Cheating Can Be Eliminated



By 03/08/2011

Print | NO PDF |

When schools—and potentially teachers—are held accountable, they have an incentive to report more positive results on student performance than actually takes place.  School administrators have been doing that for years, most notably by reporting student dropouts only if they take place during the school year, despite the fact that most students disappear from the schoolhouse between one school year and the next.  For decades we were told that 90 percent of the students were graduating from high school when the actual number was closer to 70 percent.

So it should come as no surprise when we learn from USA Today that cheating on accountability tests seems rampant in the many states they included in their investigation. Their method is very simple:  if student test scores at a school climb by three standard deviations over the course of a year, something bad is likely afoot.

But the real scandal is not the cheating by the teachers or the principals themselves, bad as that may be, but the head-in-the-sand administration by the school district.   Apparently, many school boards and district officials are quite willing to take credit for gains in performance no one in their right mind can possibly believe. Nor is the USA Today story the first cheating story to have appeared.  Freakonomics author Steve Levitt became wealthy in part because he showed how it worked in Chicago.   If both an economist and a newspaper reporting team can find out this information, why can’t those responsible for maintaining the integrity of the school system–public officials–do the same?

(Granted, startling gains in achievement might happen under exceptional circumstances, so one would have to investigate before drawing conclusions, but firemen dispatch their trucks as the alarm bell goes off, they don’t wait to find out whether the fire is serious or not. )

Of course, districts need to have a more serious monitoring system than the one USA Today put into place.  If cheaters know they will be caught if they show gains of 3 standard deviations, they will adjust their strategy accordingly.  To eliminate cheating, one needs, on test day, independent proctors to oversee the administration of the test and hand in the results.

That was standard procedure at Harvard until very recently.  It is dreadful that this sensible policy was abandoned during the university’s recent budget crisis.  Maintaining the integrity of the educational process is vastly more important than maintaining class size, and the costs are trivial by comparison.

-Paul E. Peterson




Sponsored Results
Sponsored by

Harvard Kennedy School Program on Educational Policy and Governance

Sponsored by