The Quinnipiac poll at least had a sample of more than 1,200 Ohioans. The Purple Strategies poll surveyed about 2,400 “likely voters,” but only 600 of them were from Ohio; assuming the survey was evenly divided between the genders, about 300 Ohio women participated. That’s a big enough sample to have some value, but, as this chart shows, small enough that it creates a pretty large margin of error—which makes the poll a bit of a shaky foundation for arguments about a growing gender gap. (CJR’s Brian Crowley has previously written about the problems with reporting on polling subgroups.)

This doesn’t mean that reporters shouldn’t pay attention to the specific ways in which campaigns court women voters. (As Ward wrote in his lede, “The war over women is on.”) But we shouldn’t get carried away by flashy poll numbers. And as for the idea that Ohio women will be a key voter demographic in this election, or that the gender gap is growing, only time—and some more data—will provide proof.

Meanwhile, reporters who are trying to get a sense of how female swing voters here view the campaign would do well to fan out and interview “ordinary” women around the state, not just those who are politically connected or show up at rallies. To his credit, this is just what Huffington Post’s Ward did in a follow-up Wednesday, on the first day of the Ohio State Fair.

He found a lot of disgust and disenchantment, with little enthusiasm for either candidate. That’s a sad story, but it’s consistent with a very close outcome, one in which the campaigns fight hard for every vote. That means reporters need to keep their fingers on the pulse of political battles like the gender story—but also to keep the big picture, and fuller context, in mind.

T.C. Brown covered government and politics in the Ohio Statehouse Bureau for The Plain Dealer of Cleveland for more than 17 years, and he has also written for other local, state and national publications. Brown is a founding partner in Webface, a social media communication company.