National Study of Religion & Human Origins

Jonathan P. Hill, Ph.D. Assistant Professor, Sociology Calvin College Grand Rapids, Michigan

© 2014 by Jonathan Hill

This research was made possible through a grant from The BioLogos Foundation’s Evolution and Christian Faith program.

But Half a M an/M ich ael Him beault/Flickr/C C BY 2.0

Executive Summary THE NATIONAL STUDY OF RELIGION AND HUMAN ORIGINS (NSRHO) is the first survey to

investigate the public’s beliefs about human origins in depth. The project has two primary purposes. The first is to develop a better map of the public's beliefs by disaggregating the components of the most common survey questions on human origins and focusing on each in detail. The second is to gauge the influence of social context on these beliefs. To accomplish this, survey questions were developed that focused on family, friends, church, schooling, as well as beliefs about the relationship between science and religion more generally. These dual purposes are reflected in the division of the report. The first part uses the disaggregated survey items on human origins to reconstruct various positions. Using the broadest definitions of creationist, theistic evolutionist, and atheistic evolutionist, the NSRHO finds that substantially smaller proportions of the population should be classified in each than the frequently cited Gallup findings suggest. In the NSRHO, thirty-seven percent are creationists, 16 percent are theistic evolutionists, and nine percent are atheistic evolutionists. If we consider only those that are very or absolutely certain of their views this falls to 29 percent, eight percent, and six percent. This tells us that well over half the population are at least somewhat uncertain about what they believe. When additional measurements are used to narrow these definitions, even smaller percentages of the population belong in each camp. The difference between Gallup and the NSRHO is almost certainly due to Gallup respondents being forced to choose from limited options, even when many are unsure of what they believe or maintain beliefs that do not fit into the options available.

The first part of the report concludes with an examination of why some of the public believes it is important to have the correct beliefs about human origins. Thirty themes were found in 947 open-ended responses. The most popular response for creationists invoked the authority of the Bible. In fact, all of the most popular responses for creationists involved protecting and defending the Christian faith. For theistic evolutionists the most popular responses referenced how human origins help us understand ourselves as humans or the future of humanity more generally. Other popular responses emphasized the importance of having true beliefs in all areas of knowledge, the importance of understanding and believing scripture, and having a relationship to God. Overall, theistic evolutionists were the most diverse in their responses. Atheistic evolutionists, on the hand, were the least diverse. The single most popular response referred to the superiority of science as factual and the inferiority of religion as superstitious or irrational. Another common response referred to the negative social and/or environmental consequences of not having the right beliefs about human origins (presumably because this means one has a low regard for science in general). All together, these responses help to make sense of what people believe to be at stake in public controversies over evolution and human origins. The second part of this report shifts the focus to the social context of belief. It is divided into several sections. The first sections examine how various demographic indicators and social contexts differ for creationists, theistic evolutionists, and atheistic evolutionists. Although there are differences in education, political identification, age, and region of the country, these are mostly due to differences in religious beliefs between the groups. In other words, once religion is controlled for, these differences become much smaller and many are no longer statistically significant. When we track the beliefs of close friends and family, creationists are substantially more likely to belong to networks who agree with them about human origins. They are also more likely to expect increased disagreements with family and friends if they were to change their beliefs. Likewise, creationists are more likely to belong to congregations who have settled positions that reject human evolution and to perceive disagreements with religious leaders and other congregants if they were to change their beliefs. Moreover, creationists are more likely to spend their schooling in science classrooms that did not endorse evolution. Even if this is restricted to public high schools and universities, their science classroom experience is different from others. Put simply, creationists are embedded in networks and institutions that are more effective than the other groups in reinforcing the content and importance of their beliefs. This is followed up by a section that situates these various beliefs about human origins in a broader intellectual context. The results tell us that the patterns are complex. Creationists are split in their thinking about evolution and the scientists who study it. While some think evolution is scientifically deficient and that scientists are biased in their beliefs, nearly equal numbers do not think these things. This suggests that there is a divide in some of the reasons why creationists reject evolution. Atheistic evolutionists, on the other hand, overwhelmingly believe that evolution is empirically factual and the same as any other form of scientific knowledge. In one way, both creationists and atheistic evolutionists are the same. Majorities of both groups say that science and religion are ultimately incompatible. When science and religion conflict, one group favors religious ways of knowing (primarily a literal reading of scripture) and the other group favors scientific way of knowing (trusting the mainstream scientific establishment to provide accurate information). Only theistic evolutionists oppose this conflict model in any substantial number. The remainder of the second part focuses on statistical models that predict who ends up being classified as a "confident" member of each group. These models, especially for creationists, demonstrate the power of social context combined with individual belief. In fact, every reliable pathway to creationism involves beliefs in the inerrancy of the Bible as well belonging to congregations who have settled anti-evolution positions. Confident theistic evolutionists are harder to predict overall, suggesting there are numerous, idiosyncratic pathways not well captured by the types of measurements available in the NSRHO. Atheistic evolutionists are most likely to be found outside of institutionalized

2

National Study of Religion & Human Origins

religion, have low regard for the authority of the Bible, do not regularly pray, and are predominately white. These various combinations of factors tend to lead to confident atheistic evolutionist views a disproportionately high percentage of the time. Lastly, the final section examines the factors that lead to a recent change in belief about human origins. Social context matters for this group as well, but in some ways that are different from creationists (families believe differently than them), and in some ways that are the same (they anticipate negative repercussions from family members for changing beliefs). .

The Creation o f Adam /Michelang elo/Pu blic D omain

Introduction THE NATIONAL STUDY OF RELIGION AND HUMAN ORIGINS (NSRHO) is a representative

survey of U.S. adults that examines beliefs about human origins and their social context. As the title of the study implies, there is a strong focus on understanding how religion shapes these beliefs. Nearly all social research on beliefs about human origins has relied on a small number of survey questions that have been replicated dozens of times. These existing questions provide a general picture of various positions that the American public hold and allow for tracking the stability and change in these views over time. Nevertheless, these questions have some serious shortcomings. The more detailed data in the NSRHO show that existing questions lead to oversimplified conclusions about the public's beliefs. Moreover, the social sources of these beliefs have not been adequately examined beyond simple correlations with religious beliefs, political ideologies, and educational levels. The point is not to criticize past studies; rather, it is to provide a better empirical base to understand beliefs about human origins. The survey itself is composed of several sections. The first section contains questions on religious identity, practice, and belief. These are all questions that have been replicated from other reputable national surveys. The next section contains a detailed breakdown of the various beliefs related to human origins. These include beliefs about common descent, divine involvement, and historical timeframe as well as particular questions related to the biblical account of creation (e.g., Whether Adam and Eve were historical figures and whether the days of creation should be taken as 24-hour.). This section also includes open-ended questions for respondents who have changed their beliefs about human origins in recent years to provide an account of this change, as well as for respondents to indicate why having the correct beliefs about human origins is personally important to them. The third section focuses on the influence of congregations, friends and family, and science education. Finally, the survey concludes with a section that includes questions about science and religion more generally as well as the conduct and beliefs of scientists. Standard demographic measures are also included (e.g., income, educational attainment, gender, race/ethnicity, political ideology, martial status, etc.). The full survey instrument is available in Appendix B.

44

N ati ona l Study S tu dy of of Religion Re li gion & & Human H uman Origins Ori gin s National

The survey was fielded by GfK using KnowledgePanel, a representative, address-based sample of American adults. GfK is a respected survey firm that regularly provides services for government and academic research, and KnowledgePanel is the only online, probability-based sample of U.S. adults. The initial survey was pre-tested in June of 2013 and fielded in June and July of 2013. The final sample size was 3,034. All analyses that appear in this report have statistical weights applied so that they best represent U.S. adult population. The field report from GfK, with additional information on methodology and response rates, is available in Appendix C.

A Methodological Note Results are presented graphically whenever possible throughout the report. Bar graphs typically contain error bars that represent the 95 percent confidence interval of the population estimate. Although the survey respondents are a nationally representative sample, random measurement error means we cannot know precisely the true population percentage. The error bars provide a likely range of true population values. For most bar graphs, two types of bars are shown. The bars on the top are labeled "raw percent". These represent the best possible estimate of the true population value (survey weights are applied to these percentages). The "adjusted percent" bars represent the best guess of the true population value after certain potential spurious factors are statistically adjusted for. These factors are age, gender, income, educational attainment, marital status, employment status, race/ethnicity, region of the country, and political ideology. The raw percentages are what are typically referred to in the text. Sometimes the adjusted percentages are substantially different from the raw percentages and additional explanation is warranted.

The Organization of this Report This report is divided into two main parts, each corresponding to different sections of the survey. Part I maps the landscape of beliefs about human origins. Unlike previous surveys, the NSRHO asks multiple questions about human origins. The idea is to "deconstruct" the various components that are often packaged together in survey questions on national surveys. These can then be "reconstructed" in ways that more accurately reflect the range of possible beliefs. In addition to this, respondents are always given the chance to report their degree of certainty for each item. The combination of certainty and specificity in measurement gives a more accurate snapshot of the American public than past surveys and polls. This section concludes with an analysis of open-ended responses to the questions of why it is important to have the correct beliefs about human origins. Thirty general themes are found, with the most prevalent themes unpacked in detail. Once the basic landscape is mapped, Part II focuses on the social context of these beliefs. There are three primary areas of focus: religious congregations, family and friends, and education. Numerous popular accounts either ignore social context entirely, or propose theories about how religion or education influences beliefs that do not hold up under scrutiny. The survey questions on social context in the NSRHO are almost entirely original and thus provide new empirical evidence that further situates the beliefs from Part I. This analysis is followed up by an examination of how beliefs about human origins relate to beliefs about science and scientists more generally. Are some people's beliefs about human origins sustained by meta-beliefs in the fundamental conflict between science and religion? What sorts of positions on human origins are common for those that hold to Stephen Jay Gould's non-overlapping magisteria (NOMA) view? Are scientists influenced by personal beliefs when they conduct science? The second part concludes with statistical models that use the various measures of social context and individual religious belief, practice, and identity to predict which individuals end up as confident proponents of the various views on human origins. These models identity the critical factors, in isolation and in combination, which lead to the different positions.

N atio nal Mu seum of Na tu ra l History/A dam Fo ster/Flickr/C C BY 2.0

Part 1: Constructing Beliefs about Human Origins SURVEYS AND POLLS find the American public divided over human evolution. Most, in fact, find something close to an even split between those who believe that human beings are the product of evolution and those who do not. Since the 1980s, several surveys have had some variation on a dichotomous true/false question asking whether humans have evolved from earlier species of animals. Although the results differ slightly from survey to survey, across surveys approximately 40-45 percent say this is true, 40-45 percent say this is false, and about 10-20 percent do not know or refuse to answer.1⁠ It appears as if there are two, equally-sized, warring camps in this debate. Another common strategy is to ask respondents to select from a list of “packaged” views and ask respondents to select the one that comes closest to what they belief. This is the approach that the Gallup organization has taken (and many other organizations have replicated) twelve times since 1982. The wording is as follows: Which of the following statements comes closest to your views on the origin and development of human beings: (1) human beings have evolved over millions of years from other forms of life and God guided this process, (2) human beings have evolved over millions of years from other forms of life, but God had no part in this process, or (3) God created human beings pretty much in their present form at one time within the last 10,000 years or so? 1

Roughly similar questions have been included in the General Social Survey, the Pew Religion & Public Life surveys, and the Public Understandings of Science & Technology surveys.

66

N ati ona l Study S tu dy of of Religion Re li gion & & Human H uman Origins Ori gin s National

The most popular response is, and has been in every iteration, the third option. Between 40 and 47 percent of the population select the Young Earth creationist position as closest to their view. Slightly below that is the theistic evolution position with between 31 and 40 percent. The last – atheistic evolution – is the least selected option (although growing in popularity), with between 9 and 19 percent affirming that this comes closest to their belief. Again, if we categorize these into pro- and anti-evolution camps, it appears that nearly half the nation affirms evolution and nearly half denies it. Although dozens of national surveys and polls have included questions on human origins, these two questions, with only minor variation, make up the vast bulk of attempts to use survey data to document Americans’ beliefs on the topic.2⁠ These statistics then serve as a springboard for a variety of commentaries on evolution and creationism. The media, who routinely disseminate these poll findings with each press release, use the statistic to publish articles that lament the quality of science education,3⁠ the brainwashing power of religious fundamentalism4⁠ , or the potency of right-wing political ideologies⁠5. In many ways, the cycle of press releases and journalistic articles does not serve us well if we hope to gain an understanding of why such a large percentage of Americans continue to deny human evolution. If we take a closer look at several of these conventional accounts problems begin to emerge. If poor science education is to blame, then why do Americans score near the top in

2

Another frequent survey question asks respondents to assess what should be taught about human origins in public schools. While this is obviously related to what people believe themselves about human origins, it is not an attempt to directly measure it.

3

Politt, Katha. “What’s the Matter with Creationism?” The Nation (July 2-9, 2012).

4

Blow, Charles. “Religious Constriction.” The New York Times (June 8, 2014).

5

Graham, David A. “Why Has Republican Belief in Evolution Declined So Much?” The Atlantic (December 31, 2013).

international comparisons of science literacy?6⁠ Why, in studies of high school and college biology courses, are so few students swayed to believe in evolution even though they are able to accurately describe what it is and how it works?⁠7 Why do conservative Protestants’ beliefs about science, knowledge of science, and likelihood of holding scientific-based jobs appear to be remarkably similar to other Americans (with the exception of beliefs about evolution and age of the earth) if they are being brainwashed?8⁠ All of this indicates that we are far from understanding the most basic social processes that produce beliefs

6

Miller, Jon D. “Public Understandings of Science in Europe and the United States.” Paper presented at the AAAS annual meeting in San Francisco (February 16, 2007). See the following for a summary of the findings: https://www.sciencenews.org/blog/sciencepublic/science-literacy-us-college-courses-really-count

7

Ingram, Ella L., and Craig E. Nelson. “Relationship between Achievement and Students’ Acceptance of Evolution or Creation in an Upper-Level Evolution Course.” Journal of Research in Science Teaching 43, no. 1 (2006), 7–24.; Rice, Justin W., Joanne K. Olson, and James T. Colbert. “University Evolution Education: The Effect of Evolution Instruction on Biology Majors’ Content Knowledge, Attitude toward Evolution, and Theistic Position.” Evolution:Education and Outreach 4, no. 1 (2011),137–44; Lawson, Anton E., and William A. Worsnop. “Learning about Evolution and Rejecting a Belief in Special Creation: Effects of Reflective Reasoning Skill, Prior Knowledge, Prior Belief, and Religious Commitment.” Journal of Research in Science Teaching 29, no.2 (1992),143–66; Bishop, Beth A., and Charles W. Anderson. “Student Conceptions of Natural Selection and Its Role in Evolution.” Journal of Research in Science Teaching 27, no. 5 (1990),415–27; Chinsamy, Anusuya, and Eva Plaganyi. “Accepting Evolution.” Evolution 62, no.1 (2007), 248–54.

8

Evans, John H. “Are Religious Americans Opposed to Science?” Scholars Strategy Network (2013). Available here: http://www.scholarsstrategynetwork.org/sites/default/file s/ssn_key_findings_evans_on_religious_opposition_to_sci ence.pdf

about human origins. Although the results from the NSRHO survey presented here do not settle the matter, they do help us move toward a better understanding. Such an understanding must begin with a better map of the landscape. Survey questions that force respondents to choose between sets of prepackaged positions tell us what positions Americans are drawn to under such conditions, but they do not lay out the possibility that some maintain beliefs that do not easily fit any of the positions, that some are uncertain about what they believe, or that some simply have no interest in the matter at all. It is quite possible, for example, for some creationists to reject evolution, affirm the special creation of humans, and yet not believe this occurred within the last 10,000 years (this position is commonly referred to as Old Earth creationism), yet such an option is not available to respondents on the Gallup survey. To help generate a better picture of the landscape of beliefs, the NSRHO includes separate questions about human evolution, God’s involvement, the manner in which God created, the existence of a historical Adam and Eve, belief in literal 24-hour days of creation, and the geological timeframe for the emergence or creation of humans. Respondent are always given the option to respond that they are unsure about any particular question, and after each question they are asked to rate their level of certainty. After all of these questions, respondents are asked to rate how important it is to them, personally, to have the right beliefs about human origins. If they indicate that it is very or extremely important, they are invited to write a few sentences explaining why this is so. All of these components, together, are analyzed in the first part of this report.

Constructing Creationism Creationists believe that God created humans as part of a single, miraculous act. Humans are a special creation and not the product of a long, "natural" evolutionary process. Although this belief forms the basis of creationism, creationists do not all agree about how and when this happened. Perhaps the most important distinction to make within creationism concerns the timeframe of God's activities. Young Earth creationists believe the days referred to in the creation story should

8

be taken literally, and that God created humanity in the recent past (typically within the last 10,000 years). Old Earth creationists do not take the days of creation literally (the only exception to this is the old Earth view referred to as "gap creationism" which posits millions of years of geological time in between the first two verses of Genesis), and so accept modern science's claims about vast periods of geological time while holding on to the special creation of humans. If we refer back to the Gallup polls, we should expect that between 40 and 50 percent of the population will describe themselves as Young Earth creationists (the Gallup poll does not offer an option that captures the Old Earth creationist view). What happens when we examine these beliefs separately and attempt to construct the various forms of creationism? At its most basic, creationism denies human evolution and affirms that God created humans. If we combine these two positions we find that 37 percent can be classified as creationists. If we restrict this to only those who are absolutely or very certain of their beliefs, we arrive at 29 percent of the population. Of course, we might want to add further restrictions. Creationists, both Old Earth and Young Earth, should believe that God directly and miraculously created a historical pair of individuals (Adam and Eve) that were the progenitors of the entire human race. If we add in these restrictions about how God created and the historicity of Adam and Eve, we will find that 25 percent of the U.S. population should be considered creationists. Twenty-two percent of the population fit this more restrictive definition and are certain of their views. What about timeframe and the days of creation? This, it turns out, is a point where many creationists possess much less certainty (again, this would not be clear using the standard Gallup method). Among the 25 percent in the more restrictive position, only eight percent report that they believe in the literal days of creation and the recent creation of humanity. In other words, only eight percent of the population affirms the main points of Young Earth creationism when asked these questions

National Study of Religion & Human Origins

separately.9 The remaining two-thirds of creationists do not take the Old Earth view however. If we label the Old Earth view as those creationists (using our stricter definition) who do not believe that the days of creation recounted in Genesis were literal days, and who do not believe that humans came into existence in the past 10,000 years, then only one percent of the U.S. adult population can be classified as Old Earth creationists. This is truly an uncommon position to articulate. Many creationists are simply unsure whether the days of creation were literal, and they are especially unsure about when humans first came into existence. Many Americans will select a Young Earth creationist view when presented as a package with other beliefs they are more certain of, but not many are able to clearly affirm the full position when it is deconstructed into its components.

Excursus: What about Intelligent Design? Intelligent design is often classified as a creationist position, but in many ways it falls outside of the typical categorizations. At its most basic, intelligent design attempts to make philosophical and empirical arguments for design by an intelligent agent. Intelligent design is not a position that relies on supernatural revelation or sacred texts to make its claims. Because of this, advocates for this position will typically claim their position is not the same as creationism10. In fact, some leading proponents of intelligent design, such as the biologist Michael Behe, clearly do accept human evolution11. The intelligent design position is mostly concerned with whether purely "natural" mechanisms (such as natural selection) are adequate in and of

9. Only eight percent affirm (a) that humans did not evolve from other species, (b) that God was involved in the creation of humans, (c) that God created directly and miraculously, (d) that Adam and Eve were historical figures, (e) that the days of creation were literal twentyfour hour days, and (f) that humans came into existence within the last 10,000 years. 10. Meyer, Stephen C. "Intelligent Design is not Creationism." The Daily Telegraph (January 29, 2006). 11. See p.2 of Behe, Michael J. The Edge of Evolution: The Search for the Limits of Darwinism. (New York: Free Press, 2007).

themselves organisms.

to

explain

the

complexities

of

living

Although the content of intelligent design does not fit neatly into either creationist or theistic evolutionist positions, as a social and political movement intelligent design clearly has a creationist genealogy. The first mainstream explanation of intelligent design was included in the biology textbook Of Pandas and People12 largely as a response to the legal barring of explicitly creationist positions in public school textbooks. The court’s ruling from the 2005 Dover trial came to the same conclusion. The judge’s decision declared that intelligent design was not able to “uncouple itself from its creationist, and thus religious, antecedents.”13 The public is less familiar with intelligent design than creationism. Because intelligent design takes a different approach than other beliefs about human origins, it is not included in most of this report. While creationists, both Young Earth and Old Earth, clearly fit into the intelligent design camp, and atheistic evolutionists clearly do not, it is not easy to classify theistic evolutionists.

Constructing Theistic Evolution Theistic evolution is a broad category that acknowledges God's involvement in the creation of humans while accepting scientific evidence for evolution. Apart from this there are no additional beliefs that reliably distinguish this view. For example, recalling the discussion of intelligent design, some theistic evolutionists might believe that purely natural mechanisms are not adequate to account for how God created humans and other living things, while others may believe that natural mechanisms are exactly how God created (and some might even object to the way that "naturalism" implies that God is not active). Some could even believe in the direct and miraculous creation of humans in a spiritual sense, even while acknowledging

12. Kenyon, Dean, and Percival Davis. Of Pandas and People: The Central Question of Biological Origins (Dallas: Haughton,1989). 13. Ruling available here: http://ncse.com/files/pub/legal/kitzmiller/highlights/2005 -12-20_Kitzmiller_decision.pdf

the biological connection to primates. Similarly, a belief in a historical Adam and Eve is possible, although not necessary, for theistic evolutionists. However, holding to a theistic evolutionist position does imply disbelief in certain aspects of the Young Earth creationist view. For example, to believe that humans evolved over long periods of time should rule out a literal interpretation of the days of creation in the biblical narrative. Likewise, this should rule out that humanity emerged within the last 10,000 years. Turning to the survey data, it is immediately clear that the theistic evolution position is less popular than creationism. Using the broadest categorization, respondents who (a) believe in human evolution and (b) believe that God (or an intelligent force) was somehow involved in the creation of humans, 16 percent of the population can be placed in this category. Furthermore, only half of this group (8 percent) is very or absolutely certain of both of these beliefs. So although the latest Gallup report shows that 31 percent of Americans are theistic evolutionists, it is clear from the NSRHO that a majority of these are quite uncertain of their beliefs, and many people are choosing this belief because they simply do not like the other options. Similar to what was done with creationism, we can apply a stricter definition to this group. Theistic evolutionists should not believe in literal twenty-four hour days of creation, and they should believe that humans emerged prior to 10,000 years ago. Once again uncertainty is apparent. Only five percent of the population claims humans evolved, God was involved, the days of creation were not literal, and humans emerged more than 10,000 years ago. If we require certainty on all of these points this falls to only two percent of the population. It is clear that the theistic evolution position does not come with a high degree of confidence for much of the population.

Constructing Atheistic Evolution When it comes to beliefs about human origins, atheistic evolution does not necessarily imply a disbelief in God or an Intelligent Force in the universe, it simply means that respondents believe that humans evolved and God had nothing to do with the process. When asked separate

10

questions about evolution and God's involvement in the NSRHO, approximately nine percent of the population can be classified as atheistic evolutionists. Like theistic evolutionists, this group measures out to be about half the size of the comparable category from the latest Gallup poll. This tells us that a sizable number of the Gallup respondents who choose the atheistic evolutionist option are likely unsure of their beliefs about human origins. If we include a measure of certainty, only six percent of the population reports that they believe in human evolution, they believe God had nothing to do with it, and they are very or absolutely certain about both of these beliefs. We should also expect that atheistic evolutionists should reject a belief in Adam and Eve, literal days of creation, and the recent emergence (with the last 10,000 years) of humans. Although most atheistic evolutionists reject the first two, there appears to be some uncertainty about the third. Sixty-four percent believe humans emerged more than 10,000 years ago, 22 percent believe less than 10,000 years ago, and 14 percent are unsure.

Summing up so far In sum, disaggregating the components of beliefs, allowing respondents to report that they are unsure, and having respondents rate their certainty, all show that much smaller proportions of the population are able to successfully articulate the basic components of the dominant positions about human origins, and many are uncertain about what they do believe. Using the most generous definitions, the NSRHO finds that 37 percent of the population can be considered creationists, 16 percent can be considered theistic evolutionists, and nine percent can be considered atheistic evolutionists. This leaves 39 percent of the population as unsure or holding uncommon views (such as believing humans did not evolve from earlier species while simultaneously believing that God had nothing to do with the emergence of humans). If we adopt more restrictive definitions, these numbers begin to shrink further. Still, there are some general patterns that become clear. By comparing the Gallup findings to the NSRHO, we can

National Study of Religion & Human Origins

see that although all groups are smaller in the NSRHO, theistic evolutionists and atheistic evolutionists are both proportionally around half the size in the NSRHO compared to the most recent Gallup poll (31 percent versus 16 percent, and 19 percent versus nine percent respectively). Creationists are smaller in the NSRHO, but only slightly so (42 percent in the Gallup poll versus 37 percent in the NSHRO). This suggests that creationists are more certain of their beliefs than others when presented with the forced-choice option. We can look at this still another way. The three figures (Figures 1 – 3) that follow show the percent of each human origins group, using the broadest definition, who affirm the historicity of Adam and Eve, the literal 24-hour days of creation, and the recent creation of humans (within the past 10,000) years. The historicity of Adam and Eve clearly line up as expected for creationists (who overwhelmingly affirm this) and atheistic evolutionists (who overwhelmingly reject this). Theistic evolutionists are evenly split on this. The belief in 24-hour days of creation is slightly different. Although there is a strong correlation with beliefs about human origins, a substantial number of creationists (more than one-third) reject the literal days of creation. Theistic evolutionists are also more likely to reject this view, with less than onethird agreeing. Atheistic evolutionists almost universally reject it. Lastly, the question about when humans emerged is far murkier. Young Earth creationists should hold to a belief that humans emerged in the recent past, but this is not a majority view. If we include this as part of our measure of Young Earth creationism, then this drastically shrinks their numbers when compared to the Gallup poll. The data suggest this is because many people, both creationists and evolutionists, are simply uncertain about the timeframe. Nearly one-third of creationists report being unsure about this. Atheistic evolutionists are the most certain (with 64 percent believing that humans are older than 10,000 years), but still 14 percent report being unsure. Although most creationists will select the recent emergence of humans when packaged with other beliefs they want to affirm, when these beliefs are disaggregated they are far less likely to select the option that corresponds with the Young Earth creationist viewpoint.

F IGURE 1 . Belief in a historical Adam and Eve

F IGURE 2 . Belief in 24-hour days of creation

F IGURE 3 . Belief in recent creation/emergence of humans (within last 10,000 years)

With the clear exception of the timeframe, the overall confidence of creationists is corroborated when we examine the measure of confidence in the two questions that make up the broadest definitions of these groups: belief in human evolution and belief about God's involvement. Figure 4 shows the level of certainty for each group. Creationists are the most certain (78 percent), followed closely by atheistic evolutionists (70 percent). Theistic evolutionists are the least certain, with only 52 percent reporting that they are very or absolutely sure of these beliefs.

In Their Own Words: Why is it Important to Have the Right Belief? We've seen so far that there are various levels of confidence and coherence in the public's views on human origins. Certainly one of the reasons for this is that the level of personal investment in this issue also varies substantially. Some people simply do not care about this, while others care passionately. The NSRHO contains a separate question that asks respondents to indicate if having the right beliefs about human origins is personally important to them. Forty-two percent of the population indicates that it is very or extremely important to have the right beliefs (versus somewhat, not very, or not at all important). As Figure 5 shows, this is not at all equally distributed among the various groups. Nearly two-thirds of creationists say it is important to have the correct beliefs. About one-half of atheistic evolutionists say the same, and only a little more than one-third of theistic evolutionists agree that it is important to have the correct beliefs. What is at stake for those who say that having the right beliefs is important? For the 42 percent (N= 1,285) of respondents that said having the correct beliefs was very/extremely important, the survey included an open-ended follow-up question that asked why this is an important issue to them. Fully 298 respondents chose not to provide an open-ended response, and an additional 40 respondents indicated that they did not want to answer in their written response (16 cases) or provided very unclear answers (24 cases). The remaining 947 cases are analyzed in what follows. Coding the answers required carefully reading responses to look for similarities and themes. In the end, thirty categories were developed. Categories are not mutually

12

F IGURE 4 . Percent who are absolutely or very certain of their beliefs

F IGURE 5 . Percent who report having correct beliefs about human origins is very or extremely important to them personally exclusive, and single responses frequently captured multiple themes. Appendix A, which details the various methods used in this report, provides additional information on the coding process. For each group -- creationists, theistic evolutionists, and atheistic evolutionists -- I will highlight the themes that at least ten percent touched upon, including providing some examples to get a sense of the way these answers were phrased. For the rest of the responses, I include a table that orders all the themes coded for each group from most common to least common. The only

National Study of Religion & Human Origins

requirement is that at least one percent of a group’s responses must fit the theme to be included.

The Most Common Creationist Responses Although creationists provide many types of answers to this question, responses that referenced religious faith and beliefs were dominant. The most common type of response, which 22 percent of creationists provided, referenced the importance of biblical authority. Taking scripture seriously, for those that gave this response, involves reading the account of creation in Genesis literally. Respondents wrote such things as "God's Word is the answer to all our questions...I believe in God's Word." Or, "You either BELIEVE the Bible or you don't." Or, "This is an essential belief when following the teachings of the Bible. This is a cornerstone of the one and only way to God." Many also provided some sort of variation on the "slippery slope" argument when it comes to biblical authority: "If we do not believe God's Word and his promises in their entirety, there is no basis for our faith and our salvation." Another wrote, "Without the foundation of my biblical faith the entire structure crumbles to nothing." Still another spells out the consequences of not fully believing the Bible on origins: "I believe the Bible is the inerrant Word of God. If I disagree with what the Bible says about the origins of humankind, I open the door to question other parts of the Bible, which I do not find acceptable." The second-most popular theme does not necessarily mention the Bible directly, but includes some reference to God creating humans, or God creating everything. Approximately 20 percent of creationists gave this type of response. The implication, although it is rarely made explicit, is that if one believes God creates, one cannot also simultaneously believe that evolution occurred. The following are typical responses: "We are created by God...I believe this, and plan to spend eternity with my creator when my life on earth is over." Another creationist wrote, "There is only one God, he created us and loves us and we need to respect that love!" Still another wrote that human origins are important "because we did not come from evolution. God created all beings."

The third-most common response mentions something about the respondent's faith in God or personal relationship with God (18 percent of creationists' responses can be classified as such). Again, the implication here is that faith in God is automatically opposed to believing in evolution. Respondents typically do not spell this out (although a few do). Some representative responses are as follows: "I am a person of faith, this is important." Or, "By having the right beliefs about evolution and human origins, I am able to have a personal relationship with the creator of all things: God." And another: "I believe in God. I would be devastated if what I believe was not true. I know it is." The fourth-most common response, with a significant drop in the number choosing to highlight this theme (10 percent), are creationists who mention something about the importance of personal salvation or the afterlife. Some typical responses were: "Because our belief determines our eternal destiny." Or, "I believe there is a God who watches over us and will take us, when we die, to a paradise where we will live forever in peace and harmony." Many of these responses also emphasized the importance of the Bible, linking accepting the authority of scripture to salvation: "My eternity is dependent on the truth of the Bible." Or, "I believe that the Bible is the infallible Word of God. I believe everything in the Bible is true therefore I have accepted God’s redemption for me through Jesus Christ and have accepted Him as my personal savior. Jesus is the only way to heaven. There is no other way. To me it is not optional to not believe the Bible in its entirety." All four themes are clearly linked to religious faith. In each case they at least imply that having the wrong beliefs threatens either the personal dimension with God (a relationship with God or personal salvation) or irrevocably damages key components of the Christian faith (the authority of the Bible or the belief that God is the source of all creation). In any case, they view acceptance of evolution as having dire religious consequences. The remaining themes that were mentioned by creationists are provided in Table 1 (it is important to remember that these themes are not mutually exclusive, so percentages add up to more than 100). The first

responses that are not exclusively religious mention the importance of having true or accurate beliefs (with eight percent), a belief that other groups share as well (12 percent of theistic evolutions and 16 percent of atheistic evolutionists). Protecting and defending faith is clearly front and center for creationists. Other concerns, such as negative social or environmental consequences, or that evolution conceives of humans as accidents, are not the first things that come to mind for most creationists.

The Most Responses

Common

Theistic

Evolutionist

As will be clear throughout this report, theistic evolutionists are less monolithic as a group. Compared to both creationists and atheistic evolutionists, theistic evolutionists provide a greater variety of answers to this question. Interestingly enough, two of the four types of answers that at least ten percent of this group provides, were also above ten percent for creationists. One of the four was also common for atheistic evolutionists (and used by creationists, but not as frequently). The most common answer made reference to how having the right beliefs about human origins helped to better understand oneself and/or the future of humanity. Fifteen percent of the theistic evolutionist answers could be classified as such. The following illustrates this type of response: "It is very important because you have to understand how humans were made and where they came from to understand how humans work and how they are developing/changing and how that can impact our future and the future of our children and grandchildren." Another example would be, "To know where we came from and how we got to the place we

14

Theme

%

Biblical Authority/God says so God created humans/everything Because of my faith/relationship to God Salvation/afterlife It is foundational to truth/faith To have true/accurate beliefs Provides meaning/purpose in life Understanding origins helps understand self and future Because it influences your beliefs about God or worldview Evidence or reason supports my view Guides how we live our life Desire for certainty/comfort/peace in life Wrong belief has negative personal consequences Because humans did not evolve from monkeys/apes Wrong belief has bad social/environmental consequences Pass beliefs/knowledge on to children/future generations I just believe/I just do Because I was taught this Gives glory to God Right to believe what I want Evolution says humans are accidents Humans have souls/different from animals

22 20 18 10 8 8 5 5 5 5 5 4 4 3 3 3 3 2 2 2 1 1

TAB LE 1. Frequency of themes in creationists’ answers to why it is important to have the correct beliefs about human origins

are." Yet another, "I'd like to know where I came from and hopefully it can tell me something about my purpose here on Earth." The next most frequent theme is the importance of having true or accurate beliefs (12 percent). Some examples of this include: "Knowing the truth is important to me." Or, "It is important to understand where we come from in a true, factual sense." Or, "For the sake of intellectual honesty, it is important to understand as much as possible. It affects how I view people and God."

National Study of Religion & Human Origins

Theme

%

Understanding origins helps understand self and future To have true/accurate beliefs Biblical Authority/God says so Because my faith/ relationship to God Because science and faith are compatible Believe in facts and avoiding superstition/irrational beliefs Right to believe what I want Because God created humans/everything Importance of science/scientific progress Wrong belief has negative social/environmental consequences Guides how we live our life Evidence or reason supports my view Because it influences your beliefs about God or worldview Gives glory to God Important to be considered intelligent/educated Curious/ interesting/ important Provides meaning/purpose in life Wrong belief has negative personal consequences I just believe/I just do Humans have souls/different from animals Evolution says humans are accidents It is foundational to truth/faith Salvation/ Afterlife Superiority of science for knowledge Pass beliefs/knowledge on to children/future generations Desire for certainty/comfort/ peace in life It is foundational to good science Because of my occupation [all scientists or pastors]

15 12 12 12 9 8 8 6 6 6 5 5 5 5 4 3 3 3 3 3 2 2 2 2 1 1 1 1

TAB LE 2. Frequency of themes in theistic evolutionists’ answers to why it is important to have the correct beliefs about human origins It is notable that the two most frequent responses are not explicitly about faith or religion (even though these respondents all believe that God created humans using evolution). The next two, however, are. Both of these are shared with creationists, although the content of these answers differs in important ways. The third most popular answer (12 percent) is the importance of the Bible. Some answers that referenced Christian scripture were very articulate about science and the Bible: "I think it is important to trust the Bible. It is also somewhat hard to argue with science's evidence of evolution." Another response, "I believe that Genesis shows the order in which everything was created. First the sun ignited, the earth was covered with water, the dry land appeared, life started in the ocean, crawled onto dry land, birds (dinosaurs) were created, mammals were created, and then man. I believe in evolutionary creation by God." Others were briefer: "I don't want to doubt God's

Word." Or, "The God of the Bible is your basic foundation to live by." The final most popular response (also 12 percent) referred to the respondents' faith or relationship to God. As with the previous responses, some gave quite articulate answers about the relationship between faith and science: "I believe that man came about through evolution, however I also believe that God had an important hand in this. I need to believe in God as a means of support and strength through prayer." Most, however, give responses like the following: "I need to be right with God." Or, "I believe that God loves us all and we should follow him as much as we can. As long as you believe he will be there at the end for you." Or, "Must keep God in it.". The remaining coded responses, listed in order of frequency, are available in Table 2. It is worth pointing out that the next most common response for theistic evolutionists (at 9 percent, just under the 10 percent threshold) directly mentions the compatibility of science and faith. This was almost entirely absent from the other two groups. It is also worth pointing out that, even though some answers were shared between theistic evolutionists and creationists, theistic evolutionists do not share the same concerns about the afterlife or salvation when discussing the importance of human origins (only 2 percent mention anything about this). This seems to be an important theological distinction between the two groups. One group believes salvation is at stake, while the other does not.

The Most Responses

Common

Atheistic

Evolutionist

Lastly, we can assess the most common answers for atheistic evolutionists. As a reminder, this group is not necessarily comprised of actual professed atheists, the label simply means they do not believe that God or some Higher Power was involved in the evolutionary process. Among this group, there was a single type of answer that more than one-third of all respondents provided: Having the right beliefs about human origins is important because facts are superior to superstition and irrational beliefs. The following are examples of this very common

response: "As a scientist, I base my beliefs on factual evidence and not un-provable faith." Or, "Because science is factual whereas religion is not. It's important to understand science and the world around us to continue to survive on this planet and exist with nature." Or, "I take some pride in being a rational being. To deny the clearly evident would diminish me. If someone wishes to believe in the supernatural, this is no concern of mine as long as they do not attempt to force their ‘faith’ on others." Or, "I believe in facts, not made-up fairy stories about an invisible being who magically created humans from nothing. It's a fantasy and an excuse that's used to hurt people who believe differently." All of these responses draw a sharp distinction between the sort of knowledge about the world that science gives and the sort that religion gives. A close follow-up to the fact/belief distinction is the negative social and/or environmental consequences that having the wrong beliefs about evolution will have (27 percent). Many of these responses make an argument that failure to believe in evolution will diminish the importance of science in society more generally and endanger the environment (particularly on issues of global warming) and the future of humanity. Some provided quite detailed accounts of this logic: "Having the wrong beliefs about evolution and human origins leads to the dissemination of wrong information (via individuals, churches, and in places where church and state are not kept separate, public school education on intelligent design, etc.). In turn, these beliefs often, though not always, lead to movements and policies that are discriminatory and detrimental to individuals and society: antigay-marriage, anti-abortion, etc." The following are more typical: "If you have faith in something that isn't true, it can have negative effects on society." Or, "I feel like people fail to believe in science over religion and therefore don 't take responsibility for the upkeep of the planet." Or, "I believe in science and find it frustrating that others do not. This disregard for science is affecting decisions made today regarding global warming, as one example. We must learn from our previous mistakes."

16

The next two most common responses (both 19 percent) consist of one that emphasizes the importance of science or scientific progress, and the other that emphasizes how science is a superior way to learn about the world. The following is a somewhat lengthy response, but represents these viewpoints well: “If we have a scientific method, and we use it for our understanding of our known universe, yet forgo that same method for convictions without any evidence, we're literally stalling progress in seeking knowledge, invention, and progress. Ancient religions are ancient for a reason--they were based only in philosophy, and what was concluded as truth was based on a limited understanding of science. Modern day religions are no different; the only thing that has changed is the methods to circumvent logic and reason. Having a personal belief is okay. However, to completely dismiss scientific theories, which are the highest threshold of truth in science, is harmful to the community. Gravity and evolution share the same space. You cannot reject one without rejecting both.” The next most popular response is one shared by both theistic evolutionists and creationists: to have true or accurate beliefs. Getting at the truth of the matter is important in its own right. Typical responses are as follows: "I like to understand and be correct in my beliefs of the world." Or, "It is important to know the truth about any critical human issue." Or, "I want to base my beliefs on the best science available." Lastly, 11 percent of atheistic evolutionists noted that having incorrect beliefs could have negative impacts on the individual holding these wrong beliefs. This is not as popular as the answers that specify negative societal or environmental impacts, but they still cross the 10 percent threshold. And, most responses that mentioned negative personal consequences also included negative social or environmental consequences. Some examples of this response are as follows: "Ignorance causes a person not to live to full potential." Or, "A person's views on evolution and human origin tells a lot about their critical thinking skills and thoughts on science." Or, "Because people who allow ignorant mythological beliefs replace

National Study of Religion & Human Origins

Theme

%

Believing in facts and avoiding superstition/irrational beliefs Wrong belief has negative social/environmental consequences Importance of science/scientific progress Superiority of science for knowledge To have true/accurate beliefs Wrong belief has negative personal consequences Understanding origins helps understand self and future Important to be considered intelligent/educated Right to believe what I want Evidence or reason supports my view Provides meaning/purpose in life Guides how we live our life Because it influences your beliefs about God/worldview It is foundational to good science Desire for certainty/comfort/peace in life Because science and faith are compatible It is foundational to truth/faith

34 27 19 19 16 11 9 7 6 5 4 4 4 3 3 1 1

TAB LE 3. Frequency of themes in atheistic evolutionists’ answers to why it is important to have the correct beliefs about human origins reason tend to force their ignorance on others in the classroom. Overall it lowers our society down to a third world mentality and inhibits constructive growth." The remaining classifications, in order of most frequent to least frequent, are listed in Table 3. Even when compared to creationists (and especially when compared to theistic evolutionists), atheistic evolutionists give a more limited range of answers to this question.

The Creatio n of E ve/M ich elangelo /Public D om ain

Part II: Mapping the Social Context STUDIES IN THE SOCIOLOGY OF KNOWLEDGE emphasize how our particular social location –

the bundle of social identities and group memberships that give us our unique place in broader society – shapes even our most “common sense” interpretations of the world around us.14 What seems natural and normal to us cannot be assumed to be so for others. We all belong to communities of belief that shape how we see our world. Consequently, we will work hard to maintain certain interpretations of the world against threats because our very identity and social relationships are at stake.15 Unfortunately, this perspective is nearly absent from studies of beliefs about human origins. Both popular accounts16 and scholarly studies17 work from a model that tacitly assumes better science education is the solution to the high rates of opposition to evolution in the United States. While quality science education is a worthy goal in any society, there is little indication that it makes a substantial difference in convincing Berger, Peter and Thomas Luckmann. The Social Construction of Reality: A Treatise in the Sociology of Knowledge (New York: Doubleday, 1967) 15 Magee, Michael W. and Curtis D. Hardin. “In Defense of Religion: Shared Reality Moderates the Unconscious Threat of Evolution.” Social Cognition 28, no. 3 (2010), 379-400. 16 For examples, see Coyne, Jerry A. Why Evolution is True (New York: Penguin, 2009); Pigliucci, Massimo. Denying Evolution: Creationism, Scientism, and the Nature of Science (Sunderland, MA: Sinauer Associates, 2002). 17 For an example, see Williams, James D. “Belief Versus Acceptance: Why Do People Not Believe in Evolution?” BioEssays 31 (2009), 1255-1262. 14

18 18

National N ati ona l Study S tu dy of of Religion Re li gion & & Human H uman Origins Ori gin s

individuals that their views on human origins are wrong.18⁠ At a psychological level, the science education model assumes humans are fundamentally rational actors that simply need more and better information. But this does not fit with the evidence. Recent studies in cognitive science have emphasized the importance of non-rational mental mechanisms when people come to accept or reject beliefs about evolution. One such study emphasized the importance of affective, “gut” reactions in student understanding of evolution.19 Another found that a sizable portion of the college student population values affective epistemic goals (i.e., believing what “feels” right) over goals of accuracy.20 Psychologists call this mixture of affect with reasoning “motivated reasoning”.21 While we are beginning to recognize the importance of powerful emotions in shaping what we believe about human origins, we have yet to explore how social For example, a close look at one of the most widely cited studies on the public understanding of evolution finds no impact from educational attainment on belief in evolution. See Miller, Jon D., Eugenie C. Scott, and Shinji Okamoto. “Public Acceptance of Evolution.” Science 313 (2006):765–66. Berkman and Plutzer also summarize a number of studies and suggest that intelligence, science knowledge, and educational attainment may have a weak association with beliefs about evolution, but only at the upper-tail of the distribution (i.e., only the very educated and knowledgeable seem to show a difference). See Berkman, Michael, and Eric Plutzer. Evolution, Creationism, and the Battle to Control America’s Classrooms (NewYork: Cambridge University Press, 2010).

18

19

Minsu Ha, David L. Haury, and Ross H. Nehm, “Feeling of Certainty: Uncovering a Missing Link Between Knowledge and Acceptance of Evolution.” Journal of Research in Science Teaching 49 (2011), 95121. 20 Griffin, Thomas. D. “Individual Differences in Epistemic Goals and the Acceptance of Evolution.” In D.S. McNamara & J.G. Trafton (Eds.) Proceedings of the Twenty-ninth Annual Conference of the CognitiveScience Society (Mahwah, NJ: Erlbaum, 2007), 1765. 21 Kunda, Ziva. “The Case for Motivated Reasoning.” Psychological Bulletin 108, no. 3 (1990),480.

contexts produce these powerful emotions. What contexts heighten motivated reasoning on the one hand, and what contexts defuse these emotions and open up the possibility of changing beliefs on the other? The psychology needs to be understood within a broader sociology. Without such an account, it is difficult to adequately explain how certain religiously endorsed explanations for human origins trump scientific accounts for some people but not others. In this section we will explore how friends, family, religious congregations, and schools influence beliefs about human origins. The first several sections document how the social contexts, experiences, and beliefs of creationists, theistic evolutionists, and atheistic evolutions vary. This is done using the broad definitions of each of these groups established in the previous section. The goal is to get a sense of what social factors tend to correlate with the various positions on human origins. The ultimate goal, however, is to move toward an explanatory account of how these social contexts contribute to some being firmly planted in their beliefs about human origins while others end up being open to change. The remaining sections of Part II contribute to this second goal by running multivariate statistical analyses that use measures of social context and individual identity to predict who ends up being a confident creationist, theistic evolutionist, and atheistic evolutionist. The final section uses the same methods to try to predict who has recently changed their beliefs about human origins.

Views of Human Origins by Religion, Politics, Education, Age, and Region. We begin with some basic background factors that existing research has suggested are associated with beliefs about human origins. The most obvious and wellestablished connection is religion. The figures below show how creationists, theistic evolutionists, and atheistic evolutionist vary on standard social scientific measures of religion. The first figure (Figure 6) shows the percent of each group that identifies

F IGURE 6 . Percent evangelical Protestant (determined by denomination) with an evangelical Protestant denomination.22 Forty-one percent of creationists identify as such, with theistic evolutionists (16 percent) and atheistic evolutionist (four percent) both being considerably less likely to. Similar patterns can be seen across all of the religious measures. In Figure 7, half (51 percent) of creationists attend church weekly, while only one quarter (25 percent) of theistic evolutionists, and just a sliver (two percent) of atheistic evolutionists attend church weekly. Creationists are also much more likely to report that their faith is very or extremely important in their day-to-day life (Figure 8) compared to atheistic evolutionists (79 versus six percent). Theistic evolutionists fall between the two (48 percent). Private religious practice looks nearly identical to importance of faith (Figure 9). Nearly seven out of ten creationists pray daily (69 percent) while only four percent of atheistic evolutionists pray daily. Again, theistic evolutionists fall between both groups (41 percent).

22. There are several dominant ways to measure religious tradition. The measure of evangelicalism I use here is the most widely adopted one that relies upon a detailed analysis of religious denomination: See Steensland, Brian, Jerry Z. Park, Mark D. Regnerus, Lynn D. Robinson, W. Bradford Wilcox, and Robert D. Woodberry. "The Measure of American Religion: Toward Improving the State of the Art." Social Forces 79, no. 1 (2000): 291–318.

20

F IGURE 7 . Percent attending church weekly or more

F IGURE 8 . Percent reporting faith is very or extremely important in day-to-day life

F IGURE 9 . Percent who pray daily

National Study of Religion & Human Origins

F IGURE 1 0 . Percent who believe the Bible is the Word of God without any errors (includes those who endorse a literal reading and those who say some parts are to be read symbolically) Perhaps the most important factor is beliefs about the Bible. As we already explored in Part I, one of the dominant reasons creationists reject evolution is because they perceive that it will undermine the authority of scripture. This is clear in the survey question about Bible beliefs as well. In Figure 10, we see that fully 81 percent of creationists believe the Bible is the literal or inspired word of God and contains no errors. Only four percent of atheistic evolutionists affirm this same belief. Theistic evolutionists are split down the middle with 44 percent agreeing. It is important to note that, in all of these measures of religion, the difference between the adjusted percent and the raw percent is fairly small. This is because religion is a primary factor that drives many of the differences we find between creationists and evolutionists. The factors that we turn to next, political beliefs, education, age, and region of residence are influential, but primarily because these factors correlate with religious identity, belief, and practice. Once religion is accounted for, the independent effect of these demographic factors tends to be less dramatic. To demonstrate this, the remaining graphs have an extra set of bars. The first set of bars below the raw percent control for all of the religion variables just examined. The bar below this adds the remaining standard control variables.

F IGURE 1 1 . Percent politically conservative

Let's examine political ideology (Figure 11). It is clear that creationists are much more likely to identify as political conservatives (54 percent), compared to theistic evolutionists (32 percent) and atheistic evolutionists (18 percent). But why is this? The bars that adjust for religion show these differences are primarily due to religion, and not politics. Political conservatism is associated with religious conservatism. This second set of graphs tells us that if you hypothetically made each group religiously similar, we wouldn't expect to find substantial political differences. To be clear, political conservatives are more likely to be creationists, but this is largely because they are also religiously conservative. If you have a political conservative who is not religious, then the influence of politics on beliefs about human origins is very modest. In the final set of bar graphs which adjust for all of the control variables, the expected political "gap" between creationists and atheistic evolutionists changes from 32 percent (with no controls) to 12 percent (with full controls). Similar patterns are found with other measures. Figure 12 shows that creationists are older, on average, and atheistic evolutionists are younger, on average, but this is due almost entirely to the fact that older respondents are more religious than younger. Moving on to Figure 13 which displays the percent with a bachelors degree or more, the apparent gap between creationists and atheistic evolutionists of 24 percent drops to a mere 6 percent (and is not statistically significant) once religion

F IGURE 1 2 . Percent age 50 and above

F IGURE 1 4 . Percent living in the South faith (the only exception to this is education, which is strongly associated with income in addition to religious factors). The story is almost entirely about religion, and has little directly to do with such things as politics, age, education, and geographic location.

The Influence of Friends and Family

F IGURE 1 3 . Percent with a bachelor’s degree or more and other factors are accounted for. Lastly, the percent that live in the South for each group is shown in Figure 14. Again, although it appears that creationists are disproportionately in the South, while atheistic evolutionists are not, this is almost entirely because religious individuals are also disproportionately located in this region (as the adjusted percentages demonstrate). In short, religious identification, belief, and practice are unevenly represented across the different positions on human origins. Moreover, even though creationists appear to be less educated, more political conservative, older, and more likely to live in the South, this is almost entirely due to the fact that those very same factors are associated with stronger and more conservative religious

22

Do creationists have different interactions with friends and family than evolutionists? The graphs below tell us that this is almost certainly the case. The NSRHO included two questions about how similar the beliefs of friends and family are to the respondents’ beliefs. Respondents could choose from four answer categories that ranged from "all" believe the same to "none" believe the same (they could also select that they had no immediate family or that they had no close friends). Figure 15 below shows differences between groups on how they answered the question about their immediate family. Sixty-eight percent of creationists reported that all or most of their family members shared the same beliefs as them. Similar figures for theistic and atheistic evolutionists are 20 to 24 percentage points lower. A very similar gap is evident for close friends. Figure 16 shows that fully 56 percent of creationists say all or most of their closest friends share the same belief. The comparable figure for theistic evolutionists is 38 percent and 37 percent for atheistic evolutionists.

National Study of Religion & Human Origins

F IGURE 1 5 . Percent reporting all of most of immediate family share the same beliefs on human origins

F IGURE 1 7 . Percent reporting that changing beliefs would cause increased disagreements with immediate family (only includes those who report having family)

F IGURE 1 6 . Percent reporting all of most of closest friends share the same beliefs on human origins

F IGURE 1 8 Percent reporting that changing beliefs would cause increased disagreements with closest friends (only includes those who report having friends)

While it is clear that creationists are embedded in family and friendship networks that are more likely to agree with their own views, it is not clear that the social pressure to believe one thing or another is any different in these groups. To gauge this, the NSRHO includes questions about the expected social consequences of changing beliefs for each of these groups.

creationists are substantially more likely (35 percent) to report this than either theistic evolutionists (15 percent) or atheistic evolutionists (20 percent). The pattern is true for friends as well, although it is a little less pronounced. Figure 18 shows that 28 percent of creationists report anticipating increased disagreements with close friends if they changed their position, while only 17 and 18 percent of theistic evolutionists and atheistic evolutionists report the same.

Figure 17 shows the percentage of each group who report that there would be "increased disagreements" with immediate family if they were to change their beliefs about evolution. It is worth noting that it is only a minority of each group who reports anticipated difficulties with family members over this issue. Still,

F IGURE 1 9 . Percent who belong to a congregation that rejects human evolution (churchgoing population only) In sum, creationists are both more likely to be in family and friendship networks that believe the same as them, and they are more likely to experience social pressure to maintain their current beliefs.

The Influence of Church What about the influence of religious congregations? As has already been documented, creationists are more likely to be religious on all measures (and are more conservative in their religious beliefs overall). But what happens if our analysis is limited to those who at least semi-regularly attend religious services (i.e., those who attend at least monthly)? Are the types of congregations that creationists find themselves in different from the other groups? The NSRHO includes a question asking respondents if their congregation has a settled position on human evolution. According to the survey, about half of the congregations the respondents regularly attend have a settled position against human evolution. The other half is split evenly between those that do accept it and those that do not have a settled position on the matter. But, as can see below in Figure 19, these answers are not evenly distributed between the three groups. Somewhat unsurprisingly, creationists find themselves in religious congregations where, about two-thirds of the time,

24

F IGURE 2 0 . Percent hearing about human origins during worship or education at their congregation at least occasionally (churchgoing population only) human evolution is a rejected position. Other groups all fall well below one-third. 23 Where do they hear about human origins at church? Again, slightly shy of two-thirds of creationists report hearing about human origins during formal worship or religious education at least occasionally (Figure 20). Slightly more than half of theistic evolutionists report the same, while only slightly more than a quarter of atheistic evolutionists report this. Somewhat surprisingly, atheistic evolutionists are the most likely to report hearing congregants or parishioners informally talking about human origins according to Figure 21 (although the large range of the error bars signifies that the uncertainty of this estimate; only a small number of atheistic evolutionists attend a religious service at least monthly). Creationists and theistic evolutionists are only seven percentage points apart with 61 percent of creationists 23. It should be noted that the estimates for atheistic evolutionists show a high degree of uncertainty because there are so few atheistic evolutionists who attend a religious congregation monthly or more (22 in the entire sample to be precise). It is also worth noting that being labelled an "atheistic evolutionist" does not necessarily make one a full-blown atheist. This simply means that they believe in evolution and also believe that that God had nothing to do with the emergence of humans.

National Study of Religion & Human Origins

F IGURE 2 1 . Percent who report congregants or parishioners talking about human origins at least occasionally (churchgoing population only)

F IGURE 2 3 . Percent reporting that changing beliefs would cause increased disagreements with religious leaders (churchgoing population only) with other church members if they changed their position on evolution (46 percent). Only about a quarter of theistic and atheistic evolutionists report the same. Likewise, a similar percentage of creationists (47 report) anticipate increased disagreements with church leaders (Figure 23). Theistic evolutionists and atheistic evolutionists are both significantly less likely to report anticipating increased disagreements (22 percent and 15 percent respectively).

F IGURE 2 2 . Percent reporting that changing beliefs would cause increased disagreements with other church members (churchgoing population only) reporting at least occasional talk of human origins at church, and 54 percent of theistic evolutionists reporting the same. Lastly, we can examine the same social pressure measure that we used with friends and family. For those respondents who attended a religious congregation at least monthly, they were also asked whether they believed that changing their beliefs would cause increased or decreased disagreements, or no change at all. A similar pattern to the friends and family questions appears. Creationists, according to Figure 22, report substantially higher perceived chances of disagreements

The results are consistent. Creationists are more likely to be in congregations with a settled anti-evolution position, more likely to hear about this from the pulpit (although not more likely to informally talk about this with congregants), and more likely to experience social pressure from other members and religious leaders to keep their current beliefs.

The Influence of Schooling Schools have been the primary front in the battle over evolution in the United States24. Despite this, we have almost no clear documentation of their influence on beliefs about human origins. Are creationists disproportionately likely to have schooling experiences that cast doubt upon evolution? Is there evidence that science education in schools has an impact on what people later believe? These are empirical questions that 24. See Berkman and Plutzer. Evolution, Creationism, and the Battle to Control America's Classrooms

F IGURE 2 4 . Percent who learned that evolution was the only correct account of human origins in high school

F IGURE 2 6 . Percent reporting that they did not change their beliefs about evolution after learning about it in high school from school to school26. According to Figure 24, only about one in five creationists report learning that evolution was the correct account of human origins in high school. About one-third of theistic evolutionists, and a little less than one-half of atheistic evolutionists reported the same. One might suppose that this is due to differences in school sector, with creationists attending religious schools and evolutionists attending public and private non-religious schools. But even if we restrict the sample to those who attended public high schools, the differences remain (see Figure 25).

F IGURE 2 5 . Percent who learned that evolution was the only correct account of human origins in high school (public school only) have not been adequately measured in any existing national surveys. Although the national and state standards encourage high school biology teachers to make evolution a central component in their courses25, studies have noted how actual content differs substantially from class to class and

25. See chapter 4 of Berkman and Plutzer, Evolution, Creationism, and the Battle to Control America's Classrooms.

26

Perhaps creationists and evolutionists believe what they believe today because of what they learned in high school (Figure 26). Respondents were directly asked whether they changed their beliefs because of what they learned, and it is clear that the majority did not (this only includes those who report learning about evolution27).

26. Berkman, Michael B., Julianna Sandell Pacheco, and Eric Plutzer. "Evolution and Creationism in America's Classrooms: A National Portrait." PLOS Biology 6, no. 5 (2008): 920-924. 27. Respondents were asked if they learned about evolution or not. If they did learn about evolution they were asked whether they learned it was definitely true, possibly true, or definitely false. The respondents included in this graph are the ones who learned about

National Study of Religion & Human Origins

F IGURE 2 7 . Percent who learned that evolution was the only correct account of human origins in college (population who attended college only)

F IGURE 2 9 . Percent reporting that they did not change their beliefs about evolution after learning about it in high school

F IGURE 2 8 . Percent who learned that evolution was the only correct account of human origins in college (population who attended public university only)

F IGURE 3 0 . Percent who believe only evolution should be taught in public schools

Theistic evolutionists were the most likely to have changed their beliefs, but both creationists and atheistic evolutionists were quite unlikely to report that high school science education had any impact on them (85 percent of creationists who learned about evolution, and 81 percent of atheistic evolutionists who learned about evolution). evolution at all, including those who learned that is was definitely false or only possibly true.

Respondents are even less likely to report that they learned that evolution is the only correct account of human origins in college (Figure 27). This figure restricts the population to those who attended college. Twentyfour percent of atheistic evolutionists report learning this, but only 14 percent and 11 percent of theistic evolutionists and creationists report the same. Again, if we restrict this only to those who attended public universities, the results are remarkably similar (Figure 28), but shifted upwards slightly. This low percentage is almost certainly due to the fact that many college students are not required to take a biology course. Fully

percent), while atheistic evolutionists are somewhere in between (72 percent).

F IGURE 3 1 . Percent who believe only creationism should be taught in public schools

F IGURE 3 2 . Percent who believe some combination of evolution, creationism, and Intelligent Design should be taught in public schools 66 percent of those who attended college report never learning about human evolution (not shown). The comparable figure is 33 percent during high school. Among the one-third who went to college and learned about evolution, were their beliefs swayed? Again, Figure 29 shows the majority did not alter their beliefs. Creationists were the least likely to alter their beliefs (91 percent – not surprising given they currently report reject evolution). The most malleable are those who find themselves in the theistic evolutionist category (62

28

Lastly, regardless of what they learned in high school or college, what do they believe should be taught about human origins in public schools? Unsurprisingly, these groups stay true to their own views. A majority of atheistic evolutionists (77 percent) believe only evolution should be taught (Figure 30), while creationists are almost uniformly against teaching only evolution in public schools (3 percent). Theistic evolutionists fall between the two with 28 percent. Still, a majority of creationists do not want creationism to be exclusively taught in public schools (Figure 31). Thirty-seven percent affirm this, while very few theistic evolutionists (12 percent) and almost no atheistic evolutionists (2 percent) want the same. Less than 10 percent of any group wants intelligent design to exclusively be taught (not shown). The most popular view, for both creationists and theistic evolutionists is to teach some combination of evolution, creationism, and intelligent design (Figure 32). Over half of both groups think this is the best approach. Among atheistic evolutionists only 21 percent think this approach should be taken. All together, many creationists spent their schooling in science classrooms that did not endorse evolution, while most atheistic evolutionists did. This is still true when restricted to public high schools and universities. Nevertheless, a majority of all students report that they did not alter their views about evolution, even though they were taught about it in school. This corroborates other research that finds most students are not swayed by learning about evolution in the science classroom28. 28. Ingram, Ella L., and Craig E. Nelson. "Relationship between Achievement and Students' Acceptance of Evolution or Creation in an Upper-Level Evolution Course." Journal of Research in Science Teaching 43, no. 1 (2006), 7–24.; Rice, Justin W., Joanne K. Olson, and James T. Colbert. "University Evolution Education: The Effect of Evolution Instruction on Biology Majors' Content Knowledge, Attitude toward Evolution, and Theistic Position." Evolution: Education and Outreach 4, no. 1 (2011),137–44; Lawson, Anton E., and William A. Worsnop. "Learning about Evolution and Rejecting a Belief in Special Creation: Effects of Reflective Reasoning

National Study of Religion & Human Origins

Lastly, creationists and theistic evolutionists majorities that think public schools should controversy" by presenting multiple views origins. Atheistic evolutionists, in contrast, program of teaching evolution only.

have slight "teach the of human endorse a

Relationship to Other Beliefs about Science and Scientists Beyond the immediate social context, the various positions on human origins can be set within a larger intellectual framework. What sorts of understandings of science and scientists are associated with different beliefs about human origins? A number of additional survey items help to map out some general patterns. To begin, it is worth trying to assess whether creationists are somehow more generally anti-science because they stand against the scientific mainstream, or whether they believe in the importance of science but think that evolution is somehow inadequate on a scientific level. Although there is no question that asks about the value or science more generally, or what constitutes good science, there are questions about whether evolution is fully scientific. One such question is presented in Figure 33. These bar graphs show who believes that the theory of evolution is speculation and does not have strong factual evidence to back it up. A little more than half of creationists believe this, while much smaller percentages of theistic evolutionists and atheistic evolutionists agree (18 percent and 9 percent). While there is a very clear pattern to this data, this also suggests that nearly half of creationists do not agree that evolution is factually weak, yet still report that they do not believe in it when it comes to humans. Does this mean that creationists believe that there is real dispute in the scientific community over the issue of Skill, Prior Knowledge, Prior Belief, and Religious Commitment." Journal of Research in Science Teaching 29, no.2 (1992),143–66; Bishop, Beth A., and Charles W. Anderson. "Student Conceptions of Natural Selection and Its Role in Evolution." Journal of Research in Science Teaching 27, no. 5 (1990),415–27; Chinsamy, Anusuya, and Eva Plaganyi. "Accepting Evolution." Evolution 62, no.1 (2007), 248–54.

F IGURE 3 3 . Percent who believe the theory of evolution is speculative and not factual

F IGURE 3 4 . Percent who believe scientists disagree whether evolution is the best explanation for human origins evolution? Figure 34 suggests that some do (39 percent), but a full 61 percent do not believe that there is disagreement among scientists. Again, the pattern is clear with theistic evolutionists less likely than creationists to believe this (24 percent) and atheistic evolutionists the least likely to believe this (13 percent). Still, the fact that nearly half of creationists believe that evolution is scientifically sound, and about six in ten believe that scientists agree about evolution, may come as a surprise to many.

F IGURE 3 5 . Percent who believe scientists too often let their own personal beliefs influence how they conduct science

F IGURE 3 6 . Percent who believe scientists should be free to do any research without any interference from religious morals or teachings Perhaps they recognize that scientists largely agree about these issues but have a general distrust for scientists, believing them to biased in their assumptions and personal beliefs. There is some evidence for this. The survey includes a question asking whether the respondents agree that scientists too often let their own personal beliefs influence how they conduct science. About half of creationists agree (Figure 35), a little more than a quarter of theistic evolutionists agree (28 percent) and 17 percent of atheistic evolutionists agree. Related to this, creationists are much more likely to believe that the work of scientists should be bounded by other

30

F IGURE 3 7 . Percent who believe most scientists are hostile toward religion considerations. Figure 36, shows that only 16 percent of creationists believe scientists should be free to study whatever they want, while 72 percent of atheistic evolutionists believe this (theistic evolutionists fall between the two with 35 percent). Lastly, creationists are more likely than other groups to believe that most scientists are hostile to religion (Figure 37). Forty percent of creationists believe this while only 16 percent of atheistic evolutionists believe the same (again, theistic evolutionists are between the two at 24 percent). It’s important to point out that, even though creationists are more likely to believe that scientists are hostile to religion, those who affirm this belief are still in the minority among creationists. Several of these beliefs are highly correlated with one another (with correlation coefficients between .4 and .55). People tend to answer them in similar ways. Among creationists, 23 percent agree that evolution is not scientific fact, scientists do not agree on the scientific validity of evolution, and scientists are biased. At the same time, 28 percent of creationists do not affirm any of these three statements. Clearly, there is a split over why creationists are rejecting evolution. It may be useful now to zoom out a bit and assess how the public thinks about the relationship between science and faith more generally. There are two ways we can analyze this, one in terms of the existential feeling of conflict or compatibility, and the other in terms of intellectual beliefs about the ultimate conflict or

National Study of Religion & Human Origins

compatibility of the two realms. The first, felt conflict, is measured by agreement with the following statement: "Sometimes I feel like the findings of science and the teaching of [my] religion conflict with each other." Figure 38 shows that about half of creationists and half of theistic evolutionists report that they feel this way sometimes. Nearly three-fourths of atheistic evolutionists report they feel this way. The item immediately following this in the survey asks whether the respondent believes that religion and science are, ultimately, compatible (Figure 39). Here, theistic evolutionists and creationists diverge. Only onethird of creationists report that they believe they are ultimately compatible, while more than half of theistic creationists believe this is so. So, even though creationists and theistic evolutionists both feel conflict between the two in equal proportions, their ultimate intellectual beliefs diverge. Atheistic evolutionists are the least likely (21 percent) to believe they are compatible. From this, it seems like a majority of creationists and atheistic evolutionists believe in the incompatibility of these realms, they are just choosing opposite sides in the conflict. One additional question helps us to make sense of beliefs about the relationship between science and religion. The scientist Stephen Jay Gould coined the phrase "non-overlapping magisteria" (NOMA) to describe his position that science has a domain of teaching and religion has a domain of teaching, and that these domains do not overlap.29 One is strictly about empirical facts and mechanical explanation (science) and the other is about moral value and ultimate meaning (religion). How popular is this view among the general public? Figure 40 shows the agreement with the following statement: "Science is about facts and religion is about faith. The two do not overlap." In this case, it is the atheistic evolutionists that stand out from the rest. A majority tends to agree with this statement (62 percent), while only 32 percent and 26 percent of theistic evolutionists and creationists agree.

29. Gould, Stephen Jay. "Nonoverlapping Magisteria." Natural History (March 1997).

F IGURE 3 8 . Percent who sometimes feel that science and religion conflict

F IGURE 3 9 . Percent who believe that science and religion are ultimately compatible

F IGURE 4 0 . Percent who believe that science is about facts and religion is about faith and the two do not overlap

It is not as clear how atheistic evolutionists square this belief with the other question that asks whether the two are ultimately compatible. NOMA was conceived as an attempt to make science and religion compatible by saying they are distinct, bounded fields that tackle fundamentally different sorts of questions. Yet, atheistic evolutionists who claim to agree with NOMA are actually less likely to believe in this compatibility then atheistic evolutionists who don’t agree with NOMA (analysis not shown). Recalling the verbatim responses to why having the right beliefs about human origins is important (covered in the first part of this report), the NOMA response is appears popular among several atheistic evolutionists as a way of showing the superiority of science over faith. Science has objective empirical facts, while faith is merely subjective (and therefore unreliable) beliefs. How are we to make sense of all of this? Let's begin with the creationists. Creationists are clearly not opposed to science, but many do question whether evolutionary science is actually real science. Although we have to read between the lines a bit with the survey data, perhaps Ken Ham's view that observational science is the only real empirical science, and what he calls historical science is educated speculation about the past might be being invoked here.30 Creationists are much more likely to believe that evolution is speculative and not factual. As other research has demonstrated, this does not mean creationists are any less likely to value other sorts of scientific discoveries. Creationists are also slightly more suspicious of scientists and believe their work should be bounded by moral and ethical considerations. Creationists clearly believe that values need to be taken into consideration (they overwhelmingly deny that faith and values constitute a separate realm that has no bearing on science), and, at the same time believe that the values that some scientists already have are not to be trusted. Again, science itself is not at fault here, but how scientists actually go about doing science is sometimes a problem. 30. See https://answersingenesis.org/what-isscience/deceitful-terms-historical-and-observationalscience/. Or, see https://answersingenesis.org/what-isscience/science-or-the-bible/

32

Atheistic evolutionists take the other tack. Science and faith are separate realms, scientists are to be trusted and should be free to study whatever they want, and evolution is as empirically factual and unproblematic as any other form of scientific knowledge. Scientists are not hostile toward faith; rather, they are objectively studying the world and reporting what they find. It is the people of faith who have distorted views of things by refusing to allow their subjective beliefs and values be separated from scientific fact. In one way, though, creationists and atheistic evolutionists are similar. Only a minority of both agrees that science and religion are ultimately compatible. Given the survey topics, clearly both have evolution in mind when answering this question. When they feel that science and religion conflict, one group claims that religious ways of knowing (primarily a literal reading of scripture) are superior and one group claims that scientific ways of knowing (trusting in what they are told my scientists) are superior. On this front, theistic evolutionists are different. Even though many do not affirm the ultimate compatibility of science and religion, they are substantially more likely to adopt this view than either of the other groups. Like the atheistic evolutionists, this group is not very suspicious of the scientific validity of evolution, and they are clearly less suspicious of scientists and their values than creationists. Yet, like creationists, they believe that faith and values do intersect with science. The two are not entirely separate realms. It's also worth noting that creationists and theistic evolutionists are about equally likely to report feeling conflict between faith and science. They are just less likely to believe that this conflict is ultimately real.

Who is Most Likely to be a Confident Creationist? Up until now, we have been examining the various demographic, social, and belief profiles of those who were labeled creationist, theistic evolutionist, and atheistic evolutionist. But this only gets us so far if we are trying to understand how social context shapes a position on this issue. To go forward, we need to identify the factors that work together to produce those who

National Study of Religion & Human Origins

used to estimate the influence of each factor, are available in Belonging to an evangelical denomination (compared to Catholic and those with no affiliation) Frequently attending religious services Appendix A. The first step is to Reporting that faith is important in day-to-day life assess which factors have the Praying frequently largest influence using simple Belief that the Bible is literal (no symbolism) or inspired Word of God (symbolism but no errors) bivariate models. This method Family members have same beliefs about origins Friends have same beliefs about origins tells us what the total influence Changing belief would cause disagreement with other congregants and religious leader(s) of a variable is on being a Congregation has settled position that rejects evolution confident creationist, even Identifying as a political conservative though some of this influence TAB LE 4. Factors that have the largest total influence on predicting who is a could be indirect or spurious. certain creationist (bivariate analysis from Table A1 Model 1 in Appendix A) Although there is no single confidently espouse one view or the other on human method of identifying what is a powerful predictor, the origins. How are social identities and contexts associated standardized coefficients can give us a rough sense of with arriving at a firm position on this topic? Conversely, what is important. Again, for those interested in the what factors are associated with those who are open to methodological details, please consult Appendix A. change or who have recently altered their beliefs? Important Factors for Predicting Certain Creationist

To do this, we need to employ some new methodological techniques. We need to try to predict landing in one category versus everything else. And, we can do this by using multiple measures simultaneously to identity what factors strongly and directly predict this, and what factors fall by the wayside. Moreover, we will look at how individual characteristics – religious ones in particular – are heightened or lessened depending on social context. Unlike previous sections, this section uses those who are more confident in their beliefs. The goal is not to predict those who are on the fence or muddled in their views, but those who firmly hold to one position or another. Recalling the broad definition from Part I, creationists are those that do not believe in human evolution, and who affirm that God was somehow involved in the creation of humans. If we use this definition, and add in the additional caveat that they have to absolutely or very certain of both of these beliefs, we arrive at 29 percent of the entire U.S. adult population. What factors make it more or less likely to land in this category? To begin, each individual factor that we have examined so far (demographic, political, educational, religious, and measures of social context) is assessed to see which ones have the most predictive power. The tables showing this, and the explanation of the methods

Table 4 shows which factors initially appear to be most important. The religious factors all matter (religious tradition, attendance, importance of faith, frequency of prayer, and beliefs about the Bible). So do family, friends, and congregation. And, at least when examining basic correlations, political conservatism seems to matter. But we need to see whether these hold up in a more rigorous way. A number of these factors clearly correlate with one another, and this needs to be accounted for in order to best understand what is happening. For example, perhaps one or two of the religious factors actually have a direct influence, and the other ones appear important because they tend to be associated with these. To measure this, all the variables were simultaneously entered into a model that has the variables compete with one another to see which ones most directly account for who is a creationist and who is not. This is referred to as multivariate analysis (logistic regression in this particular instance). A more detailed explanation of the methods is available in Appendix A. When we do this, we find that many of the same measures are still important predictors, but a few notable ones are absent (Table 5). Church attendance, for example, is no longer a strong predictor when other religious factors are considered.31 Likewise, political 31. This is almost certainly due to the inclusion of other measures of the congregation. Church attendance itself

conservatism is no longer Important Factors for Predicting Certain Creationist important. Two new factors Belonging to an evangelical denomination (compared to mainline Protestant, Catholic, & Jewish) Reporting that faith is important in day-to-day life emerge that were unimportant in Praying frequently the first round: whether someone Belief that the Bible is literal (no symbolism) or inspired Word of God (symbolism but no errors) is married, and whether they are Family members have same beliefs about origins white or Hispanic (compared to Friends have same beliefs about origins Changing belief would cause disagreement with other congregants and religious leader(s) black). It should be noted that not Congregation has settled position that rejects evolution all of these are equally important Being married just because they are included in Being white or Hispanic (compared to black) this table. Many of the religious TAB LE 5. Factors that have the largest direct influence on predicting who is a measures, such as being certain creationist (multivariate analysis from Table A1 Model 2 in Appendix A) evangelical (compared to certain creationism. Being married, being white or Catholic), saying faith is important, believing the Bible is Hispanic, and having friends and family believing the the literal or inspired Word of God, and having family same things about human origins do not add enough to members believe the same thing about human origins, these models to warrant including them. stand out as the most important factors. We are not quite through though. This type of model relies on every predictor competing with other predictors to best account for the outcomes. It is a good way to sort through what is directly associated with being a creationist and what is only indirectly or spuriously related. However, in the real world, characteristics, beliefs, and behaviors never compete like this. Individuals embody multiple factors simultaneously, and it is quite likely that these factors work together in complex ways that lead to beliefs about human origins. A regression model attempts to simplify reality by assuming that each part has a unique, additive, influence on the outcome. But what if we start assuming that there are complex, combinatorial conjunctions of factors and attempt to reduce these to the most parsimonious account? An alternative set of methods, known as fuzzyset/QCA attempts to do precisely this.32 We can begin with the list of factors we have already identified from Table 5. Additional analyses suggest a few of these variables (the three weakest ones) are not necessary in developing the "recipes" that lead to is not important once the aspects of the church are controlled for (whether they have a settled position on evolution and whether the respondent sense that there would be negative social repercussions for changing beliefs about human origins). 32. Ragin, Charles. Fuzzy-set Social Science. (Chicago: University of Chicago Press, 2000).

34

Through the use of Boolean algebra, we can identify the most dominant pathways to becoming a certain creationist using the remaining seven measures. Again, for methodological detail, please consult Appendix A. We can begin by examining every possible combination of factors to identify which combinations contain greater than half of the cases as certain creationists. Then, using set logic, we can reduce these combinations to their most parsimonious sets of factors. When we do this, we find five dominant pathways (Figure 41). These pathways account for 35 percent of all certain creationists. That means that a number of other less common pathways lead to becoming a certain creationist as well, but these are all less frequent, and less reliable than the pathways listed here. However, if someone belongs to one of these pathways, these models predict that there is 91 to 96 percent likelihood he or she will be a certain creationist. In other words, it's unusual to be in one or more of these pathways and not end up identifying as a certain creationist. These pathways overlap a great deal, and a few of their commonalities help us to see the absolutely essential elements. First, all five pathways include beliefs that the Bible should be taken either completely literally, or that it should be taken to be inspired and error free. But, we should recognize that, on their own, these Bible beliefs are not sufficient to produce a confident creationist (50% of those with such Bible beliefs end up being classified as

National Study of Religion & Human Origins

changing one's beliefs would result in disagreements with religious leaders and other congregants. Both of these are part of four out of the five pathways. The other measures of importance of faith, prayer, and whether family believes the same things about origins are a part of some pathways and not a part of others. Altogether, this data shows that there are some very reliable "recipes" for producing F IGURE 4 1 . fs/QCA analysis showing five pathways to being a certain creationist certain creationists. Given some theological a confident creationist), this needs to be combined with beliefs, personal religious commitment, and additional factors. congregational characteristics, you will very likely end up with someone who rejects evolution and is certain of The other factor that is essential to all five pathways is their position. belonging to a religious congregation that has a settled position of rejecting human evolution. In fact, just the Who is Most Likely to be a Confident Theistic combination of the Bible beliefs above, and belonging to Evolutionist? such a congregation, will produce a certain creationist 81% of the time. Not only this, but nearly half of certain Now that we have the basic tools to identify and analyze creationists identify with both of these components. the social correlates of creationism, we can apply these Maintaining these Bible beliefs but belonging to a methods to see what factors predict the other positions. congregation without a settled position on the topic only Confident theistic evolutionists are a fairly uncommon produces a certain creationist 38 percent of the time. group overall. Unlike confidence creationists who make This demonstrates how individual beliefs are important, up 29 percent of the population, confident theistic but only in certain contexts.33 evolutionists make up a mere eight percent. One of the Other important factors include belonging to an evangelical Protestant congregation and sensing that

33. It is important to recognize that there could be some reporting error about congregations. Individuals who are not certain creationists, or who have little interest or investment in the topic of human origins, may be less aware of their own congregation's position. Still, it is unlikely that these large differences are due to reporting error alone.

issues is that a lack of certainty is a hallmark of this position. Approximately half express some uncertainty, while less than a quarter of creationists do the same. What social characteristics are associated with belonging to this group? As with creationists, we can analyze the effect of each variable on its own as well as the unique influence of each variable when all other variables are controlled for. For the sake of simplicity of presentation, the entire statistical models are not shown here (they are available in Appendix A).

The first pathway involves identifying as a mainline Belonging to a mainline Protestant denomination (compared to evangelical Protestant) Being Catholic (compared to evangelical Protestant) Protestant, not believing that Not believing that the Bible is the literal or inspired Word of God without errors the Bible is the literal or Not identifying as a political conservative inspired Word of God without TAB LE 6. Factors that have the largest direct influence on predicting who is a errors (many still believe that certain theistic evolutionist (multivariate analysis from Table A2 Model 2 in the Bible is inspired, just not Appendix A) error free on all matters), and not identifying as a political conservative (they are either What is immediately apparent is that no variables reach moderates, liberals, or have no political identification). the level that was marked as a strong influence This group accounts for 11 percent of all confident (standardized coefficient above .3). Put simply, these theistic evolutionists and this pathway leads to variables are not nearly as powerful in predicting who identifying as a confident theistic evolutionist 20 percent ends up as a theistic evolutionist compared to a of the time. The second pathway is like the first, except creationist. the religious tradition the respondents belong to is Catholic. This accounts for an additional 13 percent of However, when we include the variables together (the the group (for a total of 24 percent of all confident multivariate "kitchen sink" logistic regression), a small theistic evolutionists), and leads to identifying as a handful emerges as important and uniquely contribute to confident theistic evolutionist about 15 percent of the 34 the model (Table 6) . The three that matter most are time. religious tradition (mainline Protestants and Catholics are both more likely than evangelicals), not affirming that the Bible is the literal or inspired Word of God without errors (although not necessarily rejecting the authority of the Bible altogether), and not identifying as a political conservative. Important Factors for Predicting Certain Theistic Evolutionist

As before, we can examine whether these matter in combination. Are there particularly likely pathways, using these three factors, to becoming a certain theistic evolutionist? Using the same fuzzy-set/QCA technique, we find that there are only two pathways (Figure 42). Not only this, but because being confident theistic evolutionists only make up eight percent of the population, these are not reliable "recipes" for producing confident theistic evolutionists like they were for creationists. In fact, a sizable majority of people in these pathways will not end up being confident theistic evolutionist. Still, these are the most likely pathways, given the measures in the survey, for ending up in this group.

34. The BIC is reduced when they are added to the model compared to a model without them. See Appendix A for more details on the methodology used.

36

F IGURE 4 2 . fs/QCA analysis showing two pathways to being a certain theistic evolutionist

These suggest there are many other pathways that are not easily captured by the survey measures at our disposal. Unlike confident creationists, who seem to follow clear pathways, confident theistic evolutionist have idiosyncratic reasons for believing as they do. Still, we can say with some confidence that Catholics and mainline Protestants who hold to certain beliefs about the Bible and are political moderates and liberals are around two and half times more likely to be confident theistic evolutionists than the rest of the population.

National Study of Religion & Human Origins

Who is Most Likely to be a Confident Atheistic Evolutionist

us that a large part of this is simply church attendance. So these congregational factors are not as important as they first appear. Lastly, both not identifying as a political conservative, and being white (compared to being black) are associated with being an atheistic evolutionist, although neither one of these are as strong as the religious factors.

When we apply our methods to confident atheistic evolutionists, we find, like the creationists (but unlike the theistic evolutionists), that several of our social and demographic measures are important predictors. In many ways, atheistic evolutionists are the flip side of creationists (although there is an Important Factors for Predicting Certain Atheistic Evolutionist important difference that I will Identifying as Catholic or nonreligious (compared to evangelical Protestants) highlight in a moment). The Infrequently attending religious services variables that turn out to be Reporting that faith is not important in day-to-day life Infrequently praying important in the bivariate analysis Not believing that the Bible is the literal or inspired Word of God without errors (without controls) are shown in Changing belief would not cause disagreement with other congregants and religious leader(s)a Table 7. Again, for Congregation does not reject evolution or does not have a position on evolutiona methodological details please Not identifying as a political conservative Being white (compared to being black) consult Appendix A. TAB LE 7. Factors that have the largest total influence on predicting who is a certain atheistic evolutionist (bivariate analysis from Table A3 Model 1 in It is clear, and perhaps not Appendix A) surprising, that atheistic Notes: aThese congregational variables likely have a strong correlation because they are contrasted evolutionists are overwhelmingly with those who do not have a congregation. Separating out the influence of church attendance from not religious (although, at least in the stance of the congregation and its leaders is only possible in the multivariate analysis (shown in the bivariate results, Catholics Model 2 of Table A3). See Table 8 where these factors are no longer important. also are more likely to identify as such when directly compared to evangelical Protestants). They Important Factors for Predicting Certain Atheistic Evolutionist identify as nonreligious, they do Identifying as nonreligious (compared to evangelical Protestant) not attend religious services, they Infrequently praying report faith is not important in Not believing that the Bible is literal or inspired Word of God without errors Not identifying as a political conservative their daily life, they pray Being white (compared to being “other” race/ethnicity) infrequently or not at all, and they TAB LE 8. Factors that have the largest direct influence on predicting who is a do not believe the Bible should certain atheistic evolutionist (multivariate analysis from Table A3 Model 2 in be taken literally and is free from Appendix A) errors. All of these are very sizable effects (above .6 for the When all of the variables are entered into a model standardized coefficients). In addition to this, those who simultaneously, five come out as being particularly do not feel pressure from religious leaders or important in directly predicting who is an atheistic congregants, and those whose congregation (if they have evolutionist (Table 8). Three of these are religious factors: one) do not reject evolution or do not have a settled being nonreligious, infrequently praying (or not praying position on evolution are more likely to be atheistic at all), and believing that the Bible is not the literal or evolutionists. It should be stated, though, that the way inspired Word of God without errors. Church attendance these congregational variables were constructed makes and importance of faith in daily life turn out to be less them correlate very highly with church attendance. This important in directly influencing whether someone is an makes it impossible to know if it is attending any church, atheistic evolutionist. Congregational factors are no or particular characteristics of those churches, that makes longer directly important in this model, once other someone more or less likely to be an atheistic factors (including church attendance) are controlled for. evolutionist. The multivariate results (discussed next), tell The remaining two factors are the same as the previous

model: not identifying as a political conservative, and being white (this time, though, contrasted with being "other” race/ethnicity).

F IGURE 4 3 . fs/QCA analysis showing one pathway to being a certain atheistic evolutionist

percent of the overall population). The combination is as follows: identifying as not religious, not holding to the authority of the Bible, infrequently praying or not praying at all, and identifying as white (Figure 43). Political identification does not emerge as an important distinguishing factor when considered in combination with the others. The first three factors make sense, but the last one may be surprising. Nevertheless, it is important. Non-whites who match the first three factors only end up being classified as certain atheistic evolutionists 17 percent of the time. The difference is substantial.

Although creationists tend to be highly religious, and Who is Most Likely to have Recently Changed atheistic evolutionists tend to not be religious, there is an their Beliefs about Human Origins? important difference between the two. Atheistic evolutionists feel no social pressure from their family and Instead of examining who is confident in their position on friends (and not really any from congregations if they human origins, we can turn this around and see what have one). This is consistent with other research that factors are associated with having recently changed finds social pressures to be an important factor for beliefs on human origins in some way. Although creationists only.35 Because evolution is the accepted respondents were asked what has changed about their position of mainstream science, it is probably unlikely to beliefs, to keep this analysis simple we will be examining require strong, direct social influences to maintain this the individual and social factors associated with changing position. Creationists have the additional burden of beliefs of any sort. For clarification, this group is less providing an alternative to the mainstream position. likely to be creationist (and definitely less likely to be Family, friends, and Important Factors for Predicting Recent Change in Belief congregations play an Frequently praying important part in making that Not believing that the Bible is literal or inspired Word of God without errors 36 alternative seem plausible. Family members do not have same beliefs about human origins Anticipate increased disagreements with family members if changed beliefs about human origins Attends a congregation that accepts human evolution Young Male

Using these five factors, we can examine if there are any distinct pathways to identifying as a TAB LE 9. Factors that have the largest direct influence on predicting who has confident atheistic evolutionist. recently changed beliefs about human origins (multivariate analysis from Table Using the same fuzzy-set/QCA A4 Model 2 in Appendix A) method, I find a single certain creationists), more likely to theistic evolutionists, important pathway. In fact, nearly two-thirds of all certain and less likely to be atheistic evolutionists. So, although atheistic evolutionists fit this pathway, and having these this group does not represent one particular transition in characteristics leads to a certain atheistic evolutionist belief, there are some general patterns that point to identity 38 percent of the time (despite only being six movement toward the theistic evolutionist position. 35. Hill, Jonathan P. "Rejecting Evolution: The Role of Religion, Education, and Social Networks." Journal for the Scientific Study of Religion 53, no. 3 (2014), 575-94. 36. See p. 45 of Berger, Peter. The Sacred Canopy: Elements of a Sociological Theory of Religion (New York: Anchor Books, 1967).

38

Again, the same methods are applied. First we will look at bivariate relationships, then we will examine what factors are most important in directly predicting who recently changed beliefs when all the factors are entered in simultaneously. Last, we will see what combination of

National Study of Religion & Human Origins

factors tend to produce high likelihoods of recently changing beliefs.

associated with being younger. It is also slightly more associated with being a man than a woman.

Overall, these variables are not strong predictors. In fact, none of the standardized coefficients reach .3 in magnitude, the level that we have been using to enter them in to a separate table. Certainly some factors are statistically significant, but the predictive power does not equal the magnitude we have seen in the other models. The effect sizes are most similar to the table for the theistic evolution position.

As before, we can combine these factors together to see if there are particular pathways to changing beliefs about human origins. The results, which are not shown, find numerous very small pathways (twelve). Not only this, but these pathways only represent a fraction of those who end up reporting that they recently changed their beliefs. This suggests that, like the theistic evolutionists, the pathways tend to be idiosyncratic and inconsistent.

Using the rules that we have established for the multivariate analysis (the "kitchen sink" model available in Appendix A), several factors do uniquely, and directly help predict who recently changed their beliefs. Table 9 lists these: Those who pray frequently, who do not believe the Bible is the literal or inspired Word of God without errors, whose family members do not have the same beliefs about human origins, who do anticipate increased disagreement with family members over changing their beliefs, who belong to congregations that accept human evolution, who are younger, and who are male.

In conclusion, it is much easier to map the social influences of creationists and atheistic evolutionists. Trying to uncover the patterns that lead to theistic evolutionist positions or to recently changing beliefs about human origins is much harder to do with the types of measures available in surveys. Nevertheless, there are individual factors that can be identified as important, even if the overall rates of predicting who ends up in these latter categories is not as precise as the former categories.

This is a complex list. And, given the nature of the question (have you recently changed your beliefs?) it is difficult to know whether some of the current measures of social context are a response to a change in belief, or helped precipitate this change. Nevertheless, we can say a few things that help to make sense of the issue. Confident creationists reported that their family agrees with them, and that they would suffer some negative consequences from their family if they changed positions. In this case, those whose families believe differently from the respondent (different from creationists) and who anticipate negative repercussions from their family for changing beliefs (same as creationists) are more likely to have recently changed beliefs. This suggests that there is some sort of ongoing conflict with family among those who recently changed their belief. At the same time, they are more likely to belong to congregations that accept evolution on average. Again, we simply do not know whether they switched congregations after having a change in belief or not. Lastly, like many belief changes, this tends to be

Conclusion The NSRHO was designed to provide a more accurate picture of the U.S. public's beliefs about human origins than existing surveys. What we find is that beliefs combine in complex ways, and those that neatly fit into the positions of creationist (including both Old Earth and Young Earth), theistic evolutionist, and atheistic evolutionist are smaller proportions of the population than reported elsewhere. Some hold beliefs that are not entirely consistent with any of the views, and many are unsure of their belief or report that having particular beliefs about human origins is not important to them. Just as significant, the findings of this study tell us that the social context of these beliefs are as important as individual factors such as religious identity, practice, and belief. Social networks of family and friends, congregations, and schools all play a role in this. Most important, especially for creationists, is the combination of certain religious beliefs and practices (especially beliefs about the Bible) and certain social contexts (especially ones related to a religious congregation). In fact, the data here suggest that the only surefire way to accurately predict whether someone is a creationist who is confident in their beliefs is to take both individual and social context into account. For those invested in the position that human evolution is compatible with orthodox Christian faith, the findings from the NSRHO tell us that persuasion needs to move beyond a purely intellectual level. Ideas are important, but ideas only persuade when individuals are in a social position that allows them to seriously consider what is before them. It's not only that such people feel constrained by family, friends, and congregation (although this might be the case for many), it's that what counts as a plausible account of human origins is reinforced by important social relationships. Many freely and gladly believe that human evolution is a patently false and dangerous position. Their own networks of coreligionists give plausibility to this position. This social context is probably especially important for creationists because their beliefs run counter to the mainstream scientific establishment. They have the burden of generating plausibility structures in the face of scientific consensus. All of this tells us that a two-pronged strategy – one that attends simultaneously to sociological and theological factors – will make the most inroads in the end.

Appendix A: Statistical Methodology in this Report This report relies on various statistical techniques and results whose details are often absent from the main body for the sake of clarity and brevity. The following four sections provide the details.

Bar Graphs in the Report The bar graphs in the report are graphical presentations of average marginal effects from logistic regression models. When the model only includes one predictor, this average marginal effect is identical to running a simple frequency table that calculates the weighted proportion belonging to the outcome in question for each subpopulation (i.e., creationist, theistic evolutionist, atheistic evolutionist). The advantage to using the marginal effects, even in the simple bivariate case, is that logistic regression estimates provide the standard error that can be used to calculate a 95% confidence interval error bars. The adjusted percentages require the logistic regression method and cannot easily be created using simply frequency tables. The average marginal effect is a method that allows all explained variation from the control variables to be removed by estimating the outcome for each case based on the model and then using a weighted average within each category of the main independent variable. A similar method would be to hold all of the control variables at their mean and estimate the influence of the main independent variable on the outcome. This is known as the marginal effect at the mean and is slightly different from what the average marginal effect calculates, but the end result is usually quite similar. The marginal effect at the mean sometimes gives problematic values when the independent variables are highly correlated with one another, for that reason I prefer to report the average marginal effect. A "do file" in the software program Stata was created that converted the average marginal effects and the 95 percent confidence intervals into the bar graphs that were used in the report.

Coding Open-Ended Responses For respondents who reported that is was very or extremely important to them to have the right beliefs about human origins, they were given the opportunity to report why this was important to them (see question HO9b in Appendix B). I initially coded all 947 valid responses using an inductive method that looks for simple commonalities between responses. Two student researchers independently analyzed the results using the initial categories I had created. Finally, I synthesized the coding from all three of us into a single spreadsheet. In the end, thirty unique themes were identified. These were merged back into the survey data software (Stata) to be sorted and counted. Although not included in this report, the themes can now be used as variables for additional modeling (either as outcomes to be predicted by other variables, or as independent variables used to predict other outcomes).

Logistic Regressions Predicting Stance on Human Origins In the latter sections of the second part of the report, logistic regression results were used to determine the most important factors for predicting who adopted views on human origins (and held them with certainty) and who had recently changed their views. Each regression had two parts. The first was a series of bivariate regressions (using only one independent variable in each) to determine the influence of each factor on the outcome separately. The second was a "kitchen sink" model that includes all of factors simultaneously to see which ones have the most direct effect on the outcome net of the other predictors.

The four tables that follow show the results from these analyses. Each table has two models, one for the bivariate results (these actually summarize numerous logistic regressions) and one for the "kitchen sink" results (these summarize one logistic regression). For each predictor, three indicators are included to help determine whether it is an important or unimportant factor. The first is the unstandardized odds ratio coefficient. If the odds ratio is greater than one, than that factor has a positive influence, if the odds ratio is less than one, that factor has a negative influence. Odds ratios can be interpreted as percent change in odds, and so are typically a little easier to interpret than log odds (the original metric for the coefficients). Asterisks and plus signs indicate statistical significance. Directly below the odds ratios in brackets are the standardized coefficients. These are the full standardized coefficients (both X and the latent Y variable are standardized to have a mean of zero and standard deviation of one). Although the coefficients are not as intuitive to interpret as the odds ratio, they are now directly comparable to one another. So, larger coefficient size means a larger influence. In the tables presented in the main body of the text, predictors that had standardized coefficients larger than .3 were included as particularly important and listed in separate tables. Lastly, below the standardized coefficients in braces is the change in the Bayesian information criterion (BIC). BIC is used as a general indicator of model fit. Sometimes variables can have a large influence, but because they capture rare subpopulations or uncommon behaviors, they do not contribute substantially to predicting the outcome. The change in BIC allows us to examine whether the variable contributes substantially to predicting the outcome. A positive BIC change score indicates that the BIC drops when the variable is added as a predictor (comparing it to a K-1 model). This is used as the primary indicator of important direct effects in the "kitchen sink" model. Our predictions are substantially improved by including these variables, and including them is worth the overall loss in parsimony. Negative values indicate the model is better off without the predictor.

Fuzzy-set/QCA Techniques These models are substantially different from standard parametric statistical analysis. The fuzzy-set/QCA (QCA stands for Qualitative Comparative Analysis) techniques are based on Boolean algebra, which uses set notation (truth values along with conjunction, disjunction, and negation operations) to represent logical propositions. The original QCA "crisp-set" analysis (where "variables" could only take on values of 1 or 0) was developed as a technique by Charles Ragin37 to systematize methods for studies that compare a small number of cases. Since then, the technique has been expanded to include "fuzzy" sets (that take on values between 0 and 1) and is increasingly used in large-N studies38. Traditional multivariate methods begin with causal simplicity. The covariates are assumed to be additive unless otherwise specified. While it is certainly possibly to model causally complex interactions, these models often become unwieldy if multiple interactions are specified involving more than two variables. The default for traditional multivariate methods is to assume additivity unless there is theoretical justification otherwise. Fuzzy-set/QCA methods begin with causal complexity, treating all cases that belong to a given set combination as a unique group. Groups are then combined and simplified based on their membership in the outcome set of interest. Beginning with complexity and reducing to simplicity allows us to uncover combinatorial patterns of set memberships that also have membership in the set outcome of particular beliefs about origins (creationist, theistic evolutionist, atheistic evolutionist). These causally complex patterns would likely be missed with traditional multivariate methods. The use of set logic in the fuzzy-set/QCA method also makes explicit the relationship between the causal factors and the outcome in a way that is often obscured by traditional statistical methods. In this particular case, we are searching 37. Ragin, Charles. The Comparative Method: Moving beyond Qualitative and Quantitative Strategies (Berkeley: University of California Press, 1989). 38. Ragin, Charles. Redesigning Social Inquiry: Fuzzy Sets and Beyond (Chicago: University of Chicago Press, 2008).

42

National Study of Religion & Human Origins

for causal combinations that consistently lead to certain beliefs about human origins. The implication is that the set of causal forces are a sufficient, but not necessarily necessary, condition of the outcome. In the language of set theory, the cases that share a set of causal conditions are a subset of the cases that share membership in an outcome condition. Although a particular combination of sets (X) consistently and reliably leads to membership in the set outcome (Y), there may be other ways to reach Y – this is what is meant by sufficiency in fs/QCA. If we were interested in finding a necessary condition or set of conditions that leads to the outcome of high charitable giving, we would expect the set of cases that share the outcome to be a subset of the set of cases that share the causal condition. The combination of set conditions (X) is necessary to lead to the outcome set (Y), and no pathway to the outcome set exists outside of these conditions. Standard statistical practice obscures the difference between these two, yet our particular analysis is strictly interested in the first instance (sufficiency) and not the second (necessity). Fuzzy-set/QCA analysis decomposes the traditional correlational analysis into components that can be more precisely connected with the research question. All of the important variables from the "kitchen sink" models were dichotomized into 0 or 1 "crisp sets" or transformed into fuzzy sets, depending on what was being measured and the conceptualization of set membership appropriate to each variable. Using the program "fuzzy39" in Stata, I began by running all combinations of set memberships to assess their membership in the outcome of interest. After assessing these, I developed an initial consistency cut point that sufficiently captured reliable pathways to the outcome set. For example, combinations of sets that produced certain creationists above 60 percent of the time were allowed to "pass through" to the next level of analysis. I also only let these sets pass when they were significantly above the consistency cut point using a Wald test for statistical significant (this eliminates instances of rare combinations that may lead to consistent membership in the outcome set by chance alone). Finally, those sets that pass through are logically reduced to a final set solution using the Quine-McCluskey algorithm.

39. Longest, Kyle C. and Stephen Vaisey. "Fuzzy: A Program for Performing Qualitative Comparative Analysis (QCA) in Stata." The Stata Journal 8, no. 1 (2008), 79-104.

Table A1. Odds ratios from logistic regression predicting certain creationist beliefs versus all others, weighted (N=2,906) Bivariate Kitchen no controlsd Sinke Religious Tradition: Mainline Protestanta 0.218*** 0.388*** [-0.215] [-0.100] {103} {14} a Religious Tradition: Black Protestant 0.263*** 0.5 [-0.137] [-0.053] {45} {-3} a Religious Tradition: Catholic 0.176*** 0.261*** [-0.322] [-0.185] {201} {52} a Religious Tradition: Jewish 0.049*** 0.148*** [-0.177] [-0.083] {50} {2} Religious Tradition: Mormona 1.267 1.022 [0.012] [0.001] {-7} {-8} a Religious Tradition: Agnostic/Atheist/None 0.022*** 0.478+ [-0.668] [-0.096] {482} {-2} Frequency of religious service attendance 1.451*** 0.938 [0.477] [-0.057] {532} {-7} Importance of faith in daily life 3.290*** 1.636*** [0.673] [0.228] {877} {30} Frequency of prayer 1.870*** 1.256*** [0.617] [0.173] {715} {19} Bible is literal or inspired Word of God, no errors 10.195*** 2.397*** [0.539] [0.146] {606} {27} Family members have same beliefs about origins 2.793*** 1.683*** [0.464] [0.161] {448} {34} Friends have same beliefs about origins 2.601*** 1.354** [0.387] [0.081] {305} {3} Changing belief would cause disagreement w/ 3.719*** 1.248 family [0.288] [0.031] {189} {-7} Changing belief would cause disagreement w/ 3.128*** 1.021 friends [0.235] [0.003] {121} {-8} Changing belief would cause disagreement w/ 10.670*** 2.086** congregants and religious leader [0.401] [0.082]

{395} {4} 2.016*** 0.585 [0.097] [-0.052] {11} {-6} b Congregation rejects human evolution 17.294*** 3.458** [0.527] [0.162] {684} {6} b Congregation does not have position on evolution 2.776*** 1.152 [0.148] [0.014] {41} {-8} Political conservatism 1.552*** 1.090+ [0.337] [0.043] {214} {-4} Lives in south 1.420*** 0.769+ [0.093] [-0.042] {9} {-4} Educational attainment 0.949* 1.029 [-0.058] [0.020] {-1} {-7} Age 1.011*** 1.011* [0.102] [0.062] {12} {-0} Family income 0.972* 0.966+ [-0.071] [-0.053] {2} {-3} Employed 0.905 1.026 [-0.027] [0.004] {-7} {-8} Married 1.826*** 1.494* [0.164] [0.067] {45} {2} Female 1.132 0.751* [0.034] [-0.048] {-6} {-2} c Black 0.826 0.496* [-0.033] [-0.073] {-6} {0} Hispanicc 1.117 1.738* [0.021] [0.064] {-7} {2} c Other race/ethnicity 0.590* 0.61 [-0.077] [-0.044] {2} {-4} Notes: a Reference category is evangelical Protestant. Religious traditions categorized as “other” and “undefined” are not shown being they are not inherently meaningful categories; b Reference category is no congregation; c Reference category is white. d Bivariate models show the influence of the independent variable on the dependent with no other variables in the models. e “Kitchen Sink” means that all variables were entered in simultaneously. Standardized coefficients are shown in brackets. Drop in BIC for K-1 model shown in braces. + p < .10, * p