Several people sent me a short essay in the New York Times, “When beliefs and facts collide,” by Brendan Nyhan.  They thought, correctly, that I’d be interested in it because it discusses the reasons why so many Americans deny palpably true science, in particular evolution and human-caused global warming. Both of these “theories” are supported by mountains of evidence (no rational scientist would deny evolution, and something like 97% of climate scientists also accept that human activities are making the Earth warmer).

So why the opposition from many Americans? Nyhan summarizes the answers briefly, but they all come from a 49-page paper by Dan M. Kahan that’s in press in Advances in Political Psychology. Kahan is a professor of law and psychology at Yale. The paper, in advance form, can be downloaded free at the link at bottom.

The title refers only to climate science, but a large chunk of the paper is about evolution, and I’ll deal with that today. Tomorrow, I hope, I’ll discuss Kahan’s conclusions about climate science.

Kahan first cites previous studies that seem to show that those who reject evolution—42% of Americans are young-earth creationists when it comes to humans, while only 19% accept purely naturalistic evolution—know just as much about evolution as those who accept it.  Then Kahan proceeds to show pretty much the same thing, but adds something else: something most of us know, but others either don’t know or don’t get.  And that is that those who reject evolution do so largely on religious grounds, not because they don’t know what evolution is. In other words, Kahan concludes that American rejection of evolution is not due to lack of information or ignorance; it’s due to adherence to one’s faith community that rejects evolution. You reject evolution because your “community” does, and you want to get along with them.  Tomorrow I’ll show that this is pretty much true for global-warming denialism as well, but the “communities” there are ones of political conservatism or liberalism, not religion.

Anyway, Kahan gave a test of science knowledge (“Ordinary science intelligence,” or OSI) to a number of people, and also measured their religiosity as “self-reported church attendance, frequency of prayer, and perceived importance of god in one’s life.” To be sure, it’s not clear from the paper whether Kahan did the polling himself or is simply using other peoples’ data. But no matter.

You might want to take the OSI test itself, as it’s shown at the end of the paper. Some of the items are easy, like the following (the percentage of correct answers and references are also given):

And others are harder, like these. The first one was the hardest for the takers, and it is a stumper. I got it, but had to think carefully. The third one you can solve in your head intuitively or with a simple algebra equation, but only 13% of people got it; most people answer “ten cents.”

There were also two evolution questions, one about your personal view:

And the other about your knowledge of what the theory of evolution maintains:

Notice that 55% of Americans agreed with evolution (a bit higher than that found in the Gallup Survey, but still disconcertingly low), but 81% of Americans at least know what the theory says. That’s still pretty low to me; how can one not know what biological evolution is?

At any rate, Kahan simply looked at how peoples’ performance on the evolution questions correlated with their general scientific knowledge as shown by their OSI score. First he showed that for the non-evolution science questions, performance on them was pretty highly correlated with overall performance.

The three graphs below show that for three of the “non-evolution” science questions, performance on those questions was well correlated with overall OSI. The relationship was not as tight for the tough mammography question, for only those who had the best scientific knowledge or probability skills could answer it correctly. These graphs show, for each category of OSI performance (X axis), what percentage of people answered the question shown correctly (Y axis). Scores on OSI were placed into 21 discrete categories, and bars are the 95% confidence intervals:

Evolution wasn’t quite the same, as acceptance of it didn’t rise as strongly with general performance on OSI:

In other words, the better you did overall, the higher your probability of accepting evolution, but that probability didn’t rise as fast with general science intelligence as did correct knowledge of other scientific ideas. Why is that?

If you break down the data by religious belief of people in each of the 21 categories, you find something interesting. Religious belief doesn’t make much of a difference in non-evolution questions, as seen by the yellow scores (of those more religious than average) versus blue scores (of those less religious than average). For example, look at the three non-evolution questions whose overall results are given above:

Religious people’s scores are pretty much in line with less religious people’s: the orange bars highly overlap the blue ones.

Now look at the evolution question:

The results are clear: in contrast to other scientific facts, even if you know more about science in general (X axis), if you’re pretty religious (orange bars), it doesn’t affect your likelihood of accepting evolution.  It does, however, if you’re less religious (blue bars). The combination of these two plots is what flattens the summary plot for this question given above.

While the authors found a slightly negative correlation between religion and OSI (-0.17), meaning that religious people don’t know quite as much about science in general as do the nonreligious, that can’t explain these results, for we are comparing performance on a single question within a group of people comprising both religious and less religious people who know the same amount about science.. The discrepancy is large; as the authors note (my emphasis):

Their performance on the Evolution item, however, is clearly discrepant. One might conclude that Evolution is validly measuring science comprehension for non-religious test takers, although in that case it is a very easy question: the likelihood a nonreligious individual with a mean OSI score will get the “right” answer is 80%—even higher than the likelihood that this person would respond correctly to the relatively simple Electron item.

it is a very easy question: the likelihood a nonreligious individual with a mean OSI score will get the “right” answer is 80%—even higher than the likelihood that this person would respond correctly to the relatively simple Electron item.

In contrast, for a relatively religious individual with a mean OSI score, the probability of giving the correct response is around 30%. This 50 percentage-point differential tells us that Evolution does not have the same relationship to the latent OSI disposition in these two groups. 

Indeed, it is obvious that Evolution has no relation to OSI whatsoever in relatively religious re-spondents. For such individuals, the predicted probability of giving the correct answer does not increase as individuals display a higher degree of science comprehension. On the contrary, it trends slightly downward, suggesting that religious individuals highest in OSI are even more likely to get the question “wrong.”

Well, do religious people simply know less about what the theory of evolution says than do other people, and that explains the discrepancy above? If so, then the lower acceptance of the theory among the religious (at every degree of OSI performance) could simply be due to their ignorance. (I mean ignorance about evolution, of course, not general ignorance.)

But that’s not the case. Below you can see the performance on the “I accept evolution” question divided up by OSI score, with and without religious belief indicated. Total data on the left, divided up by degree of religiosity on the right:

While religious people know slightly less about the theory, especially at higher levels of general science intelligence, both curves on the right go up with OSI score, and resemble curves for other individual questions on the OSI exam. As the authors note:

When the clause, “[a]ccording to the theory of evolution . . .” introduces the proposition “human beings, as we know them today, developed from earlier species of animals” (NSF 2006, 2014), the discrepancy between relatively religious and relatively non-religious test-takers disappears! Freed from having to choose between conveying what they understand to be the position of science and making a profession of “belief” that denigrates their identities, religious test-takers of varying levels of OSI now respond very closely to how nonreligious ones of corresponding OSI levels do. The profile of the item response curve—a positive slope in relation to OSI for both groups—supports the inference that answering this variant of Evolution correctly occupies the same relation to OSI as do the other items in the scale. However, this particular member of the scale turns out to be even easier—even less diagnostic of anything other than a dismally low comprehension level in those who get it wrong—than the simple NSF Indicator Electron item.

The take-home lesson: even science-savvy religious people don’t accept the truth of evolution, though they accept the truth of virtually every other scientific concept.  In other words, they uniquely reject evolution because of their religious beliefs. That rejection, to repeat, is not based on ignorance, for at all levels of OSI performance the religious know nearly as much about what evolution says as do the nonreligious. (This is borne out by previous work that asked even more about the theory of evolution.)

As I’ve been saying repeatedly, the way to eliminate creationism is not to teach people about evolution (as I tried to do in WEIT), but to get rid of the major factor that make them deny evolution: religion. Granted, WEIT was successful in changing some people’s minds (I have lots of emails attesting to that), but I suspect its main effect was simply to tell people who already accepted evolution about the kind and amount of evidence supporting it.

If we want to eliminate creationism, we have to eliminate the kind of religious belief that makes people reject evolution. (That, of course, is not all religious belief: Unitarian Universalists, I suspect, have no problem with evolution.) But to eliminate religious belief, we must eliminate the conditions that promote it, which in my opinion are dysfunctional aspects of society that make people turn to God. But that’s another theory, and one I’ve written about before.

Tomorrow we’ll compare evolution denialism to climate-change denialism.

h/t: Wendy

_________

Kahan, D. M. 2014 (in press). Climate science commuication and the measurement problem.  Adv. Pol. Psych.