Sunday, 11 January 2004

Rule One: no falsifying data

Kevin of Wizbang links an AP article that indicates some kids at CSU-Stanislaus faked survey data that helped get Scott Peterson’s trial moved out of Modesto. The Modesto Bee has more coverage here and here, that suggests the survey was based on non-random samples, did not receive IRB approval before it was conducted, and lacked effective supervision—all of which are serious no-nos for valid research. (The Bee also has a copy of the questionnaire on its website.)

The whole situation is embarrassing—not just to CSU-Stanislaus, but to anyone who takes survey research seriously. And while there can be pedagogical value to having students work on surveys, particularly in public opinion classes, there’s no excuse for the apparent lack of supervision in this case.

Update: Eugene Volokh has more. To echo Eugene, however, I will say the lack of supervision and other problems in no way absolve the students who falsified data from responsibility for their corrupt behavior. If the allegations are true, everyone involved should be frogmarched in front of disciplinary committees—the students who faked data and the professor.