When Hillary Clinton announced she’d be running for president in 2016, Cheryl Rios, CEO of Go Ape Marketing, said women were “built differently, we have different hormones.” Rios added while the fight for equal rights is “a wonderful thing,” she doesn’t think a woman should be president.

She’s not completely deluded: a 2013 study published in Psychological Science did find single women were more socially liberal, less religious, and more likely to vote for Barack Obama during the ovulatory phase of their menstrual cycle, PsyPost reported. But in a new study published in PLOS ONE, this relationship isn’t as strong. It was, to be exact, “weaker or less reliable than previously thought.”

Researchers from Brunel University London in the UK surveyed 750 women reporting regular menstrual cycles on their moral and political attitudes, as well as their menstrual cycle, health, and lifestyle factors, including but not limited to contraceptive use. Participants’ politics, or conservatism, was measured with Moral Foundations theory, “which proposes the existence of five universal psychological systems upon which moral ideology is based”: care, fairness, in-group loyalty, authority, and purity. The latter three are considered a reliable predictor of political conservatism, with purity specifically predicting “sexual conservatism.”

The results did not replicate those of the 2013 study.

“In our data,” researchers wrote, “we found no evidence of an association between menstrual cycle phase and political conservatism, either as a main effect, in the form of an interaction with relationship status, or with analyses carried out separately on the sample split according to relationship status.”

Researchers added they measured conservatism and framed relevant questions differently than the 2013 study, which may be why they failed to replicate the 2013 study exactly. Though The Raw Story reported additional studies have tried to replicate the 2013 findings and failed to do so as well.

This goes to show how important replication is in science. It's not enough to take a single study's findings at face value, especially when samples and methods vary.

“Replication is … essential for the advancement of scientific knowledge and consequently it would be helpful for future research to continue to address the extent to which previously reported menstrual cycle effects on other behaviors are robust,” researchers concluded.

Source: Scott IM, and Pound N. Menstrual Cycle Phase Does Not Predict Political Conservatism. PLOS ONE. 2015.