You all must forgive me for this blog post. You see, I am in my premenstrual phase, and so with all my insane-o premenstrual symptoms I simply cannot access the part of my brain that makes political decisions. Perhaps when I get through the devastation and physical wreckage we ladies like to think of as menstrual periods I will be able to think again. And when I do, I’ll want to vote for President Obama. I think. Ovulation always makes Romney seem like a big ole’ protective papa fighting for all of America and so, YOU GUYS, I just can’t decide.
Our political attitudes and beliefs are certainly flexible, can change over time, and can be modified by any number of factors. But can they change based on when a woman is more or less fertile during her menstrual cycle? In a manuscript accepted for publication at Psychological Science, Durante et al describe data on political behavior and religious beliefs, and explain how there are reasons to understand these beliefs as flexible and variable. When they begin to talk about reproduction, however, the paper becomes unacceptably flawed and problematic.
So to continue our great tradition, Scicurious and I are tag-teaming on this one. Make sure you read her post for her perspective that addresses more decision-making issues than I do. I’m going to be all about the lady parts.
After you read her, read on here. What’s curious about this manuscript is how at first it takes this relaxed approach that demonstrates a real understanding of the nuances and feedback loops of human nature, then the study authors try to shoehorn it into the usual universalist, reductionist framework we’ve all come to know and love from the worst of evolutionary psychology. Rather than demonstrating how our beliefs may change over time due to changing contexts and experiences, Durante et al would have you believe that women are constantly in flux depending on their time of the month.
I don’t always support gay marriage, but when I do, I am not fertile.
The study authors performed two internet survey studies on 18-34 year old, non-hormonally-contracepting American women (Durante et al. forthcoming). In both, they parsed women into single (not with anyone, casually dating) or in a committed relationship (engaged, living with someone, married) groups. In both studies women in committed relationships were significantly older and more likely to have children; in the second study women in committed relationships also had a significantly higher annual income. They also asked women about the last two dates of their menstrual period in order to classify them as being sampled in their high fertility (days 7-14) or low fertility (17-25) phase.
This means menstruating women were excluded, which they claim are so premenstrual or menstrual symptoms do not interfere with their study. Yet, they included women from days 21-25 of the cycle, which in a regular ovulatory cycle – a cycle all of these women supposedly have – this would be a time of progesterone withdrawal and in the premenstrual phase.
This is a cross-sectional sample. That means each individual was assessed only once, not more than once over her menstrual cycle. For instance, we are not looking at one group of single women assessed at low and high fertility periods, but one group of single women who happened to be in their low fertility period and another group of single women who happened to be in their high fertility period at the time of sampling.
Study 1. The first study by Durante et al looked at interactions between high and low fertility periods and feelings of religiosity. They found low fertility, single women had lower religiosity than high fertility, single women, yet low fertility, committed women had higher religiosity than high fertility, committed women.
Study 2. Durante et al asked participants about religiosity, social political attitudes (i.e., abortion, legalizing marijuana, stem cell research) and fiscal political attitudes (i.e., taxes, social security). They were then asked if they would vote on the day of the study for Obama or Romney, and were asked to which campaign they would give a $1 campaign donation. The study authors found the same relationships between religiosity, fertility phase and relationship status as in Study 1.
[caption id="attachment_627" align="alignleft" width="300" caption="Figure 3. This is figure 4A of the Durante et al, showing the "not-ovulating" and "ovulating" groups and their voting preferences."][/caption]
They found single women: less fiscally and socially conservative, more likely to vote for President Obama (79.3% versus 69.4%), and more likely to donate to Obama (79.9% versus 67.9%). All of these findings were tempered by fertile phase: as with religiosity, the single women who were sampled during high fertility were less socially conservative than those sampled during low fertility. Conversely, committed relationship women who were sampled during high fertility were more socially conservative than their low fertility counterparts. The same trends were found for voting and campaign donation choices, with the more liberal choices made by single women sampled during high fertility times, and by committed relationship women who happened to be sampled during low fertility times. Finally, when looking at all women together, women sampled during low fertility did not have significantly different political attitudes, but single women sampled during high fertility were more liberal than women in committed relationships sampled during high fertility.
Organize the data for physiological relevance, not statistical significance.
If the above is worded awkwardly, it’s because I am trying to make clear how important it is to that the reader understand where the data is coming from. In their use of language, the study authors obscured the fact that their dataset is cross-sectional and correlational. The discussion is especially worrying. Just one example:
“Two studies with relatively large and diverse samples of women found that ovulation had different effects on religious and political orientation depending on whether women were single or in committed relationships. Ovulation led single women to become more socially liberal, less religious, and more likely to vote for Barack Obama” (Durante et al, p. 15).
No. We have no way of knowing what ovulation does or doesn't do because the study authors didn't measure ovulation. Even with the authors trying to erase this reality by incorrectly categorizing women. Even with labeling all of their figures with ovulating and not-ovulating groups. And even if their methods had reflected ovulatory versus non-ovulatory states, their results show only correlations between these factors, not effects of reproductive state on attitudes, particularly because of the cross-sectional nature of the data. When you realize that each data point in this research is of a single individual sampled once, opportunistically at whatever phase of her cycle she happens to be on, the results lose a lot of their power.
Further, the way they parse the high fertility and low fertility phases of the menstrual cycle doesn't reflect what we know about conception probabilities.
[caption id="attachment_628" align="aligncenter" width="362" caption="Figure 4. The second figure of Wilcox et al 2001 which you may remember from my post on "legitimate rape." Demonstrates chances for conception through the menstrual cycle with, you know, real hormone data from real ladies."][/caption]
The average chance for conception on any given day is 3.1% (Wilcox et al. 2001). Using the data from Wilcox et al (2001), women with regular menstrual cycles have a greater than average chance of conception from days 9-18, and women with irregular menstrual cycles from days 12-22. And of course, if we wanted to define high fertility more tightly than just “higher than average” those windows would be even smaller. This doesn’t bode well for the high fertility window of 7-14 days used by Durante et al., and they overlap with the 17-25 window they consider low fertility. Changing the high/low fertility grouping criteria by just a day or two may radically change results. I know this is the case when I look at my own hormone data for my research.
If your foundational assumption is false, your conclusions have no foundation.
Unfortunately, a total misrepresentation of female reproductive biology is unsurprising given how they introduce the topic in their paper. To introduce their section on reproduction, they say:
“Women’s reproductive goals are influenced by a universal biological event – the monthly ovulatory cycle” (Durante et al forthcoming, p.5).
The monthly ovulatory cycle is not a universal biological event. It is a phenomenon experienced by non-hormonally-contracepting, non-pregnant, non-lactating adult women in their twenties and thirties who live in environments that don’t have too much energy constraint, and it’s still not a universal in that group. It doesn’t happen for most elite athletes. It doesn’t happen if you are very obese, if you have polycystic ovaries, if you are an adolescent or if you are a more mature woman. Modern women may have as many as 400 menstrual cycles in their lifetime compared to the estimated 50 of our ancestors, but nowhere near all of them are ovulatory.
Yet this study, along with most studies that try to provide an adaptive explanation for behaviors across the menstrual cycle, requires this assumption in order to move forward and analyze their results. The language of the paper frames the differences they found as being between periods of ovulation versus low fertility, discusses the ovulatory cycle despite not measuring ovulation, and therefore requires that you assume not only that it is normal to have ovulatory cycles, but that all the women of this sample are having ovulatory cycles.
The criteria of simply recruiting women with 25-35 day cycles means you have no way of knowing if your participants ovulated, what their hormone concentrations are, what a normal cycle looks like to them, or even if they had gotten pregnant that cycle.
I do not understand how it is okay to publish papers that are predicated on an assumption about ovulation and hormone concentrations, but not measure ovulation or hormones.
Test your hypothesis, my friend.*
There are two other problems with the ways Durante et al conceived this study. First, in their sample single women were younger, poorer, and had fewer children than women in committed relationships (and remember, they had somewhat loose definitions for both). Age, parent status and socioeconomic status all play roles in political preferences. When you have multiple differences between two groups, but you want to see if only one of those differences is meaningful (in this case, relationship status) there are very easy ways to control for those variables statistically.
Durante et al didn’t do this. So the differences found between the two groups could really be about income or age or being a mom, not relationship status. They didn’t actually test their hypothesis.
Finally, beliefs change over time for many people. But do they change for women over the course of a month, over and over again? And what about pregnant or lactating women? Can they even be trusted to do any important or interesting human behavior with their lack of a menstrual cycle? Or dear god, what about the adolescents? Their cycles are all over the place! It must be because we can’t predict their cycles that they don’t have the right to vote! And don’t even get me started about those postmenopausal grannies with hardly two molecules of estrogen to rub together.
This idea of cyclical political preferences is as ridiculous to me as asking if American men’s religious or political beliefs change every twenty four hours with their very dramatic diurnal pattern of testosterone (which is again, dramatic only if you live in circumstances of low energetic constraint). Just as it was once argued that women should not be world leaders in case they make rash decisions in their premenstrual phase, perhaps the morning glory should be a reason to prohibit male politicians from making policy decisions early in the morning.
An article that relies more on rhetoric and storytelling than evidence, that requires assumptions about human physiology that are patently false, that ignores intra-group differences that better explain their findings than the differences they have decided are important? No big deal, we’re writing this article totes for funzies because it’s presidential election time!
WHATEVER, WE AREN’T DOING SCIENCE OR ANYTHING.
Well, except that the article was accepted for publication in Psychological Science.
I’m voting for Obama, menstrual cycle phase be darned.
Durante KM, Arsena AR, and Griskevicius V. forthcoming. The fluctuating female vote: politics, religion, and the ovulatory cycle. Psychological Science.
Wilcox AJ, Dunson DB, Weinberg CR, Trussell J, and Baird DD. 2001. Likelihood of conception with a single act of intercourse: providing benchmark rates for assessment of post-coital contraceptives. Contraception 63(4):211-215.
ETA 10/26, ~5pm EST: I very lightly edited this after noticing some awkward sentence constructions. I took care to not alter the meaning or content.