On Wednesday, the US Office of Research Integrity released a report finding ex-Harvard professor Marc Hauser guilty of research misconduct. A scandal broke last year when Hauser, a major figure in the field of animal cognition research, was accused of questionable research practices and reprimanded by Harvard. The new report investigates several of Hauser’s studies on primate cognition that were funded by the National Institutes of Health over the last several years.
In experiments involving animal behavior, trials are often videotaped so that subtle behaviors and responses to stimuli can be analyzed more carefully. According to the report, many of the instances of misconduct had to do with the “coding,” or interpretation, of these videotapes. In two cases, Hauser miscoded the videotaped trials, claiming that the subjects responded in ways consistent with the experiments’ hypotheses; when unbiased observers watched the tapes, they found no evidence that the subjects behaved in this way. In another instance, Hauser claimed that three observers independently coded the trials, while in reality, only Hauser did so (then proceeded to fabricate a measure of “inter-observer reliability”).
In one of the most blatant instances of misconduct, it appears that Hauser did not actually conduct the research he claimed to have done, and instead, he completely fabricated data for an entire control group. The report found that in two other cases, Hauser manipulated the results obtained in the study in order to reach statistical significance.
» via ars technica