A customer exits a store displaying a campaign sign for Roy Moore, the Republican candidate for U.S. Senate from Alabama, in Centre, Ala., on Nov. 27. (Daniel Acker/Bloomberg)

The last month of the Senate race in Alabama focused on an unexpected subject: Accusations first reported by The Washington Post that Republican candidate Roy Moore had pursued sexual relationships with teens as young as 14 back when he was in his 30s. You knew this, of course; to non-Alabamians it was probably the one thing that someone only half-paying attention to the race had heard about.

Those who actually came out to vote in the state were mixed on the importance of those accusations. Slightly more than half of those who voted believed the women who leveled the accusations — eventually more than half a dozen of them. There was a split by gender, though: Women mostly believed the stories while men mostly didn’t.

A similar split occurred when voters were asked how important the allegations were to their vote. Most respondents said that they were not important or only slightly important — but among women the margin was much narrower.

The exit poll asked another question, too, an unusual one for an exit poll: Do you have children under 18 living in your household? About a third of respondents did — and within that group something interesting occurred.

Overall, men supported Roy Moore more heavily than did women. Moore won with men by more than 10 points and lost among women by nearly 20. Among men, it didn’t matter if they had children. Men with children living in their households backed Moore by 14 points; those without backed him by 15.

Among women, though, the picture was very different. Women without kids at home supported Democrat Doug Jones by 9 points. Those with kids at home? Two-thirds supported Jones, giving him a 34-point margin.

Exit polls can be a rough tool for understanding what happened in an election. In this case, there may be other demographic factors that overlay onto these results. For example, people living with young children at home probably tend to be younger themselves, and the younger voters in Alabama were the more likely they were to support Jones. (Though, of course, this doesn’t hold for men.)

Put another way, it’s not the case that we can say with certainty that the allegations against Moore were a bigger factor for women with children than for others — just that there is a correlation. Women were more likely to view the allegations against Moore as important. Women were more likely to oppose Moore. Women living in households with children were much more likely to oppose him.

There’s a reason that this unusual question was probably included in the exit poll, after all, because it seemed as if having children might inspire people to view the race differently given the high-profile allegations against Moore. Women with children living at home did indeed view the race differently than other groups — including men in the same situation.