Sigmund Freud

You may have read headlines last fall along the lines of "Have you got a hot mom? Chances are your wife will be a looker, too." That one was courtesy of New Scientist, but other media outlets, including the BBC and the Guardian, also covered a study in Proceedings of the Royal Society B that seemed to provide evidence for the Oedipus complex described by Sigmund Freud and expanded by his protégés: Men pick women who look like their mothers, and women pick men who resemble their fathers.

But Freud may have called the findings by Tamás Bereczkei and two co-authors at the University of Pécs in Hungary a parapraxis—aka the dreaded Freudian slip, a minor error in speech or action that belies one's subconscious wishes or proclivities. Earlier this month, the journal retracted the study after another scientist questioned it.

Here's what happened: After publication of the paper, Markus Rantala, a behavioral ecologist at the University of Turku in Finland, thought that the statistical relationships were suspiciously strong and noticed additional discrepancies in the results. For instance, Bereczkei's team had reported a correlation of 92.8 percent between the relative jaw width of a man's mother and that of their mates. That would mean the jaws of the mothers and the wives were practically identical.

Rantala thought that and other values were unrealistically high, and brought his concerns to the journal's attention. After reviewing Rantala's allegations, the journal appointed an independent expert who concluded last month that the paper contained "factual errors," along with incorrect analysis, "which would have altered the conclusions." Bereczkei and his co-authors agreed with the journal and retracted the study.

Bereczkei chalked up the errors to carelessness. "We worked so rapidly and were enthusiastic and wanted to publish the paper too quickly," he tells ScientificAmerican.com. "We were not cautious enough to check our database."

Rantala takes a harsher view: "It seems that the authors have fabricated data to support a Freudian view of psychology, because there is no other evidence to support it." He admits, however, it is not possible to know whether the errors were intentional or were the result of an unconscious bias in favor of Bereczkei's belief that an individual's early exposure to his or her parents' faces influences future mating preferences. Bereczkei insists that his group did not manipulate the data, and the errors were purely accidental.

Bereczkei says he has been humbled by the experience but remains "convinced that our results are good and our results support our theory." He says that "mistypes" in three data points are enough to account for the error in the jaw measurements, and the true correlation is really closer to 70 percent, still high by biological standards.

"I thought it was a really impressive study," says Lynda Boothroyd, an evolutionary psychologist at Durham University in England, calling the retraction a "shame." Boothroyd points out that there is still some evidence to support the Freudian view of mating preferences. In 2007 she found that daughters with strong relationships to their fathers tend to pick mates that have similar facial measurements. Those correlations, however, were between 20 and 40 percent.