Social Science & Medicine 211 (2018) 274–281

Contents lists available at ScienceDirect

Social Science & Medicine

journal homepage: www.elsevier.com/locate/socscimed

Knowing less but presuming more: Dunning-Kruger effects and the T

endorsement of anti-vaccine policy attitudes
Matthew Mottaa,∗, Timothy Callaghanb, Steven Sylvesterc
a
Annenberg Public Policy Center, University of Pennsylvania, 202 South 36th Street, Philadelphia, PA, 19104, USA
b
Department of Health Policy and Management, Texas A&M University, USA
c
History & Political Science Department, Utah Valley University, USA

A R T I C LE I N FO A B S T R A C T

Keywords: Objective: Although the benefits of vaccines are widely recognized by medical experts, public opinion about
Vaccines vaccination policies is mixed. We analyze public opinion about vaccination policies to assess whether Dunning-
Dunning-kruger effects Kruger effects can help to explain anti-vaccination policy attitudes.
Anti-vax Rationale: People low in autism awareness – that is, the knowledge of basic facts and dismissal of misinformation
Political psychology
about autism – should be the most likely to think that they are better informed than medical experts about the
Health policy
causes of autism (a Dunning-Kruger effect). This “overconfidence” should be associated with decreased support
for mandatory vaccination policies and skepticism about the role that medical professionals play in the pol-
icymaking process.
Method: In an original survey of U.S. adults (N = 1310), we modeled self-reported overconfidence as a function
of responses to a knowledge test about the causes of autism, and the endorsement of misinformation about a link
between vaccines and autism. We then modeled anti-vaccination policy support and attitudes toward the role
that experts play in the policymaking process as a function of overconfidence and the autism awareness in-
dicators while controlling for potential confounding factors.
Results: More than a third of respondents in our sample thought that they knew as much or more than doctors
(36%) and scientists (34%) about the causes of autism. Our analysis indicates that this overconfidence is highest
among those with low levels of knowledge about the causes of autism and those with high levels of mis-
information endorsement. Further, our results suggest that this overconfidence is associated with opposition to
mandatory vaccination policy. Overconfidence is also associated with increased support for the role that non-
experts (e.g., celebrities) play in the policymaking process.
Conclusion: Dunning-Kruger effects can help to explain public opposition to vaccination policies and should be
carefully considered in future research on anti-vaccine policy attitudes.

1. Introduction have been shown to contribute to this phenomenon (Joslyn and

In early 2017, President Trump tapped vaccine skeptic Robert
Kennedy Jr. to chair an administrative panel on the safety of vaccines.
According to Kennedy, the president – who is an outspoken critic of
vaccine science – is critical of “current vaccine policies” (Kaplan, 2017).
Although we do not yet know which specific policies have piqued the
president's suspicions, we do know that a fair number of people share
his skepticism. Nearly one in three U.S. adults oppose mandatory vac-
cine requirements for attending public schools (Joslyn and Sylvester,
2017), and less than half believe that scientists understand the health
effects of the MMR vaccine “very well” (Funk et al., 2017). Mis-
information about the link between vaccines (like MMR) and autism
Sylvester, 2017; see also Lewandowsky et al., 2017).
The prevalence of anti-vaccine (“anti-vax”) policy attitudes in both
the public and the White House is a troubling development. Vaccines
prevent the outbreak of diseases that used to be widespread, saving
millions of lives. However, while vaccination rates for measles, mumps,
and rubella (MMR) and other diseases have remained uniformly high –
at around 90% – in recent years (CDC, 2017; CDC, 2011), anti-vax
policy positions and expert skepticism might encourage support for
policy proposals that could reduce this rate. Understanding why the
public holds anti-vax policy attitudes – despite scientific consensus for
their importance – is therefore an important research question.
In this article, we propose and put to the test a novel theoretical

∗
Corresponding author.
E-mail addresses: matthew.motta@appc.upenn.edu (M. Motta), callaghan@tamu.edu (T. Callaghan), ssylvester@uvu.edu (S. Sylvester).

https://doi.org/10.1016/j.socscimed.2018.06.032
Received 25 January 2018; Received in revised form 11 June 2018; Accepted 23 June 2018
Available online 25 June 2018
0277-9536/ © 2018 Elsevier Ltd. All rights reserved.
M. Motta et al.
framework for making sense of why individuals hold anti-vax policy
attitudes. First, drawing on recent work in social psychology, we argue
that individuals low in autism awareness – which includes not only the
knowledge of basic facts about autism, but the dismissal of mis-
information about the link between vaccines (like MMR) and autism –
should be the most likely to think that they know more about the causes
of autism than medical and scientific experts. This “overconfidence”
can be thought about as a type of Dunning-Kruger effect (Dunning,
2011), in which individuals who lack expertise fail to accurately ap-
praise their own knowledge vis-a-vis experts on the subject.
Second, we suspect that people who think they know more than
medical experts will take issue with experts' role in the policymaking
process related to vaccination. Specifically, we argue that over-
confidence about the causes of autism will be associated with increased
opposition to mandatory vaccination policy (which is endorsed by most
medical professionals; CDC, 2017), with decreased support for the role
that medical experts play in informing the public and crafting vaccine-
relevant policy. Notably, while a link between Dunning-Kruger effects
and anti-vax policy attitudes has been theorized to exist in the recent
past (Camarago and Grant, 2015), we are not aware of any research
testing this claim. We conclude by noting that insights from social
psychology – including the study of Dunning-Kruger effects – can be
useful in understanding why U.S. adults oppose vaccination policies
endorsed by medical professionals.
1.1. Explaining anti-vax policy attitudes in the american mass public
Why do large segments of the American public hold anti-vax policy
attitudes despite consensus in the medical community that vaccines
prevent the development and spread of disease? Public attitudes to-
wards vaccinations are complex and driven by a mix of scientific,
psychological, sociocultural, and political factors (Larson et al., 2011;
Carpiano and Fitz, 2017). Research suggests that limited knowledge
and misinformation about vaccines play a vital role in public attitudes.
U.S. adults are generally uninformed or misinformed about the safety of
vaccines, particularly concerning their rumored link to autism. A wide
range of sources, internet blogs, celebrity activism, and various media
point to this link despite no validated scientific evidence supporting this
link (Bean, 2011; Brown et al., 2010; Kata, 2010, 2012). The ease of
access to misinformation increases skepticism about science and an
outright disregard of scientific evidence. For example, research has
shown that misinformation has led to a perception among some that
mandatory vaccinations are a result of influence from the pharmaceu-
tical industry and pushed many to pay more attention to the “risks” of
vaccines instead of their vital role to public health (Larson et al., 2011,
2014; Jolley and Douglas, 2014). Critically, people who endorse mis-
information like this have been shown to be more likely to hold anti-vax
policy attitudes (Joslyn and Sylvester, 2017).
Related to misinformation, the endorsement of anti-vaccine con-
spiracy theories may also sour public opinion toward vaccine safety and
expert research on the subject (Jolley and Douglas, 2014). Conspiracy
theories can be thought about as a special type of misinformation –
unverified (and potentially un-falsifiable) claims of malfeasance on
behalf of powerful people (Flynn et al., 2017). As Jolley and Douglas
(2014) review, prominent anti-vaccine conspiracy theories allege that
governments and pharmaceutical industries purposefully “cover up”
research demonstrating the hazards of vaccines in order to accomplish
various political and financial goals. Individuals who exhibit low levels
of generalized interpersonal trust and domain-specific knowledge
(Miller et al., 2016) may be especially likely to endorse conspiracy
theories about vaccines. Media consumption habits also likely play a
role, with online blogs and other resources influencing the dissemina-
tion of anti-vaccine information (Kata, 2012).
While limited knowledge and misinformation about vaccines (and
their potential link with autism) are associated with vaccine attitudes,
research has also pointed to the roles of ideology, religiosity, and
275
Social Science & Medicine 211 (2018) 274–281
education in this process. For example, Rabinowitz et al. (2016) found
that liberals were more likely than conservatives to support pro-vaccine
statements (also see Joslyn and Sylvester (2017) who find that a related
concept, Republican partisan identification, is associated with increased
opposition to vaccines). While ideology is correlated with attitudes
toward vaccines and vaccine-related issues in the aggregate, we offer
the caveat that ideology is not grounded in issue positions for most
individuals (Kinder and Kalmoe, 2017) and may better be thought about
as a form of social identity (Mason, 2018). Lower levels of education
(Prislin et al., 1998) and increased religious service attendance (Pelcic
et al., 2016; Ruijs et al., 2011, 2013; Shelton et al., 2013) are also as-
sociated with increased vaccine skepticism. These findings suggest that
both ideology and education influence attitudes towards vaccinations,
which appears to be consistent with the research on public attitudes
towards science more broadly (Blank and Shaw, 2015).
1.2. A new direction: expert skepticism and Dunning-Kruger effects
Largely absent from the research on attitudes towards vaccinations
is the possibility that citizens' attitudes toward medical experts also
shape the endorsement of anti-vax policy positions. We believe that this
omission is notable for two reasons. First, expert opinion tends to be
featured prominently in debates about vaccine safety. While the sci-
entific community in the United States has authoritatively concluded
that vaccines like MMR do not cause autism (Nelson and Bauman,
2003), norms of journalistic balance have lead news outlets to give
roughly equal coverage to expert-endorsed “pro-vax” stances and non-
expert endorsed anti-vax stances (Clarke, 2008; Dixon and Clarke,
2013). This strategy suggests the potential for a link between how
people think about medical experts and anti-vax policy attitudes.
Second, an important line of research details the complications ex-
perts face when attempting to communicate pro-vaccine messages with
the public. People who hold negative attitudes toward scientific experts
– a sentiment that has become increasingly common on the ideological
right (Gauchat, 2012; Motta, 2018) – tend to be less accepting of sci-
entific consensus on a variety of matters of scientific and political im-
portance (Motta, 2018). This general phenomenon has important im-
plications for vaccines. In a series of experimental studies, Nyhan,
Reifler, and colleagues found that information about vaccine safety
from the Center for Disease Control (CDC) successfully reduced mis-
perception endorsement about potential hazards of the flu (Nyhan and
Reifler, 2015) and MMR vaccines (Nyhan et al., 2014), but failed to
improve vaccination intention amongst individuals concerned about
vaccine safety. How people think about medical experts has the po-
tential to influence anti-vaccine attitudes and behavior, further under-
scoring the potential policy relevance of anti-expert attitudes.
We build on this research by suggesting that U.S. adults might also
endorse anti-vax policy positions because they believe that they are
comparatively more knowledgeable than medical experts, which we
label the Overconfidence Thesis. According to this perspective, people
who view themselves as comparatively more expert than medical pro-
fessionals should be more likely to oppose the role experts play in the
policymaking process, and the policies they support such as vaccina-
tion.
This theory has received some conceptual attention in previous
literature. For example, Camarago and Grant (2015) note that vaccine
safety skeptics often lack medical expertise themselves and argue that
the “inability of anti-vaccine enthusiasts to correctly gauge their own
skills” may help explain why people adopt anti-vax policy positions.
This argument draws on the social psychological concept of meta-ig-
norance – or the “ignorance of ignorance” (Dunning, 2011) – to suggest
that poorly informed or misinformed individuals lack the information
necessary to accurately appraise their own knowledge of the subject.
The Overconfidence Thesis can be thought about as a type of
Dunning-Kruger effect (Kruger and Dunning, 1999). The term “Dun-
ning-Kruger effect” is a label given to observations of meta-ignorance
M. Motta et al. Social Science & Medicine 211 (2018) 274–281

and its consequences in many domains, concerning assessments of both overconfidence, we think partial or total mediation is the most likely
abstract (e.g., cognitive ability) and specific (e.g., knowledge about a outcome. However, we leave this expectation somewhat open-ended
certain subject) skills. One well-documented effect is the idea that and provide formal mediation assessments while testing Hypothesis 2.
people who lack skill in some area tend to not only be poor judges of
their own skill, but also struggle to accurately assess the skills of others
2. Method
(Ehrlinger, 2008; Dunning, 2011). Kruger and Dunning (1999) find, for
example, that people who perform poorly on abstract grammar tests
To test our theoretical expectations, we fielded an online survey in
both overestimate their own abilities and underestimate the abilities of
December 2017 via Survey Sampling International. Our survey asked
those who performed comparatively better.
respondents to weigh in on several topics relevant to health policy and
Applying the study of meta-ignorance to anti-vax policy raises an
current events, including the safety of genetically modified food, atti-
important question. How should we conceptualize “skill” when ap-
tudes about plans to combat prescription opioid abuse, and climate
praising the safety of vaccines? Given the prominence of concerns about
change. The sampling firm invited 1539 people to take part in the study
a potential link between vaccines and autism in debates about vacci-
on our behalf, which had a completion rate of 85% (final N = 1310).
nation policy, we propose that an understanding of the science behind
We weighted these data to reflect population benchmarks on gender,
autism is necessary (although perhaps not sufficient) to accurately as-
race, income, and education, which we drew from the U.S. Census 2017
sess vaccine safety. To this end, we introduce autism awareness as a skill
Current Population Survey (CPS). Although the un-weighted data are
criterion, which we conceptualize to include both knowledge and the
not far off from these benchmarks, Table A1 in the appendix shows that
dismissal of misinformation about autism.
the weights improve representativeness. The institutional review
As Kuklinski et al. (2000) argue, being informed requires both
boards at Texas A&M University and Utah Valley University approved
holding factual beliefs (i.e., knowledge) and accurate beliefs (i.e., the
the study.
avoidance of misinformation) about a particular subject. We borrow
We measured autism awareness using two sets of questions: knowl-
this conceptualization, but prefer to call people failing to meet both
edge about autism and misinformation about the link between vaccines
criteria as being “low in autism awareness” rather than “ignorant” be-
and autism. First, we assessed knowledge about autism using a ten-
cause people who are misinformed are not ignorant per se. Misinformed
question true/false test developed by the National Alliance for Autism
people do hold at least some amount of information (albeit inaccurate)
Research (which merged with Autism Speaks in 2007). The resulting
about a particular subject (Flynn et al., 2017) and may hold correct
measure is a count of correctly answered items, coded to range from 0
factual beliefs in other related areas (Kahan et al., 2012; Kahan, 2015).
to 1. The test, correct answers, and more information about those an-
Our theoretical position on this issue is closest to Dunning's (2011),
swers can be found in the Supplementary Materials.
who argues that misinformation is ignorance “in disguise.” People who
This measure has the benefit of being administered in nationally
are misinformed may overestimate their expertise because they believe
representative public opinion research in the past (and therefore is
that experts' knowledge – presumably at odds with misinformation – is
useful for benchmarking our own results) and is endorsed as a knowl-
deficient in some way.
edge test by a leading autism research and advocacy organization.
Pursuant with the discussion above, we break down the
However, we recognize that this measure is not without some faults.
Overconfidence Thesis into two sets of testable hypotheses. Hypothesis 1
The true/false format of the response options, for example, may apply
suggests that individuals low in autism awareness – that is, low in
to some questions more readily than others, raising concerns about
autism knowledge (Hypothesis 1A), and high in autism misinformation
falsely scoring answers as “incorrect” that reflect some amount of
endorsement (Hypothesis 1B) – should be more likely to express over-
knowledge (see: Lupia, 2015). While we note that average levels of
confidence in their own knowledge about autism, relative to medical
knowledge on this test are nevertheless moderately high (Table 1), we
experts. This hypothesis represents our attempt to integrate Dunning-
believe that the development of an unambiguous autism knowledge test
Kruger effects into the study of citizens' attitudes about vaccines and
is an important direction for future research.
autism, as well as to provide an empirical test of Camarago and Grant's
Second, we assessed autism misinformation based on responses to
theorizing on the subject.
the following question: “Can vaccines administered to children at
We also suspect that overconfidence has important policy con-
sequences. People who believe that medical professionals know less
Table 1
than they do about autism should exhibit more reservations toward Summary statistics for key dependent and independent variables.
mandatory childhood vaccinations against MMR - an expert endorsed
Role Variable Mean SD N
policy (Hypothesis 2a) – and the role that experts play in the vaccina-
tion policymaking process more generally (Hypothesis 2b). Although DV Table 2 Overconfidence 0.29 0.32 1307
there are many paths by which experts can influence the policymaking DV Table 3 Mandatory Vaccine Opposition 0.51 0.34 1304
process, we focus on two here; their indirect role in informing the DV Table 3 Expert Role in Policymaking 0.61 0.49 1303
public and policymakers about technical issues, and their direct influ- DV Table 3 Non-Expert Role in Policymaking 0.38 0.49 1303
DV Table 3 Expert Information Trust 0.80 0.23 1309
ence in constructing public policy. We refer to negative attitudes to- DV Table 3 Non-Expert Information Trust 0.42 0.26 1306
ward these roles as “expert discounting.” Control Autism Knowledge 0.70 0.19 1310
We think that expert discounting could take one of two forms. One Control Autism Misinformation 0.41 0.31 974a
possibility is that highly overconfident people hold negative attitudes Control Ideology (Conservatism) 0.48 0.28 1304
Control Gender (Female) 0.52 0.50 1305
toward the role that scientists and other medical experts play in the
Control Degree Attainment 3.13 0.84 1310
policymaking process. Another more indirect form is the possibility that Control Race (Black) 0.10 0.30 1310
highly overconfident individuals discount experts not by harboring Control Race (Hispanic) 0.25 0.43 1310
skepticism toward scientists and medical doctors, but instead by ele- Control Income 0.57 0.26 1306
vating the role that non-experts might play in the policymaking process. Control Age 0.36 0.24 1289

We test both possibilities in the pages that follow.

Finally, while we clearly expect a direct effect of overconfidence on
anti-vax policy attitudes, we recognize that autism awareness may also
be associated with these anti-vax attitudes; either directly or indirectly
via overconfidence (i.e., mediated either partially or totally by over-
confidence). Because we expect autism awareness to be associated with
Note. Unweighted summary statistics for key variables. All variables are scaled
to range from zero to one. For additional context, we offer a comparison of our
weighted and unweighted data to national benchmarks in the Supplementary
Materials (Table S1). aIndicates that variable was administered as part of a split
ballot (N = 977) and is therefore expected to have a lower number of non-
missing observations.

276
M. Motta et al.
young ages cause them to become autistic?” Responses were given on a
four-point scale, coded to range from 0 (“They definitely can”) to 1
(“They definitely cannot”). This item was featured as part of a split
ballot question wording experiment, such that three fourths of the
sample saw “definitely can” response options, while one quarter were
exposed to “definitely do” response options (with appropriate altera-
tions to the main text). This factor reduced our sample size by 25% in
our misinformation analyses. We opted not to pool responses across
conditions, due to the presence of significant split ballot effects.
Next, we measured overconfidence using two items. The first item
asked respondents whether they think they know more or less than
“medical doctors” about the causes of autism. The second item did the
same for “scientists.” Response options were coded to range from 0 (“I
know a lot less”) to 1 (“I know a lot more”) on a six-point scale and
averaged together for the two groups. The resulting index was highly
internally consistent (α = 0.90).
To clarify, while our questions expressly asked respondents to focus
on the causes of autism, we nevertheless recognize that feelings of what
we call overconfidence may be based on more than just comparative
assessments of relative knowledge about the causes of autism. Personal
experiences with sickness, the health care system, and doctors are also
likely to matter (see: Poltorak et al., 2005). We see our research as a
starting point for understanding this form overconfidence, and welcome
future ethnographic, mixed-method, and other attempts to better un-
derstand the nature and origins of this phenomenon.
We next measured anti-vax policy attitudes based on the extent to
which respondents agreed or disagreed (a standard five-point Likert
scale) with the following prompt; “Parents should be able to decide
NOT to vaccinate their children against measles, mumps, and rubella.”
Answers were coded to range from 0 to 1, such that a score of 1 in-
dicates strongly opposing mandatory vaccinations.
We then measured expert discounting using two different sets of
measures. First, we asked respondents how much they trust several
different groups to provide them with “information about vaccina-
tions.” Responses were given on a four-point scale ranging from “Not at
all” to “A great deal,” and again coded to range from 0 to 1. Trust in
“medical experts” combines responses to two items (“The Centers for
Disease Control” and “your personal doctor; ” α = 0.60), while “non-
experts” combines response to two other prompts (“Friends/Family”
and “Celebrities; ” α = 0.76).
Second, we asked respondents to report what role (if any) various
groups should have in “making decisions about policy issues related to
the vaccination of children under the age of 16.” Responses were given
on a three-point scale ranging from 0 (“no role at all”) to 1 (“a major
role”) for two groups; “medical scientists” and “the general public.” We
re-coded these responses such that a score of 1 indicates that re-
spondents see a “major role” for a particular group and 0 indicates that
they would prefer a “minor role” or “no role at all” for that group.
Additionally, our models account for several demographic, political,
and social controls that could alternatively explain overconfidence and/
or anti-vax policy attitudes. As noted above, increased religiosity,
educational attainment, and (perhaps) ideological conservatism are
associated with negative attitudes toward vaccines. Additionally,
scholars have noted a link between of several demographic measures
and skepticism about vaccine safety. U.S. adults who are White (Prislin
et al., 1998), Hispanic (Freed et al., 2010), low-income (Armstrong
et al., 2001), and female (Freed et al., 2010) tend to be more skeptical
about vaccine safety.
Consequently, we control for the following variables, all coded to
range from 0 to 1; political ideology (a seven point scale ranging from
“extremely liberal” to “extremely conservative”); religious service at-
tendance (a five point scale ranging from “very inactive” to “very ac-
tive”); gender (a dichotomous measure of whether the respondent is
male or female); educational attainment (a nominal measure of re-
spondents' highest earned degree); race (dichotomous indicators of
whether respondents are Black or Hispanic); total yearly household
277
Social Science & Medicine 211 (2018) 274–281
income (a ten-point scale ranging from low [less than $5000] to high
[greater than $150,000]); and age (recoded to range from 0 to 1, such
that a score of 1 reflects the oldest person in the dataset).
Finally, we noted that levels of missing data in the survey were low.
Across the entire set of variables studied here, only 2.50% (N = 34) of
respondents failed to answer one or more questions (excluding the
vaccine misinformation item, which as noted above was administered
as part of a split ballot). Consequently, we did not impute missing data
on any measures.
3. Results
We began our analysis by assessing the prevalence of over-con-
fidence – that is, the degree to which individuals believe that they know
more than medical doctors and scientists about the causes of autism – in
our sample, and by describing the descriptive statistics of other key
variables (see Table 1). To facilitate interpretation of later results,
Table 1 also notes the role that each variable plays in analyses.
Table 1 presents two sets of findings that are relevant to our tests of
the Overconfidence Thesis. First, it shows that overconfidence was fairly
common in our sample. When we decomposed the overconfidence
measure into its two constituent parts (described in the methods sec-
tion), we found that more than a third of our sample believed that they
knew as much as or more than medical doctors (36%) and scientists
(34%) about the causes of autism. The idea that a substantial number of
U.S. adults think that they know more than medical experts about the
causes of autism is well-reflected in our sample. The statistics further
suggest that many respondents trust information from experts and think
that they should play a major role in the policymaking process
(Table 1). However, many also place high levels of trust on information
from non-experts (42%) and feel that non-experts should play a major
policymaking role (38%). We also noted that while our sample was
moderately well informed about the causes of autism, misinformation
endorsement was also fairly high.
Next, we tested for a link between autism awareness and over-
confidence. Because all variables were coded to range from 0 to 1,
coefficients could be interpreted as percent change in the outcome
variable, moving from the minimum to maximum value of each in-
dependent variable. Irrespective of model specification, both knowl-
edge (H1a) and misinformation (H1b) were strongly and significantly
(at the p < 0.05 level, two-tailed) associated with increased confidence
in one's own knowledge, relative to that of medical and scientific ex-
perts. By even the most conservative estimates, moving from low to
high levels of autism knowledge was associated with a 39% decrease in
overconfidence, while moving from low to high levels of misinforma-
tion endorsement was associated with a 17% increase in overconfidence
(Table 2). Still, while these differences were substantively large, they do
not provide a sense of exactly how overconfident people with low levels
of autism awareness are.
Fig. 1 plots predicted levels of overconfidence across observed levels
of autism knowledge (the left-hand panel) and misinformation (the
right-hand panel), holding all other covariates constant. Recall that a
score of 0.50 on this scale indicates that people consider themselves to
be about equally well informed about autism, compared to doctors and
scientists. People lowest in autism knowledge were projected to earn a
score of about 0.59 on this scale, indicating that they, on average,
viewed themselves as somewhat better informed than the experts. This
quantity dropped precipitously as autism knowledge increases, to about
0.16 for people scoring highest on the test. We found a similar pattern
of results for people high in misinformation, although the differences
were less stark. People who most strongly endorsed misinformation
about the link between vaccines and autism were projected to earn a
score of 0.40 on this scale, indicating that they viewed themselves as
only somewhat less informed than experts. Those who rejected mis-
information, however, were projected to score significantly lower at
about 0.19.
M. Motta et al. Social Science & Medicine 211 (2018) 274–281

Table 2 overconfidence, although we noted that this effect did not consistently
Associations of knowledge and misinformation with other factors. meet conventional levels of statistical significance across models. We
Variable or statistic Knowledge Misinformation Combined also found that wealthier and older individuals were less likely to ex-
hibit overconfidence across models.
Autism Knowledge −0.44* – −0.39* While the correlates of overconfidence are both interesting and
(0.07) (0.08)
consistent with theoretical expectations, they were less likely to be
Vax. Misinformation – 0.22* 0.17*
(0.05) (0.06)
politically consequential if they do not shape citizens' attitudes toward
Conservatism −0.06 −0.15* −0.12* anti-vax policy stances. Table 3 presents the results of several models
(0.05) (0.06) (0.06) designed to test Hypothesis 2a (Columns 1–2) - the possibility that
Religiosity 0.17* 0.16* 0.14* overconfidence increases opposition to mandatory vaccination policy -
(0.04) (0.04) (0.04)
and 2b (Columns 3–10) - the possibility that overconfidence leads to
Female −0.03 −0.04 −0.02
(0.03) (0.03) (0.03) expert discounting. We assessed knowledge and misinformation sepa-
Education 0.03+ 0.03 0.04* rately in order to preserve a large sample size (recall that the mis-
(0.02) (0.02) (0.02) information question was only administered to three-fourths of the
Income −0.16* −0.24* −0.24*
sample).
(0.07) (0.08) (0.08)
Black 0.02 −0.06 −0.08+
Additionally, given the potential for mediation in our theoretical
(0.05) (0.05) (0.05) framework, we provided formal mediation tests below each set of es-
Hispanic −0.03 −0.05 −0.05 timates. The first row performs a Baron and Kenny (1986) style test for
(0.03) (0.03) (0.03) mediation of the effect of knowledge/misinformation via over-
Age −0.33* −0.33* −0.25*
confidence. The second row uses Imai et al.’s (2011) MEDIATE software
(0.06) (0.07) (0.07)
β0 0.66* 0.40* 0.61* for Stata 13 to calculate the total effect – i.e., the direct effect of
(0.08) (0.08) (0.10) knowledge (or misinformation) plus any effects of either one explained
N 1276 951 951 through overconfidence – on each outcome variable. The third row then
calculates the proportion of that total effect that is attributable to the
R2 0.25 0.24 0.27.
mediating effects of knowledge (or misinformation) explained through
Note. Tabled are OLS coefficients (and their standard errors). Outcome variable
overconfidence. Given the observational nature of the data, it was cri-
is overconfidence in one's knowledge about the causes of autism, relative to
scientists and medical doctors. Increased scores on this scale translate to in- tical that we controlled for factors that could alternatively account for
creased confidence in one's own knowledge. Data are weighted. the effects observed at each step of the model. We therefore included all
*p < 0.05, +p < 0.10 (two-tailed). control variables in Tables 1 and 2 in these tests. Additional information
about these models can be found in the Supplementary Materials.
Several of the controls included in our models were associated with First, we found evidence consistent with the expectation that
overconfidence (Table 2). Religiosity, for example, was associated with overconfidence is associated with increased opposition to mandatory
increased overconfidence, consistent with research suggesting that vaccination policy (Columns 1 and 2; Hypothesis 1a). In model 1
more religious individuals are more likely to hold anti-expert attitudes (which controlled for autism knowledge), overconfidence was nega-
(e.g., Gauchat, 2008, 2012). Contrary to expectations (Gauchat, 2012; tively and significantly associated with support for mandatory vacci-
Motta, 2018), we found some evidence of a link between liberalism and nations. The size of this effect was also substantively large.
Overconfidence increased the predicted likelihood that people strongly

Fig. 1. Predicted effects of knowledge and misinformation on overconfidence.

Linear predictions derived from Table 1, holding all other covariates at their sample means. Shaded areas correspond to 95% confidence intervals.

278
M. Motta et al. Social Science & Medicine 211 (2018) 274–281

Table 3
The effect of overconfidence on expert policy attitudes.
Mandatory Vaccine Support Policy Role Support Trust Information About Vaccines

(Ordered Logistic Estimates) (Logistic Estimates) (OLS Estimates)

Experts Experts Non-Exp. Non-Exp. Experts Experts Non-Exp Non-Exp

Overconfidence −0.85* −0.51 0.31 0.51 0.76* 0.97* −0.01 0.02 0.21* 0.26*
(0.35) (0.40) (0.34) (0.41) (0.34) (0.38) (0.04) (0.05) (0.04) (0.04)
Autism Knowledge 0.83 – 1.28* – −1.35* – 0.03 – −0.32* –
(0.57) (0.65) (0.67) (0.08) (0.06)
Vax. Misinfo – −1.98* – −0.95* – −0.02 – −0.16* – 0.10*
(0.46) (0.42) (0.40) (0.05) (0.04)
Conservatism −0.33 −0.34 −0.83* −0.75 0.18 0.18 −0.14* −0.14* −0.11* −0.11*
(0.38) (0.45) (0.39) (0.47) (0.37) (0.45) (0.04) (0.04) (0.04) (0.05)
Religiosity −0.85* −0.59+ −0.48+ −0.28 0.35 0.26 0.04 0.05 0.10* 0.08*
(0.29) (0.34) (0.29) (0.34) (0.29) (0.34) (0.03) (0.04) (0.03) (0.04)
Female 0.01 0.15 −0.08 0.09 0.05 −0.12 −0.01 −0.01 −0.01 −0.04
(0.18) (0.21) (0.22) (0.26) (0.22) (0.25) (0.02) (0.03) (0.02) (0.03)
Education 0.29* 0.27* 0.33* 0.25 −0.00 0.01 0.03+ 0.02 −0.00 −0.01
(0.12) (0.13) (0.13) (0.16) (0.14) (0.16) (0.02) (0.02) (0.01) (0.02)
Income −0.73 −0.64 0.59 0.89 0.26 0.05 0.03 0.08 −0.04 −0.00
(0.44) (0.52) (0.53) (0.63) (0.54) (0.60) (0.09) (0.12) (0.06) (0.07)
Black −0.90* −0.60+ −0.14 0.02 0.17 0.54 −0.00 0.03 −0.02 0.00
(0.35) (0.36) (0.39) (0.45) (0.37) (0.41) (0.04) (0.05) (0.03) (0.04)
Hispanic −0.01 0.08 −0.23 −0.06 0.15 0.25 −0.01 0.04 −0.01 −0.01
(0.23) (0.24) (0.24) (0.27) (0.24) (0.26) (0.03) (0.03) (0.03) (0.03)
Age 0.24 0.20 0.95+ 1.21* 0.34 −0.03 0.18* 0.15* −0.06 −0.14*
(0.40) (0.46) (0.51) (0.60) (0.48) (0.53) (0.06) (0.06) (0.06) (0.06)
β0 – – −1.38* −0.44 −0.45 −1.12+ 0.66* 0.73* 0.64* 0.41*
(0.61) (0.60) (0.62) (0.63) (0.08) (0.07) (0.06) (0.07)
τ1 −1.46* −2.57* – – – – – – – –
(0.50) (0.55)
τ2 −0.17 −1.21* – – – – – – – –
(0.48) (0.52)
τ3 0.61 −0.45 – – – – – – – –
(0.48) (0.51)
τ4 1.82* 0.80 – – – – – – – –
(0.50) (0.52)
N 1271 948 1271 946 1272 948 1276 951 1274 950

Mediation Test Total NA NA NA Partial Total NA NA Partial Partial

Total Effect 0.20 – – – 0.17 0.36 – – 0.31 0.12
% Mediated 28% – – – 18% 36% – – 23% 36%

Note. Variables are as follows: support for mandatory MMR vaccinations (columns 1–2); support for “medical scientists” (3–4); and “the general public” (5–6) playing
a major role in vaccine policy decision making; viewing the Center for Disease Control and medical doctors (columns 7–8) and “friends and family” and celebrities
(9–10) as trustworthy sources of information about vaccines. Mediation tests are not estimated when the overconfidence variable is not statistically significant (as
mediation cannot possibly occur in such cases).
*p < 0.05, +p < 0.10 (two-tailed).

agreed with giving parents the latitude to not vaccinate their kids by
nearly 15%; doubling in size from 16% for the least overconfident, to
30% for the most.
We also found evidence consistent with the idea that overconfidence
mediates the independent effect of knowledge on anti-vax policy atti-
tudes. Supplemental Baron-and-Kenny-style mediation tests (summar-
ized in the row labeled “Mediation Test”) showed that in absence of the
overconfidence measure, autism knowledge was positively and sig-
nificantly associated with support for mandatory vaccinations, sugges-
tive of total mediation (i.e., because the term dropped from significance
when overconfidence was added to the model). Additionally, more than
a quarter (28%) of the total effect of knowledge on anti-vax policy at-
titudes was explained through overconfidence, providing strong evi-
dence of mediation.
We also found that overconfidence yielded a negative effect in
model 2 (which controls for misinformation), although it fell short of
attaining statistical significance. We suspect that this estimate fell short
for two reasons. First, when calculating the predicted probability of
strong agreement with giving parents the latitude to decide not to
vaccinate their kids, we found a smaller, but still substantively mean-
ingful, effect of overconfidence, associated with about an 8% change in
the predicted probability of support. Second, we noted about a 5%
increase in the size of the standard error term associated with over-
confidence the misinformation model, which may be attributable to its
decreased sample size. We conclude that the data are reasonably con-
sistent with Hypothesis 2a, but caution that these results may be sen-
sitive to model specification strategy.
Next, we considered the effect of overconfidence on attitudes to-
ward the role that experts ought to play in the vaccination policy-
making process (our first test of Hypothesis 2b). There were no statis-
tically significant effects of overconfidence on support for the role that
experts play in the policymaking process (Columns 3–4). However, we
did find positive and statistically significant effects of overconfidence
on support for the role that non-experts play in constructing public
policy about vaccines in both the knowledge and misinformation
models. In the former, moving from the minimum to maximum ob-
served values of overconfidence translates to an 18% increase in the
likelihood of respondents thinking that non-experts should play a
“major role” in policy-making decisions about vaccination; from 32% at
low levels of overconfidence to 50% at high levels. We found a similar
pattern of effects in the misinformation model, with overconfidence
shifting support from 31% to 53%. Both models provide evidence of
mediation, although the indirect effect of misinformation channeled
through overconfidence appears to be considerably stronger (double in

279
M. Motta et al.
size, in fact) than that for autism knowledge.
This pattern of results provides strong evidence in favor of
Hypothesis 2b. Specifically, it upholds the idea that overconfidence
does not directly depress citizens' attitudes toward experts in the pol-
icymaking process, but instead indirectly elevates the role people as-
cribe to non-experts. We uncovered a strikingly similar pattern in our
final set of analyses, focused on who citizens trusted to provide them
with information about vaccines. We tested this aspect of Hypothesis 2b
in Columns 7–10. Again, we found no significant effects of over-
confidence on trust in information provided by medical professionals
(in fact, the parameter estimates were nearly zero in both cases).
However, overconfidence was associated with large increases in trust in
non-experts. In the knowledge model, overconfidence boosted trust by
21%, from a predicted trust level of 36% at low levels of overconfidence
to 57% at high levels of overconfidence. We found an even stronger
pattern of results in the misinformation model, with overconfidence
associated with a 26% increase in trust (from 35% to 61%, respec-
tively). Again, both models reveal strong evidence of partial mediation.
Finally, we noted that the control variables included in each model
exhibited generally reasonable relationships with each outcome vari-
able. There are several highlights worth mentioning. First, ideological
conservatism was associated with decreased trust in expert information
about vaccines, as we might expect given aforementioned work on the
link between ideology and trust in scientific experts. Conservatism did
not appear to be associated with pro-vaccine policy attitudes, reflective
of a broader debate in the literature about this point referenced earlier.
Religiosity was (as expected) associated with anti-vaccine policy sup-
port, but surprisingly bore no clear relationship with trust in experts, or
support for their role in the policymaking process. Consistent with prior
research, we found that educational attainment was associated with
increased support for pro-vaccine policy but noted that other demo-
graphics like gender and income bore no clear relationship with this
outcome or any of the others.
4. Discussion
This study presented two main sets of findings. First, we found that
people who know the least about the causes of autism (Hypothesis 1a)
and who are misinformed about the link between vaccines and autism
(Hypothesis 1b) are the most likely to think that they know more than
medical professionals about the causes of autism – a Dunning Kruger
effect. Second, we showed that this “overconfidence” has important
implications for vaccine policy. People who think that they know more
than medical professionals are less supportive of pro-vaccine policies
(Hypothesis 2a) and more likely to elevate the role that non-experts
play in the policymaking process (Hypothesis 2b).
One potential conceptual objection to the present findings concerns
the possibility that overconfidence is simply a “proxy” for low levels of
trust in scientific experts. According to this view, overconfidence occurs
not because people have (erroneously) elevated assessments of their
own knowledge, but because they find experts untrustworthy. While
certainly plausible, it is important to reiterate that most respondents do
place at least some amount of trust in medical experts (Table 1). Fur-
ther, we found that overconfidence is not directly associated with de-
creased trust in scientific experts (Table 3), which we might suspect to
be the case if the two measures were truly tapping the same underlying
“distrust” construct.
Others might point out that our conceptualization of “expert dis-
counting” seems to imply a normative judgment about the influence of
non-experts in the policymaking process. We want to clarify that we,
like many others (e.g., Jasanoff, 1990; Wynne, 1992; Scheufele, 2014),
think that communication between experts and lay audiences is most
effective when both sides learn from one another. Still, we reiterate that
the outcome variables studied here point to a high degree of non-expert
policy influence. We certainly do not wish to shut ordinary people out
of the policymaking process entirely, but we do find it reasonable to
280
Social Science & Medicine 211 (2018) 274–281
consider (for example) information from celebrities to be of a generally
lower quality than that from communities of medical and scientific
professionals.
4.1. Limitations and directions for future research
A notable limitation of our study is our use of cross-sectional survey
data. This setup is useful for documenting correlational relationships
between autism awareness, overconfidence, and policy attitudes.
However, while rigorously testing our hypotheses does not necessarily
require longitudinal evidence, we cannot answer more general ques-
tions about the origins of knowledge and misinformation about autism.
We also caution that the results presented here are from a single cross-
sectional study. We welcome future attempts to replicate and extend
the results presented in this study.
Our study provides several directions for future research. First, al-
though our analysis focuses on Dunning-Kruger effects in the area of
vaccination policy, there is little reason to suspect that they are limited
only to this topic in health policy. We suspect that Dunning-Kruger
effects might be at play in influencing public attitudes on a variety of
health policies where the topic is complex and has generated conflict.
Second, future research should more closely investigate the details of
expert discounting in the policy realm. While our analysis was limited
to looking at friends/families and celebrities as non-experts, other ac-
tors – like community leaders, religious figures, and teachers – might
matter as well.
5. Conclusions
Overall, our results provide new insight into our understanding of
U.S. adults' attitudes towards vaccine policy. First, our research estab-
lishes Dunning-Kruger effects in the study of public attitudes sur-
rounding vaccination and autism. While low levels of public knowledge
about autism and misinformation about a link between vaccination and
autism have been established in prior work, our analysis is the first to
suggest that this lack of awareness can alter how citizens' view their
own knowledge, compared to that of medical experts. By showing that
individuals in our sample who are low in autism awareness are sig-
nificantly more likely to consider themselves better informed than
medical and scientific experts about the causes of autism, we add
needed complexity to our understanding of this topic.
Critically, our analysis suggests that this overconfidence could have
important policy consequences. Even after accounting for the in-
dependent effects of autism knowledge, misinformation, and several
social, political, and demographic factors, our results indicate that
overconfident individuals in our sample are less supportive of manda-
tory vaccination policy and tend to elevate the role that non-experts
should play in the policymaking process. This finding suggests that
researchers and policymakers interested in increasing support for
mandatory vaccination need to combat not only low levels of knowl-
edge and misinformation, but also the independent effect of over-
confidence.
Acknowledgements
We thank Sarah Gollust, Chris Federico, Dan Myers, Krissy Lunz-
Trujilo, Vivienne Byers, Wendy Whitman Cobb, Cindy Kam, members of
the Center for the Study for Political Psychology Proseminar, and the
anonymous reviewers for their invaluable feedback and suggestions.
Support for this research was provided by Utah Valley University's
Office of Engaged Learning.
Appendix A. Supplementary data
Supplementary data related to this article can be found at http://dx.
doi.org/10.1016/j.socscimed.2018.06.032.
M. Motta et al.
References
Armstrong, K., Berlin, M., Schwartz, S.J., Propert, K., Ubel, A.P., 2001. Barriers to in-
fluenza immunization in a low-income urban population. Am. J. Prev. Med. 20 (1),
21–25.
Baron, R.M., Kenny, D.A., 1986. The moderator–mediator variable distinction in social
psychological research: conceptual, strategic, and statistical considerations. J. Pers.
Soc. Psychol. 51 (6), 1173.
Bean, J.S., 2011. Emerging and continuing trends in vaccine opposition website content.
Vaccine 29 (10), 1874–1880.
Blank, M.J., Shaw, D., 2015. Does partisanship shape attitudes toward science and public
policy? The case for ideology and religion". Ann. Am. Acad. Polit. Soc. Sci. 658 (1),
18–35.
Brown, F.K., Kroll, S.J., Hudson, J.M., Ramsay, M., Green, J., Long, J.S., Vincent, A.C.,
Fraser, G., Sevdalis, N., 2010. Factors underlying parental decisions about combi-
nation childhood vaccinations including MMR: a systematic review. Vaccine 28 (26),
4235–4248.
Camargo Jr., K., Grant, R., 2015. Public health, science, and policy debate: being right is
not enough. J. Inf. 105 (2).
Carpiano, R.M., Fitz, N.S., 2017. Public attitudes toward child undervaccination: a ran-
domized experiment on evaluations, stigmatizing orientations, and support for po-
licies. Soc. Sci. Med. 185, 127–136.
Centers for Disease Control and Prevention, 2017. Why are Childhood Vaccines so
Important? Last updated August 18, 2017. Accessed January 3, 2018. https://www.
cdc.gov/vaccines/vac-gen/howvpd.htm.
Centers for Disease Control and Prevention, 2011. Measles - United States, 2011. MMWR
Morb. Mortal. Wkly. Rep. 2012 61 (15), 253–257.
Clarke, C.E., 2008. A question of balance: the autism-vaccine controversy in the British
and American elite press. Sci. Commun. 30 (1), 77–107.
Dixon, G.N., Clarke, C.E., 2013. Heightening uncertainty around certain science: media
coverage, false balance, and the autism-vaccine controversy. Sci. Commun. 35 (3),
358–382.
Dunning, D., 2011. 5 the dunning-kruger effect: on being ignorant of One's own ignor-
ance. Adv. Exp. Soc. Psychol. 44, 247.
Ehrlinger, J., 2008. Skill level, self-views and self-theories as sources of error in self-
assessment. Soc. Pers. Psychol. Compass 2 (1), 382–398.
Flynn, D.J., Nyhan, B., Reifler, J., 2017. The nature and origins of misperceptions: un-
derstanding false and unsupported beliefs about politics. Polit. Psychol. 38 (S1),
127–150.
Freed, L.G., Clark, J.S., Butchart, T.A., Singer, C.D., Davis, M.M., 2010. Parental vaccine
safety concerns in 2009. Pediatrics 125 (4), 654–659.
Funk, C., Kennedy, B., Hefferon, M., 2017. Vast Majority of Americans Say Benefits of
Childhood Vaccines Outweigh Risks. Pew Research Center, Washington, DC.
Gauchat, G.W., 2008. A test of three theories of anti-science attitudes. Socio. Focus 41 (4),
337–357.
Gauchat, G., 2012. Politicization of science in the public sphere: a study of public trust in
the United States, 1974 to 2010. Am. Socio. Rev. 77 (2), 167–187.
Imai, K., Keele, L., Tingley, D., Yamamoto, T., 2011. Unpacking the black box of causality:
Learning about causal mechanisms from experimental and observational studies. Am.
Polit. Sci. Rev. 105 (4), 765–789.
Jasanoff, S., 1990. The Fifth Branch: Scientists as Policy-makers. Harvard University
Press, Cambridge, MA.
Jolley, D., Douglas, K.M., 2014. The effects of anti-vaccine conspiracy theories on vac-
cination intentions. PLoS One 9 (2), e89177.
Joslyn, R.M., Sylvester, M.S., 2017. The determinants and consequences of accurate be-
liefs about childhood vaccinations. Am. Polit. Res. 1–22. http://journals.sagepub.
com/doi/abs/10.1177/1532673X17745342.
Kahan, D.M., Peters, E., Wittlin, M., Slovic, P., Ouellette, L.L., Braman, D., Mandel, G.,
2012. The polarizing impact of science literacy and numeracy on perceived climate
change risks. Nat. Clim. Change 2 (10), 732–735.
Kahan, D.M., 2015. What is the" science of science communication"? JCOM - J. Sci.
Commun. 14 (3).
281
Social Science & Medicine 211 (2018) 274–281
Kaplan, S., 2017. Vaccine critic Kennedy claims Trump tapped him to chair panel on
vaccination safety. Sci. Am. https://www.scientificamerican.com/article/vaccine-
critic-kennedy-set-to-chair-trump-panel-on-vaccination-safety1/.
Kata, A., 2012. Anti-vaccine activists, web 2.0, and the postmodern paradigm – an
overview of tactics and tropes used online by the anti-vaccination movement.
Vaccine 30 (25), 3778–3789.
Kata, A., 2010. A postmodern pandora's box: anti-vaccination misinformation on the in-
ternet. Vaccine 28 (7), 1709–1716.
Kinder, D.R., Kalmoe, N.P., 2017. Neither liberal Nor Conservative: Ideological Innocence
in the American Public. University of Chicago Press.
Kuklinski, J.H., Quirk, P.J., Jerit, J., Schwieder, D., Rich, R.F., 2000. Misinformation and
the currency of democratic citizenship. J. Polit. 62 (3), 790–816.
Kruger, J., Dunning, D., 1999. Unskilled and unaware of it: how difficulties in recognizing
one's own incompetence lead to inflated self-assessments. J. Pers. Soc. Psychol. 77
(6), 1121.
Larson, J.H., Cooper, Z.L., Eskola, J., Katz, L.S., Ratzan, S., 2011. Addressing the vaccine
confidence gap. Lancet 378 (9790), 526–535.
Larson, J.H., Jarrett, C., Eckersberger, E., Smith, M.D., Paterson, P., 2014. Understanding
vaccine hesitancy around vaccines and vaccinations from a global perspective: a
systematic review of published literature, 2007-2012. Vaccine 32 (19), 2150–2159.
Lewandowsky, S., Ecker, U.K., Cook, J., 2017. Beyond misinformation: understanding and
coping with the “Post-Truth” era. J. Appl. Res. Mem. Cognit. 6 (4), 353–369.
Lupia, A., 2015. Uninformed: Why people know so little about politics and what we can
do about it. Oxford University Press.
Mason, L., 2018. Ideologues without issues: the polarizing consequences of ideological
identities. Publ. Opin. Q. 82 (S1), 280–301.
Miller, J.M., Saunders, K.L., Farhart, C.E., 2016. Conspiracy endorsement as motivated
reasoning: the moderating roles of political knowledge and trust. Am. J. Polit. Sci. 60
(4), 824–844.
Motta, M., 2018. The dynamics and political implications of anti-intellectualism in the
United States. Am. Polit. Res. 46 (3), 465–498.
Nelson, K.B., Bauman, M.L., 2003. Thimerosal and autism? Pediatrics 111 (3), 674–679.
Nyhan, B., Reifler, J., Richey, S., Freed, G.L., 2014. Effective messages in vaccine pro-
motion: a randomized trial. Pediatrics 133 (4), e835–e842.
Nyhan, B., Reifler, J., 2015. Does correcting myths about the flu vaccine work? An ex-
perimental evaluation of the effects of corrective information. Vaccine 33 (3),
459–464.
Pelcic, G., Karacic, S., Mikirtichan, L.G., Kubar, I.O., Leavitt, J.F., Tai, C.M., Morishita, N.,
Vuletic, S., Tomasevic, L., 2016. Religious exception for vaccination or religious
excuses for avoiding vaccination. Croat. Med. J. 57 (5), 516–521.
Poltorak, M., Leach, M., Fairhead, J., Cassell, J., 2005. ‘MMR talk’and vaccination
choices: an ethnographic study in Brighton. Soc. Sci. Med. 61 (3), 709–719.
Prislin, R., Dyer, A.J., Blakely, H.C., Johnson, D.C., 1998. Immunization status and so-
ciodemographic characteristics: the mediating role of beliefs, attitudes, and perceived
control. Am. J. Publ. Health 88 (12), 1821–1826.
Rabinowitz, M., Latella, L., Stern, C., Jost, T.J., 2016. Beliefs about childhood vaccina-
tions in the United States: political ideology, false consensus, and the illusion of
uniqueness. PLoS One 11 (7), 1–19.
Ruijs, W.L., Hautvast, J.L., van der Velden, K., de Vos, S., Knippenberg, H., Hulscher,
M.E., 2011. Religious subgroups influencing vaccination coverage in the Dutch Bible
belt: an Ecological study. BMC Publ. Health 11, 102.
Ruijs, W.L., Hautvast, J.L., Kerrar, S., van der Velden, K., Hulscher, M.E., 2013. The role
of religious leaders in promoting acceptance of vaccination within a minority group:
a qualitative study. BMC Publ. Health 13, 511.
Scheufele, D.A., 2014. Science communication as political communication. Proc. Natl.
Acad. Sci. Unit. States Am. 111 (Suppl. 4), 13585–13592.
Shelton, R.C., Snavely, A.C., De Jesus, M., Othus, M.D., Allen, J.D., 2013. HPV vaccine
decision-making and acceptance: does religion play a role? J. Relig. Health 52 (4),
1120–1130.
Wynne, B., 1992. Sheep farming after Chernobyl: a case study in communicating scientific
information. In: When Science Meets the Public, pp. 43–68.