Acceptance of Evolution and Support for
Teaching Creationism in Public Schools: The
Conditional Impact of Educational Attainment
JOSEPH O. BAKER
Department of Sociology and Anthropology
East Tennessee State University

Public acceptance of evolution remains low in the United States relative to other Western countries. Although advocates for the scientific community often highlight the need for improved education to change public opinion, analyses of data from a national sample of American adults indicate that the effects of educational attainment on attitudes toward evolution and creationism are uneven and contingent upon religious identity. Consequently, higher education will only shift public attitudes toward evolution and away from support for teaching creationism in public schools for those who take non-“literalist” interpretive stances on the Bible, or to the extent that it leads to fewer people with literalist religious identities.

Keywords: evolution, creationism, religious identity, education, science and religion, public policy.

INTRODUCTION
Acceptance of evolution and support for creationism has been publicly debated since the initial diffusion of Darwin’s theory about the origin of species, particularly in the United States
(Numbers 1998, 2006). From before the infamous Scopes Trial (see Larson 1997) to the present, many Americans have resisted ideas about evolution, leading to a relatively low global ranking on public acceptance of the theory (Miller, Scott, and Okamoto 2006). Although the scientific community and legal decisions in U.S. courts have repeatedly favored the teaching of evolution and the dismissal of creationism in science classrooms over the last half-century (Larson 2003), public opinion on these topics has remained relatively stable (Plutzer and Berkman 2008). This raises the question: Why have institutional shifts in science and public education supportive of evolution failed to result in greater public acceptance?
CONTINGENT EFFECTS OF EDUCATION
Applying Gieryn’s topographical metaphor of boundary drawing in institutional science, disputes over evolution have served to demarcate the legitimate cultural domain and authority of science relative to other social institutions, especially education and religion (Gieryn 1983,
1999; Gieryn, Blevins and Zehr 1985). While advocates for institutional science have pushed creationism out of officially legitimated science (and therefore presumably public education) through legal decisions, creationism has maintained strong subcultural currency. Evans (2011)

Acknowledgments: Wave II of the Baylor Religion Survey was funded by a grant from the Templeton Foundation. The data are publicly available through the Association of Religion Data Archives. Thanks to Chris Bader, Paul Froese, Carson
Mencken, and Bill N. Duncan for helpful comments on previous drafts. An earlier version of this study was presented at the 2008 meeting of SSSR in Louisville, KY.
Correspondence should be addressed to Joseph O. Baker, Department of Sociology and Anthropology, East Tennessee
State University, P.O. Box 70644, Johnson City, TN 37614. E-mail: bakerjo@etsu.edu
Journal for the Scientific Study of Religion (2013) 52(1):216–228
C 2013 The Society for the Scientific Study of Religion

ACCEPTING EVOLUTION; SUPPORTING CREATIONISM

217

has demonstrated that conflicts involving religion and science are about morality and public influence rather than formal epistemology. Although generalized perceptions of incompatibility between science and religion are not prevalent among the American public (Baker 2012), specific topics such as evolution have come to serve as symbolic boundary markers for some religious groups. Because of this role as signifiers of institutional commitment, religious views, political identity, and educational attainment are the most salient correlates of accepting of evolution
(Evans 2011; Keeter, Smith, and Masci 2011; Mazur 2004) and supporting teaching creationism among Americans (Eckberg and Nesterenko 1985; Woodrum 1992).
A particularly salient collective and individual identity in the maintenance of issue-specific perceptions of conflict between science and religion is biblical “literalism” (Bruce 2008; Ellison and Musick 1995; Hoffmann and Bartkowski 2008; Lawrence 1989; Marsden 2006). Although often conceptualized as a singular belief, literalist perspectives represent social identities embedded in interpretative communities (Bartkowski 1996; Boone 1989; Stroope 2011). A cognitively consequential feature of literalist identities is the filtering of information through an “intratextual” interpretive lens. Hood, Hill, and Williamson (2005) argue that the defining characteristic of fundamentalist religious identity is its role as a perceptual filter; further, they suggest that for literalists:
Education and vocation are good only to the extent that they facilitate the divine intention and design for each person in relation to the community of believers. Should, for example, higher education or career advancement undermine these higher purposes, they become tools of the devil. (2005:32)

Similarly, Ammerman reported in her ethnography of a literalist community that:
Anyone who contradicts the “plain words of scripture” is doing the work of Satan, whether they know it or not.
As one man put it, “Satan attacks the word of God anyway . . . particularly Genesis and Revelation. . . . So he tries to throw out the beginning by bringing up evolution, and throw out the ending by saying that man does not owe anything to God. And then, there you come up with the atheist.” (1987:52)

The view that evolution is a slippery slope to disbelief has been, and remains, a foundational tenet of creationist arguments (e.g., Ham 2006). Due to the “functionality” of conflict for collective identity (Erikson 1966), disputes with secular culture actually bolster rather than undermine subcultural religious identities by providing symbolic outgroups against which these identities come into shaper relief (Smith 1998). Accordingly, experimental research has found that conservative religious identity inhibits the internalization of information about evolution (Lawson and
Worsnop 1992; Sinclair and Pindarvis 1998; Brossard et al. 2009).
In general, higher educational attainment results in substantially lower probability of maintaining a literalist stance toward sacred texts. Further, higher levels of educational attainment among those in one’s religious group also reduce the likelihood of affirming literalism, while amplifying the negative influence of individual education level (Stroope 2011). Yet there remain many highly educated biblical literalists. In the 2010 General Social Survey, 13 percent of biblical literalists had a bachelor’s degree or higher, meaning that approximately 4.4 percent of the U.S. population (roughly 13.5 million people) may be classified as biblical literalists with at least a bachelor’s degree.
Those who attain higher education while maintaining exclusivist religious views do so by imbedding themselves in social networks of traditionally religious people (Mayrl and Uecker
2011). Beyond social networks, selective information seeking and media consumption play critical roles. A primary trope used by organizations such as the Institute for Creation Research is that people should “decide for themselves what is true” regarding evolution and creationism.1

1 For

example, an M.D. with an M.P.H. from an Ivy League school representing ICR repeatedly told attendees of a presentation delivered at a public university that they should “decide for themselves” about evolution, rather than

218

JOURNAL FOR THE SCIENTIFIC STUDY OF RELIGION

As literalists raise their level of education, they are more likely to confront issues of evolution directly and reinforce their working knowledge of rhetorical defenses of creationism (see Toumey
1994). This selective information seeking reinforces a literalist identity and strong commitment to the moral authority of traditionalist religious communities. In essence, higher education pushes individuals toward a firmer stance on evolution and creationism (see notes 5 and 8).
In addition to seeking out of information consistent with a literalist cognitive schema, those who adhere to conservative religious authority may select private religious schooling as an alternative to negatively perceived public schools (see Sikkink 1999). Moreover, secondary education in the United States plays out on a local level, meaning that surrounding moral communities impact how, if at all, evolution gets addressed in public schools (Deckman 2002; Gibson 2004).2
As adults, students may choose to pursue higher education at institutions supporting worldviews contra evolution or avoid courses of study that challenge conservative religious views. Indeed, previous studies suggest an interactive relationship between education level and religious affiliation for predicting acceptance of evolution (Eckberg 1992; Haider-Markel and Joslyn 2008). The combination of cognitive and group processes used to maintain positions of literalism suggests that the relationship between educational attainment, religious identity, and attitudes toward evolution/creationism in the United States is unlikely to be uniform. I test two hypotheses about the impact of educational attainment on acceptance of evolutionary accounts of human origins and support for teaching creationism in public schools.
H1: Increasing educational attainment will have, on average, a positive impact on acceptance of evolutionary accounts of human origins, but this positive impact will be attenuated for those with more biblically literalist religious identities.
H2: Increasing educational attainment will have, on average, a negative impact on supporting teaching creationism in public schools, but this negative impact will be attenuated for those with more biblically literalist religious identities.
DATA AND METHODS
Sample
The data analyzed are from Wave II of the Baylor Religion Survey, a random, national sample of noninstitutionalized American adults with telephones. The first two waves of the survey were designed by Bader, Mencken, and Froese (2007) and collected by the Gallup Organization. Wave
II used a mixed-mode sampling design where potential respondents were called using randomdigit dialing and asked if they would be willing to complete a mailed questionnaire. A total of
3,500 potential respondents were successfully contacted by phone, with 2,460 agreeing to receive the questionnaire and providing a mailing address. The recruitment phase of the survey took place from September 4 to September 29, 2007. The collection of questionnaires was discontinued on
December 11, 2007. A total of 1,648 questionnaires were returned, for a contact-to-completion rate of 47.1 percent and a response rate of 67 percent for the mailed survey phase. The margin of error for estimating population parameters with the sample is ±4 percent.3

accepting the position of mainstream science. This was the standard response given to a wide variety of antagonistic questioning posed by skeptics of creationism, even when science and philosophy faculty pressed him on the ability of the lay public to evaluate evidence in scientific journals. Observational field notes taken at the event are available on request.
2 “Academic

freedom” bills at the state level further create space for positions critical of evolution and supportive of teaching creationism in public schools. For a summary of legislative activity intended to support teaching creationism at the state level, see http://ncse.com/creationism/general/academic-freedom-bills-by-state-year.

3 Potential

respondents received a $5.00 incentive in the mailing for agreeing to complete the 16-page questionnaire.
Mailings included a cover letter explaining the objectives of the survey and a number to call for questions about the

ACCEPTING EVOLUTION; SUPPORTING CREATIONISM

219

Measures
Wave II of the BRS included a battery of questions that read: “Please indicate your level of agreement with the following statements about science.” The two items in the battery on evolution and creationism were worded: “Humans evolved from primates over millions of years” and “Creationism should be taught in public schools.” The former probes levels of agreement with a firm view of evolution by asking about human origins on a geologic timescale. The creationist question is about a position on public policy rather than personal belief about creationism. Response options to these questions were on a Likert scale from “strongly disagree” to “strongly agree.” For interpretive clarity, the measures of acceptance of evolution and support for teaching creationism in public schools were condensed into binary categories of “do not agree” (0) and “agree” (1) for modeling in logistic regression.4 Forty-two percent of respondents accepted evolutionary accounts of human origins, while 44 percent favored the teaching of creationism in public schools.
Notably, accepting evolution and support for teaching creationism were not mutually exclusive, with 12 percent of the sample doing both. In addition, 27 percent of respondents neither accepted evolution nor supported teaching creationism in public schools.5
Education was measured in attainment categories and coded into dummy variables for: less than a high school diploma (reference category for modeling), high school graduate, some college or vocational degree, and college graduate. I included controls for correlates significantly associated with views of evolution and creationism in previous research, as well as standard sociodemographic controls such as income, gender, region, age, race, and urban/rural residence. Annual household income was measured in the following categories: less than $10,000; $10,001–$20,000;
$20,001–$35,000; $35,001–$50,000; $50,001–$100,000; $100,001–$150,000; and more than
$150,000 per year. Gender was coded as a dummy variable with female = 1. Age was measured in years, ranging from 18 to 96. Region of the country was divided into South = 1 and non-South = 0 based on delineations from the Census Bureau. Respondent’s race was measured with a dummy variable where white = 1. Urban/rural context was measured by asking: “Which of the following best describes the place where you now live?” Ordinal answer choices were: rural area, small city or town, suburb near a large city, and large city. Political identification was measured with “How would you describe yourself politically?” Answer choices ranged on

procedure or the survey more generally. Follow-up letters were sent thanking respondents who had returned booklets and asking for the cooperation of those who had not. Reminder postcards were also sent, as well as a second complete mailing to addresses that had not responded. Gallup created a weight based on information from the Census Bureau to achieve better representativeness for the American public using an algorithm based on the following characteristics: education, race, gender, age, and region of the country. The weight was applied in all analyses shown. In supplemental models conducted with unweighted data the relationships of interest remained similar. To assess data validity I compared the BRS II to information from the 2008 Current Population Survey conducted by the Census Bureau, as well as to data from the 2008 General Social Survey. Based on these comparisons, the BRS sample presents a reasonable estimate of population parameters for American adults. Tabled results of these comparisons and all other noted supplemental analyses are available on request.
4I

also conducted ordinary least squares and ordinal logistic regression models. The results for the variables of interest were similar to those presented for the binary logistic models. The ordinal model retains the variance in the dependent variable, but a Brant test showed the parallel regression assumption was violated. For this reason, I present binary logistic models. 5 To explore these patterns further I created four categories: evolution only supporters, creationism only supporters, those who support both, and those who support neither. Respondents with low levels of education were more likely to support neither, likely reflective of inadequate information about the topic. Those with mid-level educational attainment were more likely to support teaching creationism only, while those with a college degree were more likely accept evolution and oppose teaching creationism in public schools. Political moderates were more likely to support neither or both, with conservatives more likely to support teaching creationism only and liberals more likely to accept evolution only. There were also interesting results based on religious tradition, race, and gender. Future research should explore these belief combinations further.

220

JOURNAL FOR THE SCIENTIFIC STUDY OF RELIGION

a seven-point scale from “extremely conservative” to “extremely liberal” with “moderate” as the middle category.
Religious tradition was classified using the RELTRAD schematic (Steensland et al. 2000).
Dummy variables were created for categories of: black Protestant, evangelical Protestant, mainline Protestant, Catholic, Jewish, no religion, and “other” (which combined remaining religious groups). Evangelicals are the reference category in multivariable models due to the history of conservative Protestantism opposing evolution and supporting creationism. Attendance at religious services was measured on a nine-point scale from “never” to “more than once a week.” View of the Bible was measured with a question that asked: “Which of the following statements comes closest to your personal beliefs about the Bible?” Answer choices were: “The Bible is an ancient book of history and legends”; “The Bible contains some human error”; “The Bible is perfectly true, but it should not be taken literally, word-for-word. We must interpret its meaning”; and “The
Bible means exactly what it says. It should be taken literally, word-for-word, on all subjects.”
These answer choices were coded into dummy variables and modeled as discrete categories.
Literalists are the reference category in multivariable models.
Analytic Strategy
I used logistic regression models to estimate the probability of accepting evolutionary accounts of human origins and supporting teaching creationism in public schools for individuals with specific social characteristics. Due to a loss of over 20 percent of the cases to listwise deletion in the final models, I used multiple imputation to restore missing data, generating cases based on five Markov chain Monte Carlo simulations (see Rubin 1987).6 This resulted in five datasets with
1,648 cases. Models were conducted on each dataset separately, with the resulting coefficients pooled and analyzed to reduce the impact of artificially increased statistical power. Standardized coefficients (β) were calculated for the variables in each model using the formula available in
Allison (1999). These coefficients are reported in the main text. Constants, Beta weight coefficients, and –2 log likelihood stats were determined by averaging each of these statistics across the five models (and datasets). I calculated pseudo R2 estimates by determining the reduction in the –2 log likelihood statistic between a model with only an intercept and the one presented (see
Menard 2010:42–62).
For each outcome, I present two models: the first includes sociodemographic and religious predictors, and the second tests the interactions. Interaction terms were created by multiplying the dummy variables for education level by each of the dummy variables for Bible view. Concerning the interactions tested, since biblical literalists are the reference group for Bible view, the “lowerorder” coefficients for education in the second models represent the impact of education for this group. The lower-order coefficients for stance toward the Bible represent the impact of Bible views for those without a high school diploma, which is the reference group for education.
The coefficients for the interaction terms measure the difference between literalists and the Bible view in question concerning the difference between the specified level of education and those without a high school diploma (see Aiken and West 1991). In other words, the interaction terms test for differences across Bible view categories regarding the gap between those with less than a high school diploma and the education level in question. The unstandardized coefficients from the interaction models were used to calculate the predicted probabilities of accepting evolution and supporting teaching creationism in public schools based on education level and view of the
Bible. Due to the specificity and complication in the significance tests for the interaction terms,

6 The results

from multivariable models using listwise deletion for missing cases mirror those presented for imputed data.
The imputation procedure was conducted using the “proc mi” command in SAS, with the results analyzed using “proc mianalyze.” ACCEPTING EVOLUTION; SUPPORTING CREATIONISM

221

the overall relationships between views of the Bible and education for predicting acceptance of evolution and support for teaching creationism in public schools are best understood by examining the graphical figures rather than the coefficients. To estimate predicted probabilities, all variables in the models (other than those being “solved for”) were set to their respective means (see Long
1997).
RESULTS
Table 1 shows the binary logistic models predicting acceptance of evolution. For the general effects of education, only a college degree significantly increases the probability of accepting evolution (β = .176). Concerning the control variables entered in Model 1, women (β =
–.065), those who live in more rural areas (β = –.126), and those who attend religious services frequently (β = –.276) were all significantly less likely to accept evolutionary accounts of human origins. Political liberals (β = .231) were more likely to accept evolution. Catholics (β = .221),
Jews (β = .219), mainline Protestants (β = .158), and religious nones (β = .180) all had more than twice the odds of accepting evolution compared to evangelicals, net of other variables in the model. For view of the Bible, the “history and legends” (β = .422) and “human error” (β = .214) categories were significantly more likely to accept evolution than literalists. The difference between literalists and “history and legends” respondents was the strongest variable in the model.
There was not a significant difference between “interpretationists” and literalists. Model 2 adds the interaction terms for view of the Bible and education.7
The lower-order education and Bible view variables, along with the interaction coefficients, were used to estimate the predicted probability of accepting evolution at all combinations of
Bible view and education, as displayed in Figure 1. For nonliteralists, there is a general positive influence of education on acceptance of evolution. “History and legends” respondents move from a probability of .53 for those with less than a high school diploma to .87 for those with at least a bachelor’s degree. In sharp contrast, the predicted probability for biblical literalists falls as education increases, from .42 for those with less than a high school diploma to .12 for those with some college or a vocational degree, before rising again slightly to .18 for college graduates.
To assess whether education has a significant overall impact on accepting evolutionary accounts of human origins for individuals within each of the Bible view categories, supplemental models entered education as an ordinal variable and rotated the reference category to each of the four
Bible categories. The effect of education for literalists was negative and significant (p = .08), positive but nonsignificant for the interpretation category (p = .138), and positive and significant for the “human error” (p = .025) and “history and legends” (p = .001) categories.
Table 2 displays the results from the models predicting support for teaching creationism in public schools. For the general effects of education, higher levels of attainment had positive effects for supporting the teaching of creationism in public schools, but only a vocational degree or “some college” was statistically significant (β = .133). Concerning control variables in Model
1, residents of more rural areas (β = .069) and those with higher levels of income (β = .064) were slightly more likely to support teaching creationism in public schools, while political liberals
(β = –.228) were much less likely, the strongest variable in the model. Catholics (β = –.152),
Jews (β = –.079), mainline Protestants (β = –.152), and religious nones (β = –.196) were all significantly less likely to support teaching creationism in public schools than evangelicals. Higher levels of worship service attendance increased support for teaching creationism (β = .110). For view of the Bible, those who responded with “human error” (β = –.106) and “history and legends”

7I

replicated the main findings here using data from the 2008 General Social Survey, which also showed a significant interaction between Bible view and education level on acceptance of evolution in a similar multivariable context.

222

JOURNAL FOR THE SCIENTIFIC STUDY OF RELIGION

Table 1: Binary logistic regression models predicting acceptance of evolutionary accounts of human origins
Model 1 β OR

.105
.169
.726*
−.032
−.279†
−.008
−.177
.077
−.266**
.243***

.026
.045
.176
−.019
−.067
−.065
−.041
.039
−.126
.231

1.111
1.184
2.067
.969
.757
.992
.838
1.080
.766
1.275

.026
.967***
3.061**
.708***
1.001**
.001

.003
.221
.219
.158
.180
.000

−.172***
1.732***
1.138***
.176

b
Sociodemographics
High school diplomaa
Some college/vocationa
College graduatea
Income
Female
Age
South
White
Urban/rural
Politically liberal
Religious traditionb
Black Protestant
Catholic
Jewish
Mainline Protestant
No religion
Other religion
Religious practice
Service attendance
Bible viewc
History and legends
Human error
Interpretation
Interactions
History/legends * HS
History/legends * SC
History/legends * CG
Human error * HS
Human error * SC
Human error * CG
Interpretation * HS
Interpretation * SC
Interpretation * CG
Model statistics
Constant
–2 log likelihood
Likelihood R2
N

β

OR

−1.009*
−1.619**
−1.201†
−.082
−.316*
−.009*
−.185
.077
−.270***
.239***

−.248
−.435
−.292
−.072
−.087
−.083
−.047
.015
−.141
.215

.365
.198
.301
.921
.729
.991
.831
1.080
.763
1.270

1.026
2.630
21.349
2.030
2.721
1.001

−.026
1.087***
2.848**
.764***
1.357***
−.357

−.003
.246
.202
.168
.239
−.046

.974
2.965
17.253
2.147
3.885
.700

−.276

.842

−.157***

−.256

.855

.422
.214
.047

5.652
3.121
1.192

.109
.095
−.273

1.571
1.659
.359

.206
.370
.478
.067
.167
.205
.215
.413
.326

5.048
10.329
18.803
2.203
3.706
5.989
3.337
8.706
7.523

–
–
–
–
–
–
–
–
–
−.902
1475.39
.320
1648

< .001, ** p < .01, * p < .05, † p < .1 (two-tailed tests). than high school diploma is comparison category. b Evangelical Protestant is comparison category. c Literalist is comparison category.
*** p

a Less

Model 2 b .452
.506
−1.025
1.619†
2.335**
2.934**
.790
1.310
1.790†
1.205
2.164**
2.018*
.787
1230.77
.372
1648

ACCEPTING EVOLUTION; SUPPORTING CREATIONISM

223

Figure 1
Predicted probability of accepting evolutionary accounts of human origins by Bible view and education level

Source: 2007 Baylor Religion Survey (multiply-imputed data).

(β = –.224) were significantly less likely than literalists to support teaching creationism. As with evolution, there was not a significant difference between literalists and those who said the Bible is “perfectly true but requires interpretation.”
Model 2 adds the interaction terms, the results of which are graphed in Figure 2. Education had an inconsistent relationship with support for teaching creationism for “history and legends” respondents. There are higher probabilities of support for teaching creationism at middle levels of education (over .4), while those without a high school diploma and college graduates had the same, lower predicted probability (.19). Education had a negative impact on support for teaching creationism in public schools among the “human error” category, with those without a high school diploma (.69) having a higher probability of support than those with a high school diploma (.41).
Beyond a high school diploma, there was minimal change for education level among this Bible view category. For those who answered that the Bible is true but requires interpretation, having less than a high school diploma led to a lower level of support (.34) compared to those with a high school diploma (.57) or greater (to .62 for college graduates). For literalists, increased education had a steady positive impact on support for teaching creationism in public schools, from a probability of .5 for those with less than a high school diploma, to .75 for college graduates. Again using supplemental models to assess the impact of education within each Bible view, education had a positive and significant impact on support for teaching creationism for literalists (p = .001), was positive but nonsignificant for interpretationists (p = .179), negative and nonsignificant for the “human error” category (p = .847), and negative and significant for the history and legends category (p = .033). In general, the effects of education level on attitudes toward creationism policy are uneven, ranging from significantly positive to negative depending on views of the Bible.8

8 Results

examining the proportion of respondents who accepted evolution and supported teaching creationism in public schools within each education category separated by Bible views in a trivariate context were similar to those I obtained in the multivariable models. These analyses also revealed an overall heteroskedastic relationship between education level

224

JOURNAL FOR THE SCIENTIFIC STUDY OF RELIGION

Table 2: Binary logistic regression models predicting support for teaching creationism in public schools
Model 1 β OR

β

OR

.297
.496†
.377
.073†
−.087
−.002
.058
−.078
.132*
−.255***

.074
.133
.091
.064
−.024
−.019
.015
−.015
.069
−.228

1.346
1.642
1.458
1.076
.917
.998
1.060
.925
1.141
.775

.118
.738†
1.094*
.077†
−.010
−.001
−.056
−.086
.148*
−.256***

.029
.198
.266
.067
−.027
−.008
−.014
−.017
.077
−.231

1.125
2.092
2.986
1.080
.990
.999
.946
.918
1.160
.774

−.208
−.663***
−1.060*
−.686***
−1.093***
−.436

−.024
−.152
−.079
−.152
−.196
−.056

.812
.515
.346
.504
.335
.647

−.129
−.693***
−.787
−.694***
−1.179***
−.392

−.015
−.157
−.058
−.154
−.206
−.050

.879
.500
.455
.500
.308
.676

.110

1.070

.039

.063

1.040

−.224
−.106
−.038

.397
.569
.869

−1.424
.804
−.661

−.342
.151
−.176

.241
2.234
.516

1.102
.440
−1.094
−1.256
−1.792*
−2.171*
.823
.224
.086

.139
.070
−.178
−.108
−.227
−.249
.146
.043
.014

3.010
1.553
.335
.285
.167
.114
2.277
1.251
1.090

b
Sociodemographics
High school diplomaa
Some college/vocationa
College graduatea
Income
Female
Age
South
White
Urban/rural
Politically liberal
Religious traditionb
Black Protestant
Catholic
Jewish
Mainline Protestant
No religion
Other religion
Religious practice
Service attendance
Bible viewc
History and legends
Human error
Interpretation
Interactions
History/legends * HS
History/legends * SC
History/legends * CG
Human error * HS
Human error * SC
Human error * CG
Interpretation * HS
Interpretation * SC
Interpretation * CG
Model statistics
Constant
–2 log likelihood
Likelihood R2
N

.068**
−.924***
−.563*
−.140
–
–
–
–
–
–
–
–
–
.426
1845.79
.154
1648

< .001, ** p < .01, * p < .05, † p< .1 (two-tailed tests). than high school diploma is comparison category. b Evangelical Protestant is comparison category. c Literalist is comparison category.
*** p

a Less

Model 2 b .526
1635.55
.174
1648

ACCEPTING EVOLUTION; SUPPORTING CREATIONISM

225

Figure 2
Predicted probability of support for teaching creationism in public schools by Bible view and education level

Source: 2007 Baylor Religion Survey (multiply-imputed data).

DISCUSSION
Returning to the question raised at the beginning, the reason public opinion has not followed the movement of institutional science and education toward evolution is that the influence of education on opinions about such matters is conditioned by other factors, particularly “literalist” religious identities. Although literalism has declined since the 1950s, partly due to higher average levels of education, this decline has slowed (see Chaves 2011:33–37; Putnam and Campbell
2011:112). According to the 2010 General Social Survey, one-third of Americans are literalists.
Subcultural consumption of antievolution, procreationist media combined with selective internalization of information about evolution typically accompanies such identities. The diversity of schooling options in the United States and the influence of local, moral communities on how evolution is addressed (if at all) in public schools also play important roles (see Berkman and
Plutzer 2005, 2010).
Recurring conflict over evolution and creationism is the result of “efforts to reconcile public science—that is, publicly supported science teaching and related activities—with popular opinion” (Larson 2003:4). Americans are almost evenly divided on whether “macro,” interspecies evolution accurately explains the history of humanity. A similar split is evident regarding whether creationism should be taught in public schools. Creationist movements still face an uphill battle in bringing about wide-ranging policy or curriculum changes, as scientific elites have put up a publicly unified front against creationism (see Binder 2002; Gieryn, Blevins, and Zehr 1985; and relevant court cases: Edwards v. Aguillard 1987; Epperson v. Arkansas 1968; Kitzmiller et al. v. Dover 2005; McLean v. Arkansas 1982). In spite of the institutionalization of evolutionary perspectives in science, conflict over these issues remains prominent in public imagination,

and stance on evolution and creationism. Those with lower levels of education are more likely to be “undecided” about evolution (24 percent of those with less than a high school diploma). Those at higher levels of education are more likely to take firm positions on these issues (only 11 percent of college graduates were undecided).

226

JOURNAL FOR THE SCIENTIFIC STUDY OF RELIGION

discourse, and policy. Creationism continues to maintain a relatively high level of subcultural resonance. Meanwhile, evolution has won many court battles but has not been able to win a majority in the court of public opinion. Legal decisions about evolution have merely narrowed the limits within which local politics play out (Berkman and Plutzer 2009), and have not produced a large shift in public views.
Advocates for the scientific community have responded by calling for more thorough coverage of evolution in curricula and stinging criticism of the scientific rhetoric of creationism (e.g.,
Scott and Branch 2006); however, improved science education standards alone are unlikely to produce substantial changes in public opinion. Casting doubt on evolution, making creationist arguments compatible with a circumscribed version of evolution, and paying scant attention to evolutionary principles in the classroom are all tactics that have evolved out of the shifting bounds between science and religion, and their relative places in public education. As Evans notes, convincing conservative Protestants that scientists are not biased on issues marking cultural (and therefore moral) boundaries “may be a more intractable educational task, but it probably edges closer to the actual problem” (2011:724).
The influence of religious identity on educational trajectory and selective perception presents unique challenges for advocates of evolution. Addressing these topics in a manner perceived as an attack on creationism, and therefore religion, is likely to strengthen prior convictions of the faithful. Alternative strategies include openly discussing issues of value and philosophy in science classes (Alters and Nelson 2002; Binder 2007; Dagher and BouJaoude 1997), or addressing such topics through courses in other subjects, but these tactics also depend on the educational context in which they are implemented, as well as students’ openness. Whether education results in greater public acceptance of evolution and antipathy toward creationism is more a question of the internalization of perspectives critical of exclusivist religious schema (see Hill 2011; Reimer
2010) than one of science education per se.
Although there are many positions on evolution and creationism, ranging from young earth creationism to materialistic evolution (see Scott 2009), the data analyzed here are limited to general conceptions of these issues. The patterns of more specific positions warrant further inquiry. Concerning education, specification of the role played by different types of education
(private/public, religious/secular) would improve the current findings. Panel data (both qualitative and quantitative) tracking individuals as they progress through various paths of education is also needed to better understand how the process impacts religious identities in general (e.g., Uecker
2009), and views of evolution and human origins specifically.
Regardless of scientific consensus, the passionate divergence of views about evolution and its place in education will not subside quickly or quietly. Changes in public opinion about evolution and creationism ultimately hinge on concordant shifts in the proportion of the population with certain religious identities. To the degree that literalist identities remain intact, higher levels of education can actually lead to less acceptance of evolution and greater support for teaching creationism in public schools. Public opinion on these matters is not a straightforward issue of science education. Rather, it involves religious and political communities’ engagement with, and shaping of, a heterogeneous American education system.

REFERENCES
Aiken, Leona S. and Stephen G. West. 1991. Multiple regression: Testing and interpreting interactions. Thousand Oaks,
CA: Sage.
Allison, Paul D. 1999. Logistic regression using the SAS system: Theory and application. Cary, NC: SAS Publishing.
Alters, Brian J. and Craig E. Nelson. 2002. Perspective: Teaching evolution in higher education. International Journal of
Organic Evolution 56(10):1891–1901.
Ammerman, Nancy T. 1987. Bible believers: Fundamentalists in the modern world. New Brunswick, NJ: Rutgers
University Press.

ACCEPTING EVOLUTION; SUPPORTING CREATIONISM

227

Bader, Christopher, Paul D. Froese, and F. Carson Mencken. 2007. American piety 2005: Methods and select findings from the Baylor Religion Survey. Journal for the Scientific Study of Religion 46(4):447–63.
Baker, Joseph O. 2012. Public perceptions of incompatibility between “science and religion.” Public Understanding of
Science 21(3):340–53.
Bartkowski, John P. 1996. Beyond biblical literalism and inerrancy: Conservative Protestants and the hermeneutic interpretation of scripture. Sociology of Religion 57(3):259–72.
Berkman, Michael B. and Eric Plutzer. 2005. Ten thousand democracies: Politics and public opinion in America’s school districts. Washington, DC: Georgetown University Press.
———. 2009. Scientific expertise and the culture war: Public opinion and the teaching of evolution in the American states. Perspectives on Politics 7(3):485–99.
———. 2010. Evolution, creationism, and the battle to control America’s classrooms. New York: Cambridge University
Press.
Binder, Amy J. 2002. Contentious curricula: Afrocentrism and creationism in American public schools. Princeton, NJ:
Princeton University Press.
———. 2007. Gathering intelligence on intelligent design: Where did it come from, where is it going, and how should progressives manage it? American Journal of Education 113(4):549–76.
Boone, Kathleen C. 1989. The Bible tells them so: The discourse of Protestant fundamentalism. Albany: State University of New York Press.
Brossard, Dominique, Dietram A. Scheufele, Eunkyong Kim, and Bruce V. Lewenstein. 2009. Religiosity as a perceptual filter: Examining processes of opinion formation about nanotechnology. Public Understanding of Science 18(5):
546–58.
Bruce, Steve. 2008. Fundamentalism, 2nd edition Cambridge: Polity Press.
Chaves, Mark. 2011. American religion: Contemporary trends. Princeton, NJ: Princeton University Press.
Dagher, Zoubeida R. and Saouma BouJaoude. 1997. Scientific views and religious beliefs of college students: The case of biological evolution. Journal of Research in Science Teaching 34(5):429–45.
Deckman, Melissa. 2002. Holy ABCs! The impact of religion on attitudes about education policies. Social Science
Quarterly 83(2):472–87.
Eckberg, Douglas L. 1992. Social influences on belief in creationism. Sociological Spectrum 12(2):145–65.
Eckberg, Douglas L. and Alexander Nesterenko. 1985. For and against evolution: Religion, social class, and the symbolic universe. Social Science Journal 22(1):1–17.
Edwards v. Aguillard, 482 U.S. 578 (1987).
Ellison, Christopher G. and Mark A. Musick. 1995. Conservative Protestantism and public opinion toward science. Review of Religious Research 36(3):245–62.
Epperson v. Arkansas, 393 U.S. 97 (1968).
Erikson, Kai T. 1966. Wayward Puritans: A study in the sociology of deviance. Boston: Allyn and Bacon.
Evans, John H. 2011. Epistemological and moral conflict between religion and science. Journal for the Scientific Study of Religion 50(4):707–27.
Gibson, Troy M. 2004. Culture wars in state education policy: A look at the relative treatment of evolutionary theory in state science standards. Social Science Quarterly 85(5):1129–49.
Gieryn, Thomas F. 1983. Boundary-work and the demarcation of science from non-science: Strains and interests in professionalization ideologies of scientists. American Sociological Review 48(6):781–95.
———. 1999. Cultural boundaries of science. Chicago: University of Chicago Press.
Gieryn, Thomas F., George M. Blevins, and Stephen C. Zehr. 1985. Professionalization of American scientists: Public science and the creation/evolution trials. American Sociological Review 50(3):392–409.
Haider-Markel, Donald P. and Mark R. Joslyn. 2008. Pulpits versus ivory towers: Socializing agents and evolution attitudes. Social Science Quarterly 89(3):665–83.
Ham, Ken. 2006. Answers in Genesis. Green Forest, AR: Master Books.
Hill, Jonathan P. 2011. Faith and understanding: Specifying the impact of higher education on religious belief. Journal for the Scientific Study of Religion 50(3):533–51.
Hoffmann, John P. and John P. Bartkowski. 2008. Gender, religious tradition, and biblical literalism. Social Forces
86(3):1245–72.
Hood, Ralph W., Peter C. Hill, and W. Paul Williamson. 2005. The psychology of religious fundamentalism. New York:
Guilford Press
Keeter, Scott, Gregory Smith, and David Masci. 2011. Religious belief and attitudes about science in the United States.
In The culture of science: How the public relates to science across the globe, edited by Martin W. Bauer, Rajesh
Shukla, and Nick Allum, pp. 336–52. New York: Routledge.
Kitzmiller et al. v. Dover Area School District, 400 F. Supp 2d 707 (M.D. Pa. 2005).
Larson, Edward L. 1997. Summer for the gods: The Scopes trial and America’s continuing debate over science and religion. New York: Basic Books.
———. 2003. Trial and error: The American controversy over creation and evolution, 3rd edition. New York: Oxford
University Press.

228

JOURNAL FOR THE SCIENTIFIC STUDY OF RELIGION

Lawrence, Bruce B. 1989. Defenders of God: The fundamentalist revolt against the modern age. Columbia: University of South Carolina Press.
Lawson, Anton E. and William A. Worsnop. 1992. 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(2):143–66.
Long, J. Scott. 1997. Regression models for categorical and limited dependent variables. Thousand Oaks, CA: Sage.
Marsden, George M. 2006. Fundamentalism and American culture, 2nd edition. New York: Oxford University Press.
Mayrl, Damon and Jeremy Uecker. 2011. Higher education and religious liberalization among young adults. Social Forces
90(1):181–208.
Mazur, Allan. 2004. Believers and disbelievers in evolution. Politics and the Life Sciences 23(2):55–61.
McLean v. Arkansas Board of Education, 529 F. Supp. 1255 (E.D. Ark. 1982).
Menard, Scott. 2010. Binary logistic regression: From introductory to advanced concepts and applications. Thousand
Oaks, CA: Sage.
Miller, Jon D., Eugenie C. Scott, and Shinji Okamoto. 2006. Public acceptance of evolution. Science 313(5788):765–66.
Numbers, Ronald L. 1998. Darwinism comes to America. Cambridge, MA: Harvard University Press.
———. 2006. The creationists: From scientific creationism to intelligent design, exp. edition. Cambridge, MA: Harvard
University Press.
Plutzer, Eric and Michael Berkman. 2008. The polls—Trends: evolution, creationism, and the teaching of human origins in schools. Public Opinion Quarterly 72(3):540–53.
Putnam, Robert D. and David E. Campbell. 2011. American grace: How religion divides and unites us. New York: Simon and Schuster.
Reimer, Sam. 2010. Education and theological liberalism: Revisiting the old issue. Sociology of Religion 71(4):393–408.
Rubin, Donald B. 1987. Multiple imputation for nonresponse in surveys. Hoboken, NJ: Wiley.
Scott, Eugenie C. 2009. Evolution vs. creationism: An introduction, 2nd edition. Westport, CT: Greenwood.
Scott, Eugenie C. and Glenn Branch, eds. 2006. Not in our classrooms: Why intelligent design is wrong for our schools.
Boston: Beacon Press.
Sikkink, David. 1999. The social sources of alienation from public schools. Social Forces 78(1):51–86.
Sinclair, Anne and Murray P. Pendarvis. 1998. Evolution vs. conservative religious beliefs. Journal of College Science
Teaching 27(3):167–70.
Smith, Christian. 1998. American evangelicalism: Embattled and thriving. Chicago: University of Chicago Press.
Steensland, Brian, Jerry Z. Park, Mark D. Regnerus, Lynn D. Robinson, W. Bradford Wilcox, and Robert Woodberry.
2000. The measure of American religion: Toward improving the state of the art. Social Forces 79(1):291–324.
Stroope, Samuel. 2011. Education and religion: Individual, congregational, and cross-level interaction effects on biblical literalism. Social Science Research 40(6):1478–93.
Toumey, Christopher P. 1994. God’s own scientists: Creationists in a secular world. New Brunswick, NJ: Rutgers
University Press.
Uecker, Jeremy E. 2009. Catholic schooling, Protestant schooling, and religious commitment in young adulthood. Journal for the Scientific Study of Religion 47(4):563–84.
Woodrum, Eric and Thomas Hoban. 1992. Support for prayer in school and creationism. Sociology of Religion 53(3):
309–21.

SUPPORTING INFORMATION
Additional Supporting Information may be found in the online version of this article at the publisher’s web site:
Figure S1. Acceptance of evolution by Bible view and education level (bivariate).
Figure S2. Support for creationism by Bible view and education level (bivariate).

Copyright of Journal for the Scientific Study of Religion is the property of Wiley-Blackwell and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.