Are Women Less Skeptical?

The much-awaited survey on the religiously unaffiliated, “American Nones: The Profile of the No Religion Population,” is out and it exposes some fascinating gender differences. Or does it?

“Women are also less skeptical than men and less drawn to irreligious and anti-religious views. They are more likely to reject a secular upbringing,” Kosmin says.

Well, that’s just from the researcher talking to the press, and you know how that can go. How about the report summary?

Whereas 19% of American men are Nones only 12% of American women are Nones. (Fig. 2.1)

Oh, my Gawd! Men are 60% more skeptical than women!

And from the report itself (detailed pdf):

There are a couple of additional findings worth noting here. Looking at retention by gender, Nones are more likely to retain men than women: 66% of men who reported no religion at age 12 were Nones at the time of their participation in ARIS 2008, but only 47% of females who reported no religion at age 12 remained Nones. Of those who reported having a religion at age 12, 15% of men left while only 9% of women did. It appears that American women have a greater affinity for religion than men. And conversely men have greater affinity for secularity than women.

Well, that’s easy to say, but of course, the study isn’t directly measuring skepticism or the appeal of secularity or religious views. It’s measuring social behavior. It’s measuring affiliation.

So what happens if we look at affiliation? 19% and 12% give us 81% of American men who are religiously affiliated versus 88% of American women. Which means we have 9% greater affiliation among men than women. That’s still a significant difference, and it’s the difference in what is actually being measured by the survey.

Of course, religion isn’t the only major type of affiliation that people are expected to have. In fact, Kosmin and his group looked at the interaction of political affiliation and and religious affiliation. Unfortunately, they didn’t report their results by gender, but a 2004 Pew study did. 59% of men and 65% of women identified themselves as affiliated with a political party. In other words, women showed about 10% greater rates of political affiliation.

Huh. About the same, then. Who’d have guessed?

Okay, now that I’ve got you thinking about what the survey does and doesn’t measure, and the complexity of the situation, it’s time to take a step back. When Kosmin is talking about regarding skepticism in the original quote isn’t the base statistics on affiliation. He’s talking about a smaller survey of beliefs among the non-affiliated, in which 58% of women indicated a belief in a higher power or personal god, compared to 46% of men.

Yet Kosmin’s statement isn’t that women are less skeptical of religion. Nor does it address the interaction between those beliefs and the ways in which women move into and out of religious affiliation (if males with no belief stayed among the unaffiliated while the women with no belief left for social/affiliation, rather than belief-based reasons…). It doesn’t, as my Episcopalian atheists friends would be the first to tell you, even address belief among the religiously affiliated by way of comparison. In no way can it be said to address gender differences in skepticism in general.

While it is possible to conduct studies in which women show less skepticism, it’s also possible to choose topics on which men are less skeptical. One survey that covered a range of topics found no difference by gender. If you want to really address the question, you simply have to look at a broad range of areas requiring skepticism. I recommend Podblack’s threepart series on the topic.

Whatever you do, don’t take Kosmin’s word for it.

Update: I also recommend this post at Pandagon and the comments, which delve into the question of gender differences in affiliation.

Are Women Less Skeptical?

16 thoughts on “Are Women Less Skeptical?

  1. 1

    My newly proposed study will measure whether or not men or women change their position on a fictitious situation when provided with varying degrees of information.When I get the results I hope to prove that men are less reliable than women.That makes sense, right?

  2. 2

    As an anthropologist, I would have predicted that women are more affiliative in general while men are more coalitionist. Well, that's not a prediction, but an observation seen across all cultures. Human universal status. So, these results don't surprise me much. I agree that this is primarily about social behavior and affiliation and not skepticism. That is not to say it is not also true that women are less skeptical. But I'm skeptical of that.

  3. 3

    Dan, in the context of this post, that might even be too direct a relationship between what you're studying and your results.Greg, I wasn't really surprised at what I found when I went looking for political affiliation data either. Well, I was surprised at how closely the percentage difference tracked between the two measures, but I'm not assuming that's more than coincidence at this point.

  4. 4

    but I'm not assuming that's more than coincidence at this point.I don't think its a coincidence, because my internal model of human nature predicts both for the same reasons.

  5. 5

    To be more clear, I'm not assuming that seeing the same magnitude "affiliation bump" is more than coincidence. While I think the same basic push to affiliation is acting in both cases, I'm not assuming the situation isn't more complex than that, that we wouldn't see different sized bumps in different situations or different types of affiliations.

  6. 7

    While phrasing as 81% v. 88% is relevant to debunking/discounting the strong claim made in the survey, the phrasing of 12 to 19 has more relevance when one considers what one would expect of the group and social dynamics within groups of such people. A group that's about evenly divided in gender is going to have different group dynamics than one which is strongly skewed in one direction. However, these numbers mean that in a given section of skeptics (assuming these ratios) one would expect that about 60% will be male. That's unlikely to have that substantial an impact on group dynamics.

  7. 8

    Strangely, digging through my anecdotal history, I can recall exactly one out of dozens of theist proselytizers I've argued with over the years being a woman (self identification helps with that though). I'd suggest maybe three quarters were obviously males (or pretending to be males) based on making suggestions about their gender, and the rest were ambiguous by never having said so. Anecdote = anecdote. Anecdotes = data! 😀

  8. 9

    Thanks very much for linking to my blog! 🙂 After working on those posts over a period of time (and still finding an audience for my writing), I'm glad they came in use to someone who was after more information. Great summary of this paper as well, I've downloaded it to peruse it more thoroughly – you've got me off on a great start. 🙂

  9. 10

    Tracy: Affiliative = sharing part of your identity with others in a common social structure that has various benefits and obligations. Coalitional = dynamically cutting deals with individuals or sets of individuals in a Machiavellian manner.

  10. 11

    Thanks Greg. Are there methods for quantifying affiliative vs. coalitionist behavior, or do you base your predictions on anthropoligical evidence? Also, I would be interested to know if different regions of the brain become active in these different contexts, and if either correlates with brain activity seen during various kinds of religious experiences (e.g. prayer vs. confession).

  11. 12

    Joshua, I thought about trying to break some of those things apart, but without the actual data…. I would love to get my hands on the data, though.Jason, if I recall correctly, creationism is one of the predominantly male irrationalities.Kylie, I was thoroughly impressed by how many times I wanted to know about gender differences on a particular belief, Googled it, and found one of those posts in the top results. I gave up in very short order trying to find something you hadn't covered.

  12. Ian

    A few other random thoughts1. Church is where you take the children to give them a proper moral foundation.* Since child-raising is women's work, more women attend church.2. Church is where you go when you're retired and have the time to get involved in your community. Women live longer than men, so again, you're likely to see more of them there.3. It's also where you go when you need support from your community. Women earn less than men, they live longer than men, so chances are that they're more likely to need assistance from their community age.4. Many churches provide services to needy children in their community, which draws their parents into affiliation with the church. Again, the idea that child rearing falls into the women's sphere, coupled with the fact that single parents and custodial divorced parents are far more likely to women makes it more likely for women to be involved in churches (or other religious organisations).*Not that I endorse these generalisations or anything.

  13. 14

    Thanks Stephanie! Isn't it weird how many of the 'mainstream' skeptical blogs don't often link to me though? Despite all the Google hits, it's rare you'll find much beyond the Open Laboratory Best of Science Blogging link to one of my posts (a fun blog-post on the psychology of romance and plagiarism of a nature journal, of all things!)However, if you want me to research anything to add to the series, I'm more than happy to. 🙂 Helps me sort out my thoughts for my thesis. 🙂

  14. 15

    So among female skeptics, are a higher proportion of them members of groups than male skeptics? And as freethinker/secular/skeptic groups become more prevalent, will more women affiliate with them and then identify as skeptics?

  15. 16

    Theo, I don't have any firm answers for you. One thing I can say, which lends some credence to the idea that church affiliation is tied to family responsibilities, is that since MN Atheists added a meetup group and more kid-friendly activities, female participation has jumped. So I would suspect that the types of groups available for participation make a difference as well.

Comments are closed.