Last week, the Pew Research Center released the results of a survey on American device ownership. Much of it was intuitive. A growing number of adults have a smartphone, though there are disparities across age groups and income levels. Over 90 percent have some kind of cellphone, and ownership of laptops and gaming consoles remains steady. But there was one item that stuck out. According to Pew, a larger percent of women own video game consoles than men.

Research from the Entertainment Software Alliance and other groups has confirmed that a large number of women play games of all kinds, even if the percentages for different genres vary. But the latest Pew survey is a confusing piece of evidence — and not necessarily something we should be citing when we talk about gaming's gender gap.

First, the numbers. In 2015, Pew surveyed around 1,900 adults, about a third of them over landline phones and two-thirds over cellphones. Among other questions, it asked respondents: "Please tell me if you happen to have each of the following items, or not. Do you have a game console like Xbox or PlayStation?" According to the results, 40 percent of adults answered yes. When you broke down the data further, 37 percent of men said they did, and 42 percent of women. That's a change from 2010, when researchers asked the same question and found that 45 percent of men and 40 percent of women had a console.

The problem occurs when we try to generalize this to a meaningful social trend. Unlike cellphones or laptops, game consoles aren't really personal electronic devices. They're closer to televisions. So if one person in a household buys an Xbox or PlayStation, it's fairly likely that another adult in that house will report "having" the item, whether or not they purchased it or use it. If Pew calls a single-gender household or a person living alone, the data splits up neatly. But as of 2012, only around 15 percent of American households were women living alone, and 12 percent were men living alone. (Married couples, with or without children, made up around 48 percent.)

Pew gives us a good raw number for overall console prevalence, and things like race and income level tend to be shared in a household. But the gender data doesn't quite answer the question we're asking by proxy: whether a larger proportion of women than men are buying and playing console games. I asked Pew's research team about the breakdown, and the response was noncommittal — they're capturing numbers, after all, not doing sociological analysis. "We strive to write our questions in straightforward language so that people can easily answer them," read the response. "In some cases like this, there may be interpretation left to the respondent. If they think they 'own' a game console, whether or not they bought it, then we capture them as game console owners."

A survey focused specifically on this could have included more specific followup questions. Further splitting up the data — counting only single-person households, for example — could yield interesting results even without more information. Unfortunately, Pew wasn't able to give me anything more specific than the details that it published last week.

It's possible that Pew's numbers solidly reflect who is interested in console gaming, even if "a greater percentage of women" means something slightly different than "more women." Whether they're accurate or not, it's a great jumping-off point for further research. But there's enough ambiguity that it's a fragile factoid to be passing around, and one that might not actually tell us much about men, women, and games.