The prominent journal Science published a study last December that has turned out to be fraudulent. The study purported to show that a single “respectful conversation” could change people’s minds about a subject on which they have strong views. The study claimed that when gay canvassers discussed same-sex marriage, they changed views far more effectively than when straight canvassers conducted the discussions.
It was newsworthy, said the Chronicle of Higher Education recently, because it suggested that “empathy, even among strangers, can override people’s deeply held political beliefs.” The Chronicle also called it a “blockbuster” study.
The study had as authors a respected political scientist, Donald P. Green of Columbia University, along with a UCLA graduate student, Michael LaCour. Last week, Dr. Green asked that the study be retracted.
The fraud was discovered by two graduate students (recently interviewed by the Chronicle). They had tried to replicate the results. When they were unable to confirm that the research had even been conducted, they went to Dr. Green.
Science has issued a statement of concern as it investigates. As for the impact on higher education, there’s much to be said about 1) the willingness of a prominent scientist to accept findings without asking for actual data; 2) the flaws of the peer review process; 3) the enthusiasm of the press for these particular findings…etc.