Replication of a famous Stanley Milgrim experiment in social psychology, which turned out to be far harder for the investigators than the subjects. (New York Times )