Umbrellas, soaked heads, and the ecological fallacy

This is the second of four posts on religiosity and life history. In yesterday’s episode, I noted that the evidence in favor of viewing religiosity as driven by fast life history comes from group-level data. Poorer nations tend to have more churchgoers; American states and racial groups with more teen births and lower socioeconomic status contain more churchgoers.

A key risk here is the ecological fallacy—that is, the mistake of drawing individual-level conclusions from group-level data. It seems natural to conclude, for example, that if states with higher teen birth rates also have higher average church attendance rates, then surely churchgoers are more likely to be teen parents. But that’s the fallacy. You really don’t know what’s happening at the individual level until you check individual-level data.

Here’s a hypothetical. Imagine that you were curious about the effectiveness of carrying umbrellas (vs. not carrying umbrellas) in preventing people’s heads getting soaked by rain. Seems like it’s pretty obvious, but you want to see some numbers. You don’t find any individual-level data on it, but you manage to find averages in various cities for (1) the percentage of people who habitually carry umbrellas and (2) the percentage of people who often get their heads soaked by rain. So you decide to correlate those two city-level variables.

Here’s the thing: This would probably produce a positive correlation between umbrella-carrying and head-soaking! There would be cities like Seattle and Glasgow where lots of people carry umbrellas and also lots of people frequently get caught unprepared in the rain. And there would be cities like Los Angeles and Barcelona where hardly anyone carries umbrellas and also people hardly ever get their heads soaked by rain.

The ecological fallacy would be to use this group-level correlation to conclude that carrying umbrellas makes individuals more likely to get head-soaked—if they just stopped carrying umbrellas, their heads would be dry.

Of course that’s silly. The real answer is that there are places where it hardly ever rains and thus there’s little need to carry an umbrella to keep your head dry. But in other places, it rains a lot. There, it’s the umbrella-carriers who have drier heads than non-carriers. The right data here is individual-level data that measures average local rainfall along with umbrella-carrying and getting head-soaked. That data will show what the city-level data can’t.

So, we know that places with increased levels of fast life history are also places that contain more churchgoers. But that’s simply not enough information to know whether individuals with fast life history are the same individuals attending services. Outcomes like teen births and low education could be the head-soaking or the rain, while church attendance could be the umbrella.

This scenario is related to what Baumard and Chevallier (2015) were suggesting. In their view, it’s the intermingling of fast and slow life histories that causes the slow folks to seek out religion. And as we head into highly developed societies—to return to the analogy—it hardly ever rains, so you just don’t find many individuals carrying umbrellas.

We’re getting closer now. What we need is to look at both the individual-level and group-level predictors of church attendance. Baumard and Chevallier think it’s going to be slow individuals in groups that also have lots of fast individuals. But now we run into another problem, one that infects many current discussions of life history in evolutionary psychology. The problem involves the relationship between fast/slow life history and unrestricted/restricted sociosexuality. This will be the subject of tomorrow’s post.

[Update: The next post in this series is here.]

Is church attendance fast or slow?

This is the first in a four-post series on church attendance and fast/slow life history patterns (which I explain below). As discussion of fast/slow life history has become popular in evolutionary circles, we’re currently in a situation where some smart people claim that fast life histories drive religiosity and other smart people claim that slow life histories drive religiosity. It’s an unusually stark disagreement. In this post and following two, I’m going to lay some groundwork before providing the resolution in the fourth post.

What are fast and slow life history patterns? As Del Giudice, Gangestad, and Kaplan (2016) explain: “At the broadest level of analysis, the life history strategies of different species can be arranged on a continuum from ‘fast’ (early maturation and reproduction, fast growth, small body size, high fertility, short lifespan, and low investment in offspring quality) to ‘slow’ (late maturation and reproduction, slow growth, large body size, low fertility, long lifespan, and high investment in offspring).”

Within species, there are group-level and individual-level differences in fast/slow patterns. As applied to humans, typical discussions note that some groups face more dangerous, unpredictable, harsh environments, leading to faster life history patterns—earlier age at first child, higher completed fertility, reduced investments in embodied capital (e.g., less time spent pursuing formal education), earlier sexual maturation, earlier sexual debut, less stable mating relationships, and so on—while safer and more reliable environments encourage slower patterns.

What does this have to do with religiosity? On the one hand, the secularization hypothesis has long noted that increased economic development in societies over the past century has tended to lead to reduced religiosity in those societies (McCleary & Barro, 2006). Because economic development plays a key role in producing slower life history patterns, the thought is that faster life histories are associated with more religiosity and slower life histories with less religiosity. Further, state-level analyses from the U.S. have shown solid relationships such that states with higher rates of teen births, lower household incomes, and higher STD rates are also states with higher levels of church attendance (Hackman & Hruschka, 2013; Strayhorn & Strayhorn, 2009). In addition, African Americans have both faster life history patterns than white Americans as well as higher rates of church attendance. So it looks pretty good for the fast/religious slow/secular idea, right?

Not so fast. According to Baumard and Chevallier (2015), religiosity could be a tool employed by slow life history strategists to fight off threats from fast life history strategists. Taking a longer view, they see the initial rise of world religions as coinciding with the widespread appearance of slower life history strategies, while the recent decline in religiosity in developed countries could be the result of a waning threat, given how rare genuinely faster patterns have become in these developed countries. They use their slow/religious fast/secular thesis to draw together lots of findings regarding religiosity—its association with prosociality, family values, and various forms of delayed gratification (frugality, work ethic, etc.).

So who’s right? We’ll get there. Before that, I need to make a couple of broad points. To set up tomorrow’s episode: Did you notice earlier that all the evidence for the fast/religious slow/secular hypothesis comes from group-level data?

[Update: The next post in this series is here.]

Achen, Bartels, abortion, and researcher degrees of freedom

Let’s say you want to argue that women watch more TV than men do. You find a sample that measures gender and time watching TV. You look for the relationship and it’s not there—men and women are watching about the same overall amount of TV. But don’t fret, here’s where researcher degrees of freedom can save you!

For example, there might be variables that cover what kinds of programs the sample members are watching and there might be a category for sports. You can eliminate from the sample all the time spent watching sports! You really can. Who’s going to stop you? Just argue that, you know, watching sports isn’t really watching TV. Talk about how watching sports is some special phenomenon. Cite a few books about sports. Anyway, your problem is solved. Now you can present an analysis showing that women watch more TV than men do, proving that you were right all along.

I’ve been reading the new book “Democracy for Realists” by Achen and Bartels and came across a claim made in the intro that I was surprised by: “As the Democratic and Republican parties took increasingly clear, opposing stands on [abortion], partisan identities came into conflict with gender identities. We show that this conflict was resolved in quite different ways for women and for men. A substantial number of women gravitated to the party sharing their view on abortion, reflecting the deep significance of the issue for women. Men, on the other hand, more often changed their view about abortion to comport with their partisanship—in effect, letting their party tell them what to think about one of the most contentious moral issues in contemporary American politics.”

Yeah, well, you know where this is going.

Turns out that the later section with the abortion analysis suddenly starts taking about how Catholics “differ” and are “unique and complex, requiring separate study.” And then it happens—the researcher degree of freedom: “Hence we focus here on non-Catholics only.” They threw out the Catholics. Move along; nothing to see here. ¯\_(ツ)_/¯

They’re using the Youth-Parent Socialization Panel Study (YPSPS), which I’ve got as well. So I did some checks. And, sure enough, the pattern they find among non-Catholics is basically the reverse of the pattern among Catholics (where now men more than women are picking parties based on their abortion views). And in neither group (non-Catholics or Catholics) is the gender difference significant at a level that you ought to put much faith in. And if you use the whole sample, it’s totally clear that there is no significant general gender difference of the kind announced in the intro of the book.

The real story, as far as I can see from my quick check, is that from 1982 to 1997 in the YPSPS there’s a decent portion of people switching-abortion-opinions-to-match-prior-party-affiliations and perhaps even more switching-party-affiliations-to-match-prior-abortion-views. And, if you believe in 3-way interactions (and it’s not clear that you should), there may be something going on where churchgoing and Catholic men are relatively likely to switch parties while non-churchgoing and non-Catholic men are relatively likely to switch abortion views. But there is no general gender difference of the sort claimed by the intro text.

Achen and Bartels have done and continue to do plenty of good work. In this instance, though, they were kicking around a p-hacky sack and ultimately hid the ball. The gender-difference-in-abortion-and-party-switching tale will be one of those things—yet another one of those goddamn things—that will be believed and repeated for years to come, even though it was never true.