It’s palpably obvious that acceptance of evolution is impeded by religion. The data are many, including surveys of different denominations, a strong negative correlation among countries between their degree of religiosity and their inhabitants’ acceptance of evolution, the statements of religious people themselves to the effect that evolution threatens their view of scripture, morality, or self-worth, and the fact that creationism throughout the world is always connected with religion. And then there’s this, from an analysis by David Masci at the Pew Forum in 2007:
When asked what they would do if scientists were to disprove a particular religious belief, nearly two-thirds (64%) of people say they would continue to hold to what their religion teaches rather than accept the contrary scientific finding, according to the results of an October 2006 Time magazine poll. Indeed, in a May 2007 Gallup poll, only 14% of those who say they do not believe in evolution cite lack of evidence as the main reason underpinning their views; more people cite their belief in Jesus (19%), God (16%) or religion generally (16%) as their reason for rejecting Darwin’s theory.
It also seems obvious that religion impedes acceptance of not just evolution, but science in general—at least that brand of science, like stem-cell research or work on global warming—that threatens religious views. That conclusion has just been buttressed by a new paper by Darren E. Sherkat in Social Science Quarterly, “Religion and scientific literacy in the United States.” Sherkat’s analysis plainly shows that even excluding issues of evolution, religion in America plays a substantial role in reducing science literacy. (I’m not sure if this paper is behind a paywall. If it is, email me and I’ll send it to you.)
Sherkat took data from the 2006 General Social Survey (GSS) collected by the National Opinion Research Center (NORC) here at the University of Chicago, a survey of 4,510 randomly chosen Americans who were asked questions about their race, income, immigrant status, geographic region of residence, gender, urban or rural home, and so on. To a randomly sampled subset of 1,863 of these individuals, NORC gave a 13-question science literacy exam. Here’s what people were asked:
The GSS employed a 13-question science examination covering: (1) understanding experimental control groups; (2, 3) two questions about probability regarding disease in a brief vignette; (4) knowledge of the core temperature of Earth; (5) understanding that radioactivity is not simply manmade; (6) knowledge of male determination of sex in human reproduction; (7) understanding that lasers are light waves and not sound waves; (8) knowledge that electrons are smaller than atoms; (9) understanding that the Earth revolves around the sun and not the other way around; (10) that a revolution of the earth going around the sun takes a year; (11) that the universe began with a huge explosion; (12) that continents have drifted over time, and continue to move; and (13) understanding that antibiotics do not kill viruses. A question about evolution was eliminated, since the purpose is to see if religious factors have a bearing on scientific understandings outside that controversial realm. The scale approximates one developed by Miller (1998) for the measurement of civic scientific literacy. A reviewer suggested that sectarians and fundamentalists might answer the “big bang” question correctly by interpreting it through the lens of their distinctive faiths; however, that should minimize rather than augment their differences from others.
The GSS also surveyed people about their religious identification and how they interpreted the Bible:
Religious identifications are classified into five broad groups following prior research on U.S. religion (Roof and McKinney, 1987; Sherkat, 2001): (1) sectarian Protestant identifications (Baptists, Pentecostals, Churches of Christ, Nazarenes, etc.); (2) other Protestants (mostly mainline groups such as Methodists, Presbyterians, Episcopalians, and Lutherans); (3) Catholics; (4) non-Christians (including Jews, Buddhists, Moslems, Hindus, and other faiths); and (5) no religious identification. Religious beliefs are gauged using a question identifying whether respondents believe (1) “The Bible is the actual word of God and should be taken literally, word for word”; (2) “The Bible is the inspired word of God, but not everything in it should be taken literally”; and (3) “The Bible is an ancient book of fables, legends, history, and moral precepts recorded by men.” The first answer to this question is commonly used as an indicator of religious fundamentalism.
Sherkat then did statistical analyses of this data to see which factors affected science literacy, and also performed a multivariate analyses to see which factors were important independent of the others. The results, especially for the effect of religion, were striking:
- The percentage of correct answers on the science exam was strongly (and statistically significantly) affected by religious beliefs. Those who take the Bible as the literal word of God scored 54% correct, those who see the Bible as “inspired by God” got 68% correct, and those who see the Bible as a “book of fables” got 75% correct. This classification explained 13% of the total variation in science literacy.
- Dividing up people by religious identification rather than by how they regarded the Bible, we also see strong effects on science scores. Sectarian Protestants scored 55% correct, Catholics 65%, “other Protestants” and non-Christians 68%, and nonbelievers (yay!) 72%. The difference between sectarian Protestants and the others is statistically significant, as is the difference between Catholics and everyone else, though the difference between Catholics and “other Protestants” is a small 3%. All together, these religious identifications explain 15% of the variation in science literacy.
- To put these figures in perspective, race accounts for 9% of the variation in science literacy, education for 20%, income 9%, and gender 4%. Sherkat concludes that “religious factors are as important for predicting scientific proficiency as are many common sociological characteristics such as race, education, income, and gender.”
- One must, of course, control for cross-correlation of factors (for example, perhaps sectarian Protestants are less educated than nonbelievers, and do worse solely because of that) by performing multivariate analysis. When one does this, we still find that sectarian Protestants have significantly lower science literacy than do “mainline” Protestants and nonbelievers. Catholics, too, remain significantly lower in science literacy compared to other Protestants and nonbelievers, but Catholics now don’t differ from sectarian Protestants in their lower degree of science literacy. Remember, this analaysis measures the effect of religious affiliation with all other factors held equal, and these factors include whether or not one has a fundamentalist interpretation of the Bible.
- Biblical interpretation by itself is also significantly associated with science literacy with other factors (like religious identification) held equal. Fundamentalists are less science literate than those who see the Bible as inspired by God, who in turn are less science literate than those who see the Bible as a book of fables.
- Perhaps contrary to popular belief, the South is not a hotbed of science ignorance by virtue of geography alone. When one removes religious factors (for the South harbors more sectarian Protestants and fundamentalists), the South isn’t associated with less science literacy, though rural areas remain less science literate than urban ones.
The gap between sectarians and fundamentalists and other Americans is quite substantial. Indeed, only education is a stronger predictor of scientific proficiency than are religious factors. . . .Scientific literacy is low in the United States relative to other developed nations, and this research suggests that religious factors play a substantial role in creating these deficits. This study adds to a growing body of research demonstrating the importance of religious commitments for structuring stratification outcomes, and pointing to the negative impact of sectarian Christian commitments for life chances.
Catholic deficits in scientific literacy are less pronounced, and mostly arise after controls for education. This suggests that while Catholics have achieved considerable gains in educational attainment (Keister, 2007), their scientific proficiency does not match their educational position. It is possible that Catholic scientific disadvantages are a function of limited scientific offerings in Catholic colleges and high schools. However, the lack of a significant interaction between educational attainment and Catholic identification suggests that Catholics’ social networks may de-emphasize scientific knowledge, and channel intellectual curiosity into other pursuits.
Sherkat, D. E. 2011. Religion and scientific literacy in the United States. Social Science Quarterly 92:1134-1150.