How religiosity and science literacy interact: Evolution & science literacy part 2

This is the second of two posts on science literacy and evolution.

And religion.

And liberal democratic society as the naturally congenial but sometimes precariously raucous—or maybe better, simultaneously congenial and precarious because naturally raucous—home for science.

And how the common misunderstanding of what public “disbelief” in “evolution” truly signifies can actually interfere with popular dissemination of scientific knowledge.  Plus compromise norms of respect for cultural pluralism that are essential to the practice of liberal democracy.

See? Get it?

Okay, well, in the last post I described the vast body of long established but persistently–weirdly–ignored work that social scientists have amassed on the relationship between public “disbelief” in evolution and public understanding of evolution and other basic elements of science.

That work shows that there  isn’t any relationship. What people say they “believe” about evolution is a measure of who they are, culturally.  It’s not a measure of what they know about what’s known to science.

Indeed, many people who say they “believe” in evolution don’t have the foggiest idea how the modern synthesis hangs together. Those who say they “disbelieve” are not any less likely to understand evolutionary theory–but they aren’t any more more likely to either.

That so few members of the public have a meaningful understanding of the workings of genetic variance, random mutation, and natural selection (the core elements of the modern synthesis) is a shame, and definitely a matter of concern for the teaching of science education.

But it’s a problem about what people “know” and not what they say they “believe.” What people say they “believe” and what they “know” about evolution are vastly different things. That’s what the ample scientific evidence on public understandings of science show.

In this post I want to add a modest increment of additional evidence corroborating this important point.

The evidence has to do specifically with the relationship between religion, science literacy, and belief in evolution.

The evidence is from a survey of 2,000 US adults recruited and stratified in a manner designed to assure national representativeness.

The survey instrument included the NSF science indicators.

It also contained various measures of religiosity, including regularity of church attendance; regularity of prayer; and perceived “importance of God” in one’s life. These cohered in a manner that enabled them to be formed into a reliable “religiosity” scale.

And the survey contained an item that Gallup and other pollsters routinely use to measure the public’s “beliefs” about evolution.

What do these data show?

Well, I’ll state in summary form what I regard as the findings of interest, and then supply the supporting details:

Okay. Here are the specifics.

1. In general, religiosity (measured, as I said, by aggregating items on church attendance, frequency of prayer, and perceived personal importance of God) is correlated negatively with science literacy.

But the effect is modest. The large overlap in the density distribution plots to the left makes it clear that the portions of population “above” and “below average” in religiosity (“AARs” and “BARs,” let’s call them) both comprise individuals of a wide range of scores on the NSF science literacy battery.

Or at least they do when one leaves Evolution and Big Bang out of the tally, as the NSF itself decided to do in 2010, and & as I have here. To make the science literacy scale more reliable and discerning, I’ve added items from the Indicators’ “science process” battery, which tests knowledge relating to probability and validity of experimental methods.

Consider, though, how AARs and BARs scoring in the top 50% of the science literacy test so measured respond to Evolution and Big Bang:

The difference in the percentages of the two moderately “science literate” groups who answer “true” to these questions is stunningly high.

Now one can use even more intricate statistical tests—ones involving, say, Cronbach’s alpha, factor analysis, and structural equation modeling—to convincingly show that Evolution and Big Bang are not measuring the same latent proficiency in acquiring scientific knowledge as are the remaining NSF Indicator items.

But nothing more intricate than this discrepancy in the performance of modestly science literate AARs and BARs is necessary to see that these two items aren’t a valid measure of science literacy in the former.

2. The NSF Indicators test of science literacy is far from perfect, but I think it’s reasonable to infer that people who do above average have acquired more understanding of basic science knowledge than those who score below average.

I doubt that a majority of BARs who score in the top 50% of the NSF Indicator battery (sans Evolution and Big Bang and avec the process items) know the basic elements of the theory of evolution, including the role that genetic variance, random mutation, and natural selection play in it.

But I think more of them are likely to understand those things than BARs who score in the bottom 50%.

By the same token, there’s reason to believe that AARs who score in the top 50% on the NSF science literacy test are more likely to have acquired an elementary knowledge of evolutionary theory than those—BARs or AARs—who score in the bottom 50%.   

Nothing in how the above-average science literacy AARs answer the Evolution item furnishes any reason to doubt this. How they respond to that item, I’ve just pointed out, is not, for them at least, a measure of what they know about science.  And in any case, as has been established by researchers on multiple occasions, there’s zero correlation between whether one says one “believes in” evolution and whether can give a passable account of the modern synthesis.

3. Now let’s consider what we can learn from the responses to the “popular opinion poll” item on beliefs in evolution.

That item asks respondents to indicate “which one of the following statements comes closest to your views on the origin and development of human beings—”

• Humans developed over millions of years from less advanced forms of life, but God guided this process

• Human beings have developed over millions of years from less advanced forms of life, but God had no part in this process; or

• God created human beings pretty much in their present form at one time within the last 10,000 years or so.”

Let’s call these responses “Theistic Evolution,” “Naturalistic Evolution,” and “Young Earth Creationism,” respectively.

Theistic Evolution was the most popular response but by was supported by only a plurality (38%). Young Earth Creationism was second and Naturalistic (or “Godless”) Evolution third but the proportions who selected each differed by only a slight amount (32% vs. 29%, respectively).

These numbers, by the way, differ a bit from what Gallup tends to report. The percent selecting Theistic Evolution is in consistent with that. But Godless Evolution runs closer to Young Earth Creationism than it does in Gallup polls.

What to make of this? Well, I’ll write a blog soon about the validity of on-line public opinion samples. But suffice it to say that based on the predictive accuracy of surveys conducted by YouGov, the premier on-line survey firm that recruited the sample for this study, and surveys conducted by Gallup in the 2010 and 2012 elections, YouGov is probably getting closer to the “true” general population values.

What we are interested in, though, is how science literacy and religiosity influence selection of these responses.

Consider first the relationship between these responses & science literacy.

Maybe not shocking but note that support for Naturalistic peaks at only about 55% even among the most science literate. The relationship between support and for that position and science literacy, moreover, is “U”-shaped—higher at both the low and high ends. This relationship was confirmed by a multinomial logistic regression with appropriate quadratic terms; the fitted values from that regression are what I’m graphing (these plots are very true to what one would see in the “raw” data).

Now add religiosity. The following plots contrast the probabilities that AARs and BARs will select one or another of the response to the popular pollster item. They are derived from the same multinomial logistic regression, which confirmed that the impact of science literacy on the probability of selecting one response or another varies depending on level of religiosity.

It’s clear that the “U”-shaped relationship between science literacy and believing in Naturalistic Evolution is being driven by BARs.

In other words, BARs are more likely to believe in Naturalistic Evolution as they become either extremely science literate or extremely science illiterate!

Is this a surprise? Well, I wasn’t expecting this. My inspection of the data was pretty much exploratory, without strong hypotheses.

But I was reminded of a finding in what I regard as one of the very best studies of how high-quality instruction in the teaching of evolutionary theory generates improvements in knowledge but not changes in belief.

In the study, Anton Lawson and collaborators found that high school students, particularly those scoring highest in critical reasoning skills, readily acquired knowledge of various aspects of evolution through instruction, but that acquisition of such knowledge did not produce a corresponding shift in belief among the students who began as nonbelievers.

Nevertheless, the subgroup of such students who did back away from two particular beliefs hostile to naturalistic evolution (that the “living world is controlled by a force greater than humans” and that “all events in nature occur as part of a predetermined master plan”) consisted of the students who scored the lowest in critical reasoning skills.

Speculating on why, Lawson et al. noted that “experience tells us that people change their beliefs for other than rational reasons. For example, hearing the opinion of an acknowledged authority figure could cause one to change a belief. Perhaps intuitive [students] are more likely than reflective students to change their beliefs for this reason.”

Lawson et al. don’t themselves explicitly suggest this, but a consistent conjecture might be that students who are higher in critical reasoning skills might be more inclined to push back on identity-threatening “beliefs” (even while taking on more knowledge) than those who are less reflective. That would be consistent with findings that motivated reasoning can be amplified by science literacy and cognitive reflection.

Someone should do a study to test that hypothesis!

4.  For AARs, in contrast, an increase in science literacy does not predict belief in Naturalistic Evolution. On the contrary, it seems to predict a slight decrease, although the effect is pretty much zero for all but those AARs whose scores are quite low.

So much for the idea that “disbelief” in evolution is a sign of low science literacy.  It isn’t.  “Disbelief” is just as consistent with being high in science literacy as low.

The only thing “disbelief” in Naturalistic Evolution reliably signifies is that one is religious.  This is consistent with the hypothesis that evolution “beliefs” are actually measures of cultural identity (as reflected in religiosity).

This conclusion is strongly corroborated by the relationship between science literacy and the increased probability of believing in Theistic Evolution among AARs. Offered the opportunity—as they aren’t in the NSF Science Indicators science knowledge battery—to select a position simultaneously consistent with “belief” in evolution and religious identity, the most science literate AARS grab hold of it!

5. Indeed, those same subjects—AARs who score high in science literacy—are less likely to espouse Young Earth Creationism than their less science literate counterparts.

What does this tell us? I suppose other interpretations are possible, but I’d say that AARs high in science literacy are in fact eager to affirm their “belief” in evolution, so long as they can be presented with a means of doing so that doesn’t denigrate their cultural identities.

Not surprisingly, BARs also less likely to express support for Young Earth Creationism as they become more science literate.

Support for Young Earth Creationism is associated disproportionately with being simultaneously above average in religiosity and below average in science literacy.

Some concluding thoughts:

1. “Disbelief” in evolution doesn’t reflect a deficiency in science literacy or shortcomings in science education in our society.  

I think it is very reasonable to think members of our society are not as science literate as they should be, and also that our education system must do better in imparting scientific knowledge to citizens generally.

But it’s wrong to think that the level of “disbelief” in evolution is evidence of those things.  It’s wrong to think that because that view is contrary to empirical evidence.

The evidence that many researchers have compiled and that I’ve added to in a very modest way here show overwhelmingly that an individual’s unwillingness to profess “belief” in evolution doesn’t indicate science illiteracy or her unfamiliarity with the rudiments of evolutionary theory.

It measures her expression of her cultural identity. What saying “I don’t believe in evolution” means, culturally speaking, is that one belongs to a community whose members subscribe to a particular set of understanding on best way to live.

2.  Those dedicated to the critical task of promoting scientific literacy, including public knowledge of the best scientific understanding of evolution, should not be focusing on what percentage of the population says they “believe” in evolution.

They shouldn’t be focusing on that because that information tells us nothing about how much scientific knowledge or even knowledge of evolution the public has.  Those who want to test how well society is doing in imparting knowledge of evolution should be measuring instead what fraction of the population can give a cogent account of genetic variance, random mutation, and natural selection. It’s pitifully small, among both those who say they “believe” in evolution and those who say they don’t.

But even more important, those who want to promote public acquisition of scientific knowledge should avoid making professions of “belief” in evolution their aim because doing so is much more likely to deter than promote acquisition of basic scientific knowledge.

People who have a religious identity—who include plenty of science literate people and people capable of becoming even more so—see profession of “belief” as denigrating their cultural identities.  Naturally, then, they will see the demand that they not only learn but publicly affirm their “belief” in evolution as an attack on their community by members of another who harbor a shared understanding of the best life hostile to theirs.

They’ll resent that.  And with good reason. It’s appropriate–absolutely essential, even–that a liberal democracy oblige those who furnish the public good of education to impart to people of all cultural identities the best available understanding of how the universe works, including the career of life on earth.  But citizens who make it their business to force others who have cultural views different from theirs to submit to purely symbolic rituals of identity-abnegation are engaged in a noxious, fundamentally illiberal form of conduct.

Such behavior, moreover, predictably breeds motivated resistance to acquiring knowledge of what science knows. Fear of the loss of status associated with “assenting” to facts symbolically linked to the identity of a rival cultural group is exactly what blocks citizens from converging on the best scientific evidence on issues climate change, nuclear power, the HPV vaccine, and other culturally contested policies.

In their study of how effectively imparting knowledge of evolutionary theory does not produce “belief,” Anton Lawson & William Worsnop conclude:

Of course, every teacher who has addressed the issue of special creation and evolution in the classroom already knows that highly religious students are not likely to change their belief in special creation as a consequence of relative brief lessons on evolution. Our suggestion is that it is best not to try to do so, not directly at least. Rather, our experience and results suggest to us that a more prudent plan would be to utilize instruction time, much as we did, to explore the alternatives, their predicted consequences, and the evidence in a hypothetico-deductive way in an effort to provoke argumentation and the use of reflective thought. Thus, the primary aims of the lesson should not be to convince students of one belief or another, but, instead, to help students (a) gain a better understanding of how scientists compare alternative hypotheses, their predicated consequences, and the evidence to arrive at belief and (b) acquire skill in the use of this important reasoning pattern-a pattern that appears to be necessary for independent learning and critical thought.

This is a sensible prescription for those who (very appropriately!) want to promote the widest dissemination of basic science knowledge in the general public.

But it also happens to be a prescription consistent with the basic liberal injunction to respect the entitlement of individual citizens to freely use their own reason both to understand what is known by science and to decide for themselves what constitutes a virtuous life.

The convergence of the two is not any sort of accident.  It reflects a deep truth about the reciprocal affinity of science and political liberalism.