Friday Weird Science: Estrogen for President!

Oct 26 2012 Published by under Friday Weird Science, Uncategorized

Today's Friday Weird Science is a synchro-blogging with the ever fabulous Kate Clancy at Context and Variation. Because when you see a paper this...special...well sometimes it takes two of us. Make sure you head over and check out her awesome coverage!

I don't know about you all, but when I do anything, I do it with my hormones. This results in issues sometimes, like when my hormones tried to take the SAT (they never can bring a #2 pencil), but I'm confident that when it comes to the voting booth in November, my estrogen will represent my interests.

Or at least, it will depending on what phase of the menstrual cycle I'm on. Otherwise I might accidentally vote for Obama or something.

Durante and Arsena. "The Fluctuating Female Vote: Politics, Religion, and the Ovulatory Cycle" To be published in Psychological Science, 2012.

No one will deny that our values and political affiliations can change over time. The question is both how and why. Do people always become more conservative as they grow older? Are changes in political affiliation due to things like higher economic status, changes in religiosity, increased education, social pressure, or other factors? Or, in the case of women, do we just vote based on how likely we are to get knocked up? I mean, that last one sounds like biology! That's got to be the best reason.

The authors of this particular study wanted to look at the idea that religious feelings and political orientation were affiliated with "reproductive goals" in women (aka, how likely your eggo is to be preggo). They administered two internet surveys, the first to 275 women, and the second to 506 women, none of whom were on hormonal contraceptives. They asked the date of the last menstrual period, the expected date of the next one, and the average length of their menstrual cycle, and using this data, divided the women up into high and low fertility groups. They asked if the women were single or in a committed relationship.

Then they asked a series of questions. The first study looked specifically at measures of religion, asking questions like " (1) “How much do you believe in God?” (Not at All – Very Much); (2) “I see myself as a religiously oriented person” (Strongly Disagree – Strongly Agree); and (3) “I believe that God or a Higher Power is responsible for my existence” (Strongly Disagree – Strongly Agree)." The authors used these questions to come up with a score of "religiosity".

What they show is that single women show lower religiosity when ovulating, while married women show higher. The authors don't really explain this data, but in the introduction they hypothesize that ovulation makes women seek out more reproductive benefits, so I guess this would mean that when you are married and ovulating, you want some from your mate because...religion, and when you are single, you want some from, well, people, because...less religion?

Anyway, on to the politics.

In the second survey, the authors asked for the same menstrual cycle data and single vs married data, and then asked political questions, dividing their questions into social and economic issues (apparently no one cares about the views of your menstrual cycle on foreign affairs).

The questions for economic issues included things like "taxation policy, corporate regulation, economic standard of living, and privatization of social security", while the social issues included things like "legalizing marijuana, equal rights, and stem cell research".

The authors found no differences with regard to economic issues based on high vs low fertility or committed vs single status (hormones don't need your dirty money, apparently). But when they came to social issues, they saw some differences.

Specifically, they saw that single women who were high fertility were more liberal, while committed women who were high fertility were less liberal.

They then asked the women in the study, if they were walking into a voting booth today, who would they vote for, and then they asked which political party they would like to receive a $1 donation.

When in the low fertility portion of the cycle, committed and single women had the same preference. But when ovulating, single women preferred Obama more, while committed women preferred Obama less.

From these studies, the authors conclude that "Women’s voting preferences were mediated by their ovulatory-induced changes in
political orientation."

The problems I have with this paper are many and varied. And I'm certainly not the only one. Not only should you check out Kate's post, but you should also check out the coverage at Retraction watch, which covered the story of the original CNN report of this study being taken down (sorry, CNN, the internet is forever). The study as reported by the media was sketchy enough, but once you have access to the full paper, the holes really come through.

To begin:

1. Women in committed relationships were more likely to have children, more likely to be in a higher socio-economic class, and were older. The authors never controlled for these incredibly important factors in their data.

2. Let's just be clear here, Romney ain't winning with women no matter which reproductive phase they vote in. The authors state that "married women tend to vote Republican, whereas
single women tend to vote Democrat." and "They also reveal a potential reason for the female divide in the 2012
presidential election, in which single women strongly prefer the more liberal candidate, while married women prefer the more conservative candidate." I'm not sure where they get this belief, but it's certainly not from their data. "Non-ovulating" women rated about the same in measures of Liberal and voting no matter what the relationship status, only the "ovulation" status showed a difference. Even when support for Obama was at its lowest point, it was still 60% in married women who were ovulating (and over 75% in non-ovulating). At its highest in single ladies, it's at over 85% (and around 75% in non-ovulating). The first is a comfortable win and the second is a landslide. So it looks like the incumbent need not fear for the menstrual cycle on election day.

3. The authors never asked how informed any of the women were about the political views they were rating.

4. The authors rated "high fertility" as cycle days 7-14 and "low fertility" as cycle days 17-25. They cut out the menstrual period itself, and some days before and after, as well as days in between "high" and "low" fertility. They cut out over a third of the menstrual cycle. Over a third! And this is where it gets sticky because...

5. Ovulating. You keep using that word. I do not think it means what you think it means.

The authors start out classifying as "high fertility" and "low fertility", but then switch on over to "ovulation". And this is a mischaracterization to say the least. Most "high fertility" times of the cycle are not all that fertile, and it certainly does not mean you are ovulating. For more on this, I recommend you check out Kate's post.

6. They ask "could the hormones associated with ovulation account for some of the discrepancy between single versus married women?"

I'm not sure what that means. First off, they didn't take any hormone levels. It was an internet survey. No hormone levels. Secondly...why would the hormones associated with ovulation account for DIFFERENCES between married and single women? Wouldn't they, rather account for some of the similarities? Are they stating that estrogen and progesterone have entirely different effects on behavior due to whether or not there's a committed relationship? They can suggest that and that's fine, but there's no data to back that up, and no studies have been done on this (and certainly none that are cited). And it certainly runs contrary to much of the behavioral literature on the topic.

6. Let me repeat. The authors here say that political views differ based on ovulatory phase...but they have no proof of ovulation. It was an internet survey at a single time point (what we call a cross-sectional study). They did not take hormone levels. The likelihood of ovulation actually much lower than you'd think. This point, and the one I am about to list below, make me think these data are mostly artifact.

7. I'll repeat again. The authors here say that political views differ based on ovulatory phase...but they do not examine the same women over time. This is a cross-sectional study, a single time point. They did not ASK any of these women whether their political views differed as a function of ovulatory phase because they did not ask them during different phases of the cycle.

8. Finally, the social beliefs that they asked about...well many of them are not exactly things that you believe in shades of grey. Certainly issues like stem cells have shades of grey that can be influenced when rating on a 1-7 scale (as required by the study), but marriage equality? This is not something you believe in, but only at, like, a 4. You either believe that 'marriage is between a man and a woman' (as stated in the survey), or you don't. And this view is very likely to have been acquired and adhered to strongly. You don't just randomly decide to not believe in gay marriage so much today. Similarly, abortion issues are something that many women have thought very deeply about, and which many hold strong issues on. You don't just suddenly start rating your pro-life beliefs as a 2 rather than a 6.

I could keep going, but I don't think I need to. The paper bases its hypotheses on assumptions about the menstrual cycle for which they have no proof and ignores some major confounds in the data. But, as Kate Clancy's not like they're doing science or anything. And what do we know? We're the ones stuck voting with our ovaries.

19 responses so far

  • [...] SciCurious’s report “Friday Weird Science: Estrogen for President!“ [...]

  • A says:

    I think one of the most important role of women in the modern world is be part of the whole of economy, but we don't know enough about it, and that is just the problem.

    Just begin to ask a few simple questions:

    Do you know how much we use and purchase? and where does it all go? or where does it all come from? Are we commodities, if so, what is my value? who assigns that value? based on what? is the actual proportion of hours worked relative to income a direct projection of population relative to available or circulating money?

    Yeah, how does that go? thanks for reading this little note.

    Ah, I think all the politicians look good, but what will or should matter is the above.

  • Mark B. says:

    It was nice to see someone actually critique the study instead of just rant about a researcher linking "hormones" to "voting behavior". Thank you.

    One quibble with point 8: Political scientists often note that ,most "regular" people don't have firm answers on many issues, including gay marriage. For many people politics and the surrounding issues are only a very very tiny sliver in their life and so they may very well score a 4 on a seven point scale. People who are highly education and who pay a lot of attention to politics, however, are very clearly polarized on these types of issues. So just because it is difficult for you, someone who is clearly educated and thinks deeply about issues, to think about scoring a 4 doesn't mean that holds for someone else as well.

    And as someone who has asked similar questions (and even more black/white), there is often a lot of variation in the measures compared to what I would anticipate.

  • [...] blogger Scicurious’s take on the study here. Share this:TwitterFacebookDiggRedditEmailStumbleUponLinkedInTumblrPinterestLike this:LikeBe the [...]

  • becca says:

    I mentioned on twitter that the notion of ovulation being involved is over-interpreting the data, but hormones being involved isn't as much, because there are a lot of data on how hormones are different for the two slices of the cycle they picked.
    Also, I *don't* have an a priori distrust of the hypothesis that estrogen/progesterone are interpreted completely differently based on social context, precisely because of the oxytocin literature. I think it's great the study brought up the idea, even if it is sorely lacking in testing it.

    As far as your objection 7), I agree but I also see this as potentially a high-cost suggestion. You might need a bigger N to compensate help give you the numbers to control for the political landscape changing over a month or two.

    As far as your objection 8), I'm really curious what a scatterplot of the answers they got actually looks like. Because other people *do* see shades of grey, given where they see the Overton window on the subject. For gay marriage, you could imagine someone ranking themselves as "a little in support" if she truly wanted to let states decide for themselves, but would prefer to live in a state where it was legal (I think this opinion is dumb, but I know it can be honestly held). For abortion, the shades of grey are many. "Opposed to abortion in all cases, even if the pregnancy kills you" is different from "Opposed to for sex selection and when the choice is based on lack of financial resources (+ in support of better financial resources for pregnant women)" which itself is different from "Support for access to abortion in every case; i.e. I believe my 14 year old daughter should be able to have an abortion without me knowing to prevent incarceration of her skeezy high school history teacher who knocked her up"

    The real question is- in states with early voting, what phase of the menstrual cycle do women pick to vote? Aside from "not on cramps-from-hell day" (when applicable). And if it's not random, does it vary by how people view voting (i.e. if people are excited to support a particular candidate vs. if they feel it's a civic duty to be endured like jury duty)?

    I will say that, the indirect finding that religiosity and social values, but not economic ones, co-vary is plausible and *important* for political strategy. I don't myself much care for social conservative/economic liberals, but there's a real niche for them.

  • Pascale says:

    What I'm trying to figure out is why the hell anyone did this study.

  • pyrope says:

    Was there an upper limit age cutoff in this study? It would be interesting to calculate the hormone levels of post-menopausal women.

  • joe arrigo says:

    If women vote with their ovaries, them men must vote with their testacles. Is that study in the hopper?

  • Neuroskeptic says:

    Oestrogen has failed to deliver the change it promised. Glutamate for President! (and GABA for VP)

  • [...] Retraction Watch who called it “poor science, poorly reported.” The blogger Scicurious wrote that the study “bases its hypotheses on assumptions about the menstrual cycle for which they have no proof [...]

  • [...] action by promoting this…thing. You can call it a study, but I wouldn’t. Just read this blog post over at scientopia describing all that is wrong with this “study” that claims women’s ovaries [...]

  • [...] first, a trickle of criticism of the work itself has begun, such as a post by Scicurious, discussing the work. Scicurious is very concerned about the method of assessing ovulation. As she [...]

  • Paul says:

    Having read the paper I have to say: while some of the criticisms above are valid, it's really not such a bad study. In particular there's a solid and quite surprising result here: a woman's stated religious and political views depend on the point in her reproductive cycle on which they are polled. In the case of religiosity this result is replicated in a second independent dataset (admirable and quite unusual in this field). The effect size is not small.

    None of the arguments I have read here or elsewhere provide reasons to think this result is artifactual. There's nothing wrong with a cross-sectional study, unless you can suggest an alternative mechanism for generating these results except that an individual's views change over time. It's true they didn't have a direct measure of ovulation, but that seems to miss the point: clearly no-one means to claim that menstruation or ovulation directly cause changes in political views: the correlation must be mediated by other factors currently unidentified.

    Identifying those factors is going to be an important part of making sense of this result, and it seems some people have unfortunately jumped straight to the simplest interpretation, i.e. hormones change political views. It's reasonable to criticise that interpretation, but I'd be interested to hear what SciCurious and others think the alternatives are - because it seems there's a genuine result here that deserves to be taken seriously.

    • scicurious says:

      Thanks for your thoughtful comment Paul. While I agree that there is nothing wrong with a cross-section study, I think that the conclusion that "Women’s voting preferences were mediated by their ovulatory-induced changes in political orientation." is too strong for the results of a cross-sectional study of this type. They saw differences in views, when different women were polled at different points in the cycle. These are not ovulatory-induced changes. If you want to talk about ovulatory-induced changes, I think you need to perform a longitudinal design and actually measure ovulation. I also have a real problem with the confounds of age, child bearing, and socio-economic status differences that were not controlled for, and could well account for the effects.

      I do think that hormone changes might influence the STRENGTH of political views (though probably not the direction, despite the claims from the authors that married women preferred conservative candidates, all subjects in the study leaned liberal), but I feel this study has far too many flaws to be conclusive. I would strongly prefer a longitudinal study on political attitudes with concurrent hormone levels. I also think the social politics questions used were probably not the best ones, they are ones where people who do hold views, hold very strong ones. I think issues with more nuanced views might be better for showing changes correlating with reproductive cycle.

  • According to, "Ovulation is the event that sets the pace for every woman to get pregnant or avoid pregnancy. It is simply the monthly release of an egg or ovum from any of the two ovaries, usually triggered by luteinizing hormone (LH). The egg subsequently travels down the fallopian tube through to the uterus, popularly known as the womb."

    Having said that, I chose to stand on the side of the critics of this survey. As for me, it's a calculated attempt by the authors of the survey/researchers to gain attention at this time.

    It's an exercise that is demeaning and disrespectful to ovulating women generally. The purported findings tend to suggest that women are not rational thinkers when ovulating.

    I think ovulation has no significance, as a contributory factor, to a woman's normal decision making process, except in a circumstance where a woman has problem during ovulation, to the extent she is psychologically influenced by such problem.

    Whether you are a man or woman, any activity or circumstance can influence your decision-making process. For instance, the decision a man or woman will make while having sex may be a little different than in a normal situation.

    So, this survey is making no new finding in any way.

    Women are great decision-makers, whether ovulating or not! That's my view.

  • [...] There are many other problems with the study as well; for a more thorough debunking, check out this post by Scicurious and this post by Kate [...]

  • [...] of Retraction Watch who called it “poor science, poorly reported.” The blogger Scicurious wrote that the study “bases its hypotheses on assumptions about the menstrual cycle for which they have no proof and [...]

  • [...] Friday Weird Science: Estrogen for President! [...]

Leave a Reply