped them that much (but they certainly could have used their nights to do some revising and research, if needed). How this could possibly be a misunderstanding, though, I can't imagine, despite the testing coordinator's insistence that this mistake would have been very easy to make.
Granted, it's not outright deception, as in the case of one hapless principal:
The highest-profile case involved Maureen Booth, who was principal at Sequoya Elementary School in the Scottsdale district. District officials claimed Booth changed Stanford 9 scores so her teachers could receive incentive pay. Booth denied any wrongdoing, and the district dropped the allegations in December in exchange for her resignation. The state's case against her remains open.
Regardless, when "misunderstandings" are this widespread (and isn't it interesting that such misunderstandings almost always happen in a way that can only benefit the student), something is wrong.
Alice Finn Gartell is the attorney for the Arizona Education Association, which represents about half of Arizona's 46,000 teachers. Gartell said cheating allegations fall into three categories: intentional cheating; sloppy or accidental cheating, such as misunderstanding directions; and teachers who think it's acceptable to help a little, such as pointing to a question and asking a student, "Have you checked on these numbers?"
Ten years ago, Gartell said, there were no such allegations, but in the past three years she has averaged five to eight cases a year.
"There is a lot of pressure on teachers to improve test scores, and many more different tests," Gartell said. "It's difficult emotionally on most teachers, but very few cheat."
Arizona State University researcher David Berliner estimates that while outright cheating is rare, stretching the rules is far more common. He estimates that about 15 percent of the time, teachers allow students extra time or teach something that will help while the kids are taking the test.
"It is breaking the procedures in such a way that students score better and thus it compromises the validity of the test," said Berliner, former dean of ASU's College of Education.
Exactly. This is why psychometricians like myself get so bent out of shape about even small instances of cheating. We spend a lot of time fighting the notion that such tests are by definition invalid; when teachers intentionally or carelessly invalidate tests through this kind of behavior, no one wins. Because there is such a strong emphasis on testing these days, teachers should recieve a thorough training that doesn't leave wiggle room. If anything, I think the situation is not "high-stakes" enough - at least, not enough to convince teachers that these kinds of "misunderstandings" are unacceptable.
Posted by kswygert at July 12, 2004 01:44 PM