Gaining trust as well as respect in communicating to

motivated audiences about science topics

Susan T. Fiske
1
and Cydney Dupree
Psychology and Public Affairs, Princeton University, Princeton, NJ 08544
Edited by Baruch Fischhoff, Carnegie Mellon University, Pittsburgh, PA, and accepted by the Editorial Board April 3, 2014 (received for review October
31, 2013)
Expertise is a prerequisite for communicator credibility, entailing
the knowledge and ability to be accurate. Trust also is essential to
communicator credibility. Audiences view trustworthiness as the
motivation to be truthful. Identifying whom to trust follows
systematic principles. People decide quickly anothers apparent
intent: Who is friend or foe, on their side or not, or a cooperator
or competitor. Those seemingly on their side are deemed warm
(friendly, trustworthy). People then decide whether the other is
competent to enact those intents. Perception of scientists, like
other social perceptions, involves inferring both their apparent
intent (warmth) and capability (competence). To illustrate, we
polled adults online about typical American jobs, rated as Ameri-
can society views them, on warmth and competence dimensions,
as well as relevant emotions. Ambivalently perceived high-compe-
tence but low-warmth, envied professions included lawyers, chief
executive officers, engineers, accountants, scientists, and research-
ers. Being seen as competent but cold might not seem problematic
until one recalls that communicator credibility requires not just sta-
tus and expertise but also trustworthiness (warmth). Other research
indicates the risk from being enviable. Turning to a case study of
scientific communication, another online sample of adults described
public attitudes toward climate scientists specifically. Although dis-
trust is low, the apparent motive to gain research money is dis-
trusted. The literature on climate science communicators agrees
that the public trusts impartiality, not persuasive agendas. Overall,
communicator credibility needs to address both expertise and
trustworthiness. Scientists have earned audiences respect, but not
necessarily their trust. Discussing, teaching, and sharing informa-
tion can earn trust to show scientists trustworthy intentions.
public images
|
scientist stereotypes
L
ong ago, Aristotle knew that communication is not just about
logic and knowledge, but also about emotions and values
[differentiating among logical argument (logos), demonstrating
character (ethos), and evoking emotion (pistis) (1) and refs. 2 and 3].
As science communicators, we are interested in public belief forma-
tion, so we can benefit from understanding the complementary roles
of audience respect for scientists expertise and trust in our character.
Scientists as communicators have earned audiences respect,
but not necessarily their trust, as the evidence will suggest. This
Perspectives article begins with climate science as an example of
potential misunderstanding between scientists and their audi-
ence, and then examines the science of communicator credibility
more generally, showing that trust is a critical factor. Next the
article describes how people decide which groups to trust, and
how scientists rate on that dimension. Finally, returning to cli-
mate scientists as a case study, the public has some specific
opinions about how they are and are not trustworthy. Conclusions
suggest that scientific communication can be more effective by
drawing on both dimensions of communicator credibility.
Public Beliefs and Affect
Consider climate change communications as a case study in
science communication. Public responses to climate change
suggest both good news and bad news for scientists trying to
convey the best available evidence. The discouraging news is that
scientists and the public are isolated from each other (4). They
inhabit distinct information environments; for instance, sheer at-
tention to political news versus science and ecological news predicts
disparate climate risk perceptions and policy support (5). Despite
scientific consensus on climate change trends, the public is of two
minds, with much nay saying and extreme skepticism, but at the
same time being worried and having alarmist imagery (6). How
can science communicators reconcile the gap between public
ambivalence and scientific consensus? Two social psychological
factors come into play here.
First, scientists may misunderstand the sources of lay beliefs:
People are no idiots. The publics issue with science is not nec-
essarily ignorance (7). The public increasingly knows more than
before about climate changes causes (8). Psychology under-
graduates at least can judge both science and nonscience arguments
by the amount and reliability of their evidence (9). Different lay
people hold different models of science, some more classical
(seeking a single, true answer) versus others more Kuhnian
(acknowledging multiple answers to negotiate and debate,
accepting scientific uncertainty) (10). Hence, potential divides
between scientists and the public are not merely about sheer
knowledge in any simple way.
The second, often-neglected factor is the other side of atti-
tudes. Attitudes are evaluations that include both cognition
(beliefs) and affect (feelings, emotions) (11). Acting on attitudes
involves both cognitive capacity and motivation (12). Attitudes
show an intrinsic pressure for consistency between cognition and
affect (11), so for most attitudes, both are relevant. When atti-
tudes do tilt toward emphasizing either cognition or affect,
persuasion is more effective when it matches the type of attitude
(13). In the domain of climate change, for example, affect and
values together motivate climate cognition (4, 14). If public
attitudes have two sidesbeliefs and affectwhat is their re-
spective role in scientific communication?
Communicator Credibility
Science communicators try to persuade the public that they are
honest brokers of scientific evidence, that is, that they are credible.
In an attitudes research literature spanning decades, communi-
cator credibility demonstrably has two components. Expertise
is only one crucial prerequisite for communicator credibility (15).
Perceived expertise entails the knowledge and ability to be accurate.
This paper results from the Arthur M. Sackler Colloquiumof the National Academy of Sciences,
The Science of Science Communication II, held September 2325, 2013, at the National Acad-
emy of Sciences in Washington, DC. The complete program and video recordings of most pre-
sentations are available on the NAS website at www.nasonline.org/science-communication-II.
Author contributions: S.T.F. designed research; C.D. performed research; C.D. analyzed
data; S.T.F. reviewed persuasion literature; C.D. reviewed communicating climate change;
and S.T.F. and C.D. wrote the paper.
The authors declare no conflict of interest.
This article is a PNAS Direct Submission. B.F. is a guest editor invited by the Editorial Board.
1
To whom correspondence should be addressed. Email: [email protected].
This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.
1073/pnas.1317505111/-/DCSupplemental.
www.pnas.org/cgi/doi/10.1073/pnas.1317505111 PNAS | September 16, 2014 | vol. 111 | suppl. 4 | 1359313597
Audiences who are operating on automatic (not thinking very
hard) are instantly inclined to agree with experts. However, ex-
pertise can also trigger close scrutiny, if the audience is moti-
vated and has the opportunity to think hard (16). Scientists
presumably can expect their expertise to have these types of
effects, depending on the audience.
The other feature of communicator credibility is trust, also
according to decades of attitudes research (16). Audiences judge
the communicators apparent trustworthiness as the communicators
inferred motivation to be truthful. Trust most often makes audi-
ences automatically believe in the messages validity (16). We know
less about whether scientists are usually accorded trust, making this
an open question that our illustrative data will begin to evaluate.
The classic communicator credibility literature provides useful
information regarding whom audiences choose to trust. First,
people trust similar others, that is, people they categorize as
being in their own group (17). People assume people like us
share their goals and values, so they are motivated to support
them as being on their own side. When scientists convey a par-
ticular political orientation, for example, they alienate people
whose politics disagree. Given liberal political tendencies in
universities, conveying those politics is likely to undermine some
audiences feelings of similarity to a scientist communicator.
Second, audiences also trust the sincerity of persistent mi-
norities (11), which may help explain the impact of vocal but
minority viewpointssuch as extreme climate change deniers
but that is a topic for another article. Primarily, people focus
their trust on their own group; trust being a socially adaptive
strategy for getting along in ones group (18, 19).
Identifying Whom to Trust
To fundamentally social beings, trust always matters. Peoples
social interactions are ruled by a known set of five repeatedly
cited and supported core motives that together sketch some main
drivers of social behavior, including trust (18, 19). Most central
of the five motives is the motivation for belonging to a stable set
of face-to-face relationships (at the level of dyads and small
groups) (20). Arguably, belonging has survival value; humans
health and wellbeing suffers under isolation, and stable rela-
tionships protect against mortality (21).
From the fundamental motive for belonging follow other core
motives, two more oriented to cognition and two more oriented
to affect (18, 19): On the cognitive side, to survive and thrive
within a group, people need to acquire socially shared un-
derstanding. They also need some sense of control; that is, they
need to see some contingency between what they do and what to
expect in return. Communicator expertise, for example, con-
tributes to these two cognitively oriented motives by providing
valid information about consensus beliefs (socially shared un-
derstanding) and about social norms (peoples reports about
social rules, such as littering).
On the affective side are two other core motives: People also
adapt better to group life when they have self-esteem adequate
to maintain participation in the group and when their default
motive is to trust their own group members. Reasonable self-
regard and interpersonal trust lay the groundwork for more effi-
cient and effective interpersonal interactions (18, 19). Together,
these core motives suggest that audience priorities are not only
cognitive, but also affective.
This last core social motive, identifying whom to trust, follows
systematic principles (22, 23). People decide quickly anothers
apparent intent: Who is friend or foe, on their side or not, or
a cooperator or competitor. If seemingly on their side, people
deem the other individual or group as warm (friendly and
trustworthy). Upon determining intent for good or ill, people
then decide whether the other is competent to act on those
intents. These rapid perceptions express a warmth competence
space [see Fig. 1 for American data on societal groups (23)].
Within these warmth-by-competence cultural maps of groups
public images (22), the societal ingroups, reference groups, and
their allies typically appear in the top right cluster. These groups
are seen as high on both dimensions, perceived warmth and
perceived competence: middle class, Christians, whites (and in
other samples, Americans). Affectively, people report feelings of
admiration and pride toward these groups.
In the opposite end of this diagonal, judged as low on both the
warm and competence dimensions, appear poor people, homeless,
welfare recipients (and in other samples, drug addicts, immi-
grants). Affectively, people report disgust and contempt toward
these groups. (In the middle cluster are groups whose middling
average ratings do not differentiate among these respondents
expressing ignorance, indifference, or opposing extremes that
cancel each other in the aggregate. See refs. 24 and 25 for examples
of decomposing this apparent indifference.)
Compared with existing research on stereotypes, the in-
novative feature of our research using this warmth-by-compe-
tence map (23, 24) is the ambivalent diagonal, groups judged as
high on one dimension but low on the other. Groups judged as
nice but incompetent include older and mentally or physically
disabled people (group names in Fig. 1 are respondent pro-
vided). Affectively, people report pity and sympathy toward
them. Groups seen as competent but cold include rich people,
men, Jews, Asians, and to some extent, professionals. People
report envy and jealousy toward them.
Our program of research has gathered warmth-by-competence
maps sampled across place, time, and social entities. This space
usefully describes a variety of human social perceptions: over US
samples, both convenience [students, adults (23)] and repre-
sentative [adults (24)]; over dozens of countries, using their own
social groups (26, 27); over time [Italian Fascists (28); American
students since 1933 (29)]; over levels of analysis, from individuals
(30) to subtypes of women and men (31), immigrants (32), gay
men (25), and African Americans (33); and even over other
intent-having entities, such as animal species* and corporations
(34, 35). These warmth-by-competence data suggest the bold
conclusion that the dimensions are universally descriptive of
group images.
Competence
W
a
r
m
t
h
Homeless
Asians
Blacks
Blue-collar
Christians
Disabled Educated
Elderly
Gay men
Hispanics
Poor
Jews
Men
Rich
Muslims
Native Americans
Professionals
Retarded
Students
Women
Young
Whites
Middle
class
Welfare
recipients
High
Low
Low
High
Fig. 1. Five-cluster solution describing American societal groups. Adults
completed a survey rating previously nominated common groups each on
their public images of being warm and trustworthy, as well as competent
and capable. Group mean ratings on the two dimensions are subjected to
statistical cluster analysis (23, 24). Note that these data map reported images
of social groups, not the accuracy of those images. (Reprinted with permission
from Ref. 23.)
*Sevillano V, Fiske ST, Perceived Dimensions of Animals. Society for Personality and Social
Psychology Annual Meeting, January 1719, 2013, New Orleans, LA.
13594 | www.pnas.org/cgi/doi/10.1073/pnas.1317505111 Fiske and Dupree
What About Scientists as Communicators?
Perception of scientists, like other social perception, presumably
involves inferring intent (warmth) as well as capability (compe-
tence). In the context of this Perspectives article, these percep-
tions may help to understand how the public responds to science
communicators. Consistent with our previous methods, we tested
this idea in two phases (see SI for details). We first asked online
samples of adult volunteers to list some typical American jobs,
and then we selected the most commonly mentioned 42 jobs,
which included scientists, researchers, professors, and teachers.
We then polled a new online adult sample about these jobs,
asking for warmth and competence ratings according to how
American society views them. This technique yields peoples
reports of the public images of groups. It also reduces social
desirability biases that would occur in reporting ones own in-
dividual stereotypes. The data thus generate lay theories of the
cultures shared images. Participants rated the various job
holders standing on American societys warmth and competence
perceptions, as well as reporting Americans typical emotional
reactions to the job-holder groups. The warmth competence
data then underwent cluster analysis, generating the map shown
in Fig. 2.
In the top right of Fig. 2the high-warmth, high-competence,
pride cornerappear professionals who are also caring: nurses,
teachers, and doctors. They are rated as having the image of
being warm and trustworthy and capable and competent. They
are both trusted (warm) and respected (competent). The most-
reported emotions are admiration and pride.
In the bottom leftlow-warmth, low-competence, disgust
cornerappear prostitutes, garbage collectors, dishwashers, and
fast-food workers. They are rated low on both dimensions and
higher on disgust and contempt than other groups are. That is,
they are neither trusted nor respected.
The remaining corners of the space are ambivalent in the
sense of having mixed images, scoring high on one dimension
and low on the other. The ambivalent, high-warmth, low-com-
petence, pity quadrant (upper left in Fig. 2) is empty in this
dataset, perhaps being reserved for the unemployed (who land
there in other datasets), because the current data points all in-
volve groups that have a job.
The fourth corner lists the ambivalently perceived high-com-
petence, low-warmth, envied professions: lawyers, chief exec-
utive officers, engineers, accountants, scientists, and researchers.
They earn respect but not trust. Being seen as competent but
cold might not seem problematic until one recalls that commu-
nicator credibility requires not just status and expertise (com-
petence) but also trustworthiness (warmth). People report envy
and jealousy toward groups in this space. These are mixed
emotions that include both admiration and resentment (23, 24).
Science communicators arguably need to know about this pos-
sible type of response to them.
What is more, other research indicates the risk inherent in
being enviable. People rated other types of envied outgroup
members (e.g., rich people). Participants reported feeling slightly
good and not so bad, when mundane bad events (splashed by a
taxi, sitting on chewing gum) happen to these groups. This mali-
cious glee at others misfortunes is known as Schadenfreude. The
same emotion also appears in electromyography of the cheeks
smile muscles (zygomaticus major) when bad things happen to
envied people. All other quadrants (ingroup, pitied groups, dis-
gusting outgroups) elicit more happiness and smiling to good
events rather than bad ones (36).
In sports rivalries, likewise, Schadenfreude toward envied
group members is reflected in neural reward centers, also cor-
related with self-reports of aggression (taunting, food throwing,
fistfights) toward envied rival teams fans (37). Being enviable
may entail respect, but also entails some degree of dislike and
Fig. 2. Warmthcompetence ratings of commonly mentioned jobs. Four-cluster solution describes American job holders. Adults completed an online survey
rating previously nominated common jobs on their public images of being warm and trustworthy, as well as competent and capable. Jobs mean ratings on
the two dimensions are subjected to statistical cluster analysis (see SI text). Note that these data map reported images of job groups, not the accuracy of those
images. Although not the focus here, the middle cluster averages are relatively neutral on both warmth and competence, as well as emotions (see SI text).
Fiske and Dupree PNAS | September 16, 2014 | vol. 111 | suppl. 4 | 13595
distrust. Therefore, even if scientists are respected as competent,
they may not be trusted as warm. Recall that the warmth/trust-
worthiness judgment assesses the others perceived intent for
good or ill (friend or foe). Scientists, in this view, may seem
not so warm. Their intent is not necessarily trusted, maybe
even resented.
We began this illustrative research to exemplify how audiences
may view scientists as communicators, not just as scientists.
Science communicators could be less at risk for the envy syn-
drome than pure scientists and researchers. Note in Fig. 2 that
professors and teachers images appear warmer than those of
scientists and researchers. Scientists whose job involves teaching
and communicating may seem warmer and more trustworthy,
seeming to show worthy intentions. Audiences as spontaneous
intent detectors may attend to the apparently trustworthy goals
of teaching (professors) and caring (doctors).
Climate Scientists Presumed Agendas
Focusing in on our case study of scientific communication, we
asked another online sample of adults to describe public atti-
tudes toward climate scientists specifically, to provide a clearer
picture of the publics apparently mixed feelings (see SI text). On
a seven-item scale of distrust, climate scientists averaged a dis-
trust mean of 2.16 (below the midpoint on a five-point scale).
For the most part, distrust of climate scientists runs low, but not
at the floor of the scale. And of course, responses vary, with
some more distrustful than others.
The slight distrust toward climate scientists might seem to
contradict the earlier data showing that scientists seem less warm
than many other job holders. However, the current result is open
to different interpretations. First, this result reflects rating cli-
mate scientists per se, who might be viewed more positively than
generic scientists. Second, climate scientists are rated here in
isolation from judging other groups, and absolute ratings (here)
may differ from relative ones (the earlier data). Third, this is
a scale of distrust, not trust (as before), and participants tend to
avoid rating other people negatively as individuals (19). The
more individuated (specific) the person, the more reluctant
people are to express negativity. Finally, the seven items on
this scale differ from the two items on the earlier scale (SI text).
So strictly speaking, this result says that climate scientists by
themselves are distrusted less than the distrust-scale midpoint,
whereas the earlier result says that generic scientists are trusted
about at the warmth/trustworthiness-scale midpoint. People are
inclined to trust other people on average, usually scoring above
a positive scales midpoint (18, 19), so neither result is cause
for celebration.
In these ratings of climate scientists (all items being derived
from pilot work on scientists alleged motives), the seven survey
items reflecting distrust include alleged motives to lie with sta-
tistics, complicate a simple story, show superiority, gain research
money, pursue a liberal agenda, provoke the public, and hurt big
corporations. Among these individual distrust items, the only
one averaging above the midpointgain research money (M =
3.58)stands out as the Achilles heel: the publics clearest
reported reason to distrust climate scientists motives. Never-
theless, none of the distrust ratings are at the floor.
Turning to the scale of reasons to trust, climate scientists seem
to fare well, at least among this online sample (which is slightly
better educated, slightly more liberal than the US population).
On a three-item, five-point scale of trust, climate scientists av-
eraged 4.35 on motives to educate the public, save humanity, and
save the environment. Clearly, these reasons for trust suggest a
constructive approach to the public, balancing expertise (com-
petence) with trustworthiness (warmth), together facilitating
communicator credibility.
The literature on climate science communicators supports
these recommendations to focus on trust-inducing approaches
such as education, humanity, and the environment. The public
tends to trusts impartiality, rather than persuasive agendas (38).
Although scientific communicators may stress persuasion, de-
liberation would be better (39). Communicating uncertainty is
essential to building credibility (40, 41), and trust best predicts
attention to scientific experts (42).
The road to communicating climate science starts with some
advantages. As noted, the public does have some knowledge.
Regarding climate change, the current public mostly understands
the changes, understands humanitys role, and advocates changing
emissions standards (43, 44). Climate science communicators
have effectively conveyed much evidence, which should encourage
their continuing to educate and communicate. Our data additionally
suggest conveying trustworthy motives as well as expert information.
Conclusions
Science communication, like other communication, needs to
convey communicator warmth/trustworthiness as well as com-
petence/expertise, to be credible. Peoples attitudes combine
cognition and affect, that is, beliefs and values. Hence, communi-
cator credibility needs to address both expertise and trustworthiness.
Our illustrative data are limited by not being a representative
sample. Nevertheless, they suggest that scientists may have the
respect, but not necessarily the trust of the public. This gap can
be filled, we suggest, by showing concern for humanity and the
environment. Rather than persuading, we and our audiences are
better served by discussing, teaching, and sharing information, to
convey trustworthy intentions.
ACKNOWLEDGMENTS. The illustrative research was supported by the
Princeton Institute for International and Regional Studies Communicating
Uncertainty Project. C.D. was supported by a National Science Graduate
Fellowship and the Princeton Joint Degree Program in Psychology and
Social Policy.
1. Aristotle (1959) Ars Rhetorica, ed Ross WD. (Oxford Classical Texts, New York).
2. Lamb M, Lane M Climate Change and Non-Ideal Justice, eds Heyward C, Roser D
(Oxford Univ Press, New York).
3. Allen DS (2004) Talking to Strangers: Anxieties of Citizenship Since Brown v. Board of
Education (Univ of Chicago Press, Chicago).
4. Weber EU, Stern PC (2011) Public understanding of climate change in the United
States. Am Psychol 66(4):315328.
5. Zhao X, Leiserowitz AA, Maibach EW, Roser-Renouf C (2011) Attention to science/
environment news positively predicts and attention to political news negatively predicts
global warming risk perceptions and policy support. J Commun 61(4):713731.
6. Smith N, Leiserowitz A (2012) The rise of global warming skepticism: Exploring af-
fective image associations in the United States over time. Risk Anal 32(6):10211032.
7. Kellstedt PM, Zahran S, Vedlitz A (2008) Personal efficacy, the information environ-
ment, and attitudes toward global warming and climate change in the United States.
Risk Anal 28(1):113126.
8. Reynolds TW, Bostrom A, Read D, Morgan MG (2010) Now what do people know
about global climate change? Survey studies of educated laypeople. Risk Anal 30(10):
15201538.
9. Corner A, Hahn U (2009) Evaluating science arguments: Evidence, uncertainty, and
argument strength. J Exp Psychol Appl 15(3):199212.
10. Rabinovich A, Morton TA (2012) Unquestioned answers or unanswered questions:
Beliefs about science guide responses to uncertainty in climate change risk commu-
nication. Risk Anal 32(6):9921002.
11. Eagly AH, Chaiken S (1998) Handbook of Social Psychology, eds Gilbert DT, Fiske ST,
Lindzey G (Oxford Univ Press, New York), pp 269322.
12. Fazio RH (1990) Multiple processes by which attitudes guide behavior: The MODE
model as an integrative framework. Adv Exp Soc Psychol 23:75110.
13. Edwards K (1990) The interplay of affect and cognition in attitude formation and
change. J Pers Soc Psychol 59(2):202216.
14. Swim JK, et al. (2011) Psychologys contributions to understanding and addressing
global climate change. Am Psychol 66(4):241250.
15. Petty RE, Cacioppo JT (1981) Attitudes and Persuasion: Classic and Contemporary
approaches (WC Brown, Dubuque, IA).
16. Petty RE, Wegener DT (1998) Handbook of Social Psychology, eds Gilbert DT, Fiske ST,
Lindzey G (Oxford Univ Press, New York), pp 323390.
17. Brewer MB, Brown RJ (1998) Handbook of Social Psychology, eds Gilbert DT, Fiske ST,
Lindzey G (Oxford Univ Press, New York), pp 554595.
13596 | www.pnas.org/cgi/doi/10.1073/pnas.1317505111 Fiske and Dupree
18. Fiske ST (2002) Motivated Social Perception: The Ontario Symposium, eds Spencer SJ,
Fein S, Zanna MP, Olson J (Erlbaum, Mahwah, NJ), Vol 9.
19. Fiske ST (2013) Social Beings (Wiley, New York), 4th Ed.
20. Baumeister RF, Leary MR (1995) The need to belong: Desire for interpersonal at-
tachments as a fundamental human motivation. Psychol Bull 117(3):497529.
21. House JS, Landis KR, Umberson D (1988) Social relationships and health. Science
241(4865):540545.
22. Fiske ST, Cuddy AJC, Glick P (2007) Universal dimensions of social cognition: Warmth
and competence. Trends Cogn Sci 11(2):7783.
23. Fiske ST, Cuddy AJ, Glick P, Xu J (2002) A model of (often mixed) stereotype content:
Competence and warmth respectively follow from perceived status and competition.
J Pers Soc Psychol 82(6):878902.
24. Cuddy AJC, Fiske ST, Glick P (2007) The BIAS map: Behaviors from intergroup affect
and stereotypes. J Pers Soc Psychol 92(4):631648.
25. Clausell E, Fiske ST (2005) When do the parts add up to the whole? Ambivalent ste-
reotype content for gay male subgroups. Soc Cogn 23(2):157176.
26. Cuddy AJC, et al. (2009) Stereotype content model across cultures: Towards universal
similarities and some differences. Br J Soc Psychol 48(Pt 1):133.
27. Durante F, et al. (2013) Nations income inequality predicts ambivalence in stereotype
content: How societies mind the gap. Br J Soc Psychol 52(4):726746.
28. Durante F, Volpato C, Fiske ST (2010) Using the Stereotype Content Model to examine
group depictions in Fascism: An archival approach. Eur J Soc Psychol 40(3):465483.
29. Bergsieker HB, Leslie LM, Constantine VS, Fiske ST (2012) Stereotyping by omission:
Eliminate the negative, accentuate the positive. J Pers Soc Psychol 102(6):12141238.
30. Russell AM, Fiske ST (2008) Its all relative: Social position and interpersonal percep-
tion. Eur J Soc Psychol 38(7):11931201.
31. Eckes T (2002) Paternalistic and envious gender stereotypes: Testing predictions from
the stereotype content model. Sex Roles 47(3-4):99114.
32. Lee TL, Fiske ST (2006) Not an outgroup, but not yet an ingroup: Immigrants in the
stereotype content model. Int J Intercult Relat 30(6):751768.
33. Fiske ST, Bergsieker HB, Russell AM, Williams L (2009) Images of Black Americans:
Then, them, and now, Obama! Du Bois Rev 6(1):83101.
34. Kervyn N, Fiske ST, Malone C (2012) Brands as intentional agents framework: Warmth
and competence map brand perception. J Consum Psychol 22(2):166176.
35. Malone C, Fiske ST (2013) The HUMAN Brand: How We Relate to People, Products,
and Companies (Wiley/Jossey Bass, San Francisco).
36. Cikara M, Fiske ST (2012) Stereotypes and Schadenfreude: Behavioral and
physiological markers of pleasure at others misfortunes. Soc Psychol Pers Sci
3(1):6371.
37. Cikara M, Botvinick MM, Fiske ST (2011) Us versus them: Social identity shapes neural
responses to intergroup competition and harm. Psychol Sci 22(3):306313.
38. Rabinovich A, Morton TA, Birney ME (2012) Communicating climate science: The role
of perceived communicators motives. J Environ Psychol 32(1):1118.
39. Johnson BB (2012) Climate change communication: A provocative inquiry into mo-
tives, meanings, and means. Risk Anal 32(6):973991.
40. Jensen JD (2008) Scientific uncertainty in news coverage of cancer research: Effects of
hedging on scientists and journalists credibility. Hum Commun Res 34(3):347369.
41. Karmarkar U, Tormala ZL (2010) Believe me, I have no idea what Im talking about:
The effects of source certainty on consumer involvement and persuasion. J Consum
Res 36(6):10331049.
42. Bleich S, Blendon R, Adams A (2007) Trust in scientific experts on obesity: Implications
for awareness and behavior change. Obesity (Silver Spring) 15(8):21452156.
43. Krosnick JA, MacInnis B (2013) Does the American public support legislation to reduce
greenhouse gas emissions? Daedalus 142(1):2639.
44. Terwel BW, Harinck F, Ellemers N, Daamen DD (2009) Competence-based and in-
tegrity-based trust as predictors of acceptance of carbon dioxide capture and storage
(CCS). Risk Anal 29(8):11291140.
Fiske and Dupree PNAS | September 16, 2014 | vol. 111 | suppl. 4 | 13597
Supporting Information
Fiske and Dupree 10.1073/pnas.1317505111
The present study examines perceptions of people holding certain
jobs and occupations. Our goal was to examine the group
structure of job stereotypes to gain insight into the perceptions
surrounding members of the scientific community. Such insight
could assist us in gaining traction on how scientists can more
effectively communicate science.
Review of Stereotype Content Model Hypotheses.
i) Perceptions of jobs will vary, such that certain jobs will clus-
ter and occupy distinct locations in the stereotype content
model (SCM) space. Competence and warmth will differen-
tiate job perceptions.
ii) Similar to previous SCM research applied to other social
groups (15), most job groups will receive ambivalent stereo-
types of warmth and competence, as defined by low ratings
in one dimension accompanied by high ratings on the other
dimension.
iii) Social structure will influence the stereotypes applied to out-
group jobs; more competence ratings will be ascribed to
those outgroup members who hold jobs perceived to be high
status and powerful, and more warmth ratings will be as-
cribed to those job holders who are not seen as competition.
Preliminary Study: Selecting Job and Occupation Groups. Methods.
Participants. The preliminary study used Amazon.coms Mechan-
ical Turk online marketplace (www.mturk.com/mturk/welcome);
48 participants (32 men; mean age = 30.08 y, SD = 10.77) were
recruited to complete a brief questionnaire about various social
groups at standard MTurk rates. The Princeton University In-
stitutional Review Board approved both studies in this research,
including informed consent.
Questionnaire and procedure. Participants encountered an open-
answer questionnaire given the following instructions: Off the
top of your head, what various groups of people do you think
todays society categorizes based on occupation or job? Par-
ticipants saw 16 blank lines and were asked to provide at least 8
and no more than 16 groups. On the following page, participants
saw similar instructions, except here they were asked to consider
and list jobs that are seen as low status in todays society. This
question has been included in previous SCM studies during
group selection phases to allow for the emergence of groups
representing pure antipathy (2). On the third page, participants
listed up to four jobs that they currently or recently held. This
allowed for the inclusion of the respondents own ingroups. The
questionnaire ended with basic demographics questions.
Results and discussion. Across the three questions (general groups,
low-status groups, own groups), participants listed a total of 372
jobs. A coder consolidated those jobs that were viewed as similar
in function. The most frequently listed job was mentioned 50
times. We selected groups to include in the following survey if
they were listed more than 8 times (i.e., by at least 15% of
participants). This criterion yielded a set of 42 jobs to include in
the main survey. The most frequently mentioned job in the set
(janitor) was listed 50 times, and the least frequently mentioned
jobs in this set (actor, bus driver, researcher, food service worker,
professor, office clerk, prostitute) each were mentioned 8 times.
Four jobs in the final set relate to our primary interest in per-
ceptions of science communicators: professors, researchers, and
scientists.
Survey: Rating Warmth and Competence. Methods.
Participants. Participants who had not participated in the pre-
liminary study were recruited to complete the survey in the same
manner as before; 116 participants (65 men; mean age = 37.91 y,
SD = 13.57) completed the main survey for standard MTurk
pay rates.
Questionnaire and procedure. To prevent fatigue, each participant
was randomly selected to rate one-third of the total groups; this
yielded three participant samples, each of which completed one of
three surveys. Each version included 14 randomly selected jobs
that participants rated on items measuring warmth, competence,
status, and competition. To ensure the random presentation of
our particular jobs of interest, each condition included one of the
main science-related jobs most relevant: scientist, professor, or
researcher.
Participants read that we were interested in the way different
occupations are viewed by others in American society. They then
rated each of their 14 target groups on perceived warmth and
competence items using a five-point scale (1 = not at all; 5 =
extremely). To do so, participants were asked, To what extent
do most Americans view holders of this job as competent?
Warm? Capable? Trustworthy?
To judge perceived status, participants rated the prestige and
economic success associated with each group. For competition,
participants responded to questions anticipating the impact on
their own group if each job group received special tax breaks, held
incompatible values and goals, or gained additional resources.
Participants were also asked about typical emotional responses
to each group: pity, envy, contempt, pride, sympathy, admiration,
disgust, and jealousy.
As in the preliminary study, all participants listed up to four
jobs that they currently or recently held. Finally, they completed
basic demographics forms before debrief and the online pay-
ment process.
Results. We tested the differentiation of warmth and competence
among job stereotypes by calculating the mean ratings applied to
each job group. After averaging each job groups item rating
across participants, scales of warmth, competence, competition,
and status were created by averaging the group means of the
relevant items (see Table S1 for scale items). The job then be-
comes the unit of analysis.
Perceived warmth and competence differentiate job groups. Drawing
from the two-step methodology used in previous SCM research
(2), cluster analyses tested job group differentiation on warmth
and competence scales. First, Wards method of hierarchical
cluster analysis (minimizing within-cluster variance and maxi-
mizing between-cluster variance) identified a plausible number
of clusters according to agglomeration statistics, which supported
a four-cluster solution. Next, the k-means solutions indicated
which groups belonged in each cluster (Fig. 2).
Some stereotypes include mixed competence and warmth. Although
the majority of stereotypes are typically ambivalent (high on one
dimension and low on the other), in this dataset, only the high-
competence, low-warmth (envy) ambivalent cluster appeared.
No discernable pity cluster (high warmth, low competence)
appeared. As noted in the main text (What About Scientists as
Communicators?), however, unemployed people (elders, chil-
dren, people with disabilities) often inhabit the space but were
missing, due to the focus on job holders.
Warmth and competence ratings have social structure correlates. Here
the SCM hypothesizes that status positively predicts competence,
and competition negatively predicts warmth. Testing this pre-
Fiske and Dupree www.pnas.org/cgi/content/short/1317505111 1 of 2
diction requires computing correlations between stereotype
dimensions (warmth, competence) and social structure dimen-
sions (status, competition) in a group level analysis. As predicted,
competence and status correlate significantly (r = 0.91, df = 40,
P < 0.0001). Also as predicted, warmth and competence were
negatively related, but only marginally (r = 0.26, df = 40,
P = 0.09). Finally we examined other ones, predicting that
statuswarmth and competitioncompetence correlations would
not reach significance. For the nonpredicted structuretrait re-
lationships, the statuswarmth prediction was supported, in that
status did not significantly correlate with warmth (r = 0.25, df = 40,
P = 0.11). Unexpectedly, competition and competence did sig-
nificantly relate (r = 0.39, df = 40 P < 0.05). This could be be-
cause of the groups tested here: All successful job holders show
some level of competition, especially given the current struggle
with the economy.
Distinct emotions target distinct quadrants. Each groups rating on
four emotions combined two emotion scales, as in previous work
(6, 7). Envy + jealousy were predicted to be highest for the high-
competence, low-warmth cluster; disgust + contempt were pre-
dicted to be highest for the lowlow cluster, and pride + admi-
ration were predicted to be highest for the highhigh cluster.
Because the dataset produced no high-warmth, low-compe-
tence cluster, pity + sympathy had no predictions. Also, the
middle cluster was not predicted to show any distinct emo-
tions. Note that its ratings as neutral on warmth and compe-
tence could result from true indifference or from polarized
ratings, whereby participant opinions differed sharply but aver-
aged to neutral.
Between-cluster contrast analyses examined each emotion
separately (i.e., by column; Table S1) and found, as predicted, a
significant difference between envy + jealousy reported toward
groups falling into the high-competence/low-warmth cluster
compared with the average envy + jealousy of all other clusters
[F(1, 38) = 15.31, P < 0.001]. Also as predicted, contrast analysis
found significant differences between pride + admiration re-
ported toward groups falling into the high-warmth/high-compe-
tence cluster compared with the average pride + admiration
reported of all other clusters [F(1, 38) = 28.42, P < 0.000].
Contrast analyses did not reveal significant differences between
disgust + contempt reported toward low-warmth/low-competence
groups compared with all others, due to the high disgust + con-
tempt reported toward groups falling into the high-competence/
low-warmth cluster. These data suggest that groups seen as high
competence and low warmth are on the receiving end of not only
feelings of envy, but also those of disgust.
Brief Survey About Climate Scientists Perceived Motivations. Methods.
Participants. Participants who had not participated in either
previous study were recruited to complete the brief survey in the
same manner as before; 52 participants (27 men; mean age =34.81 y,
SD = 11.04) completed the brief survey for standard MTurk
pay rates.
Questionnaire and procedure. Participants read that the survey
focused on thoughts about different social groups. Then, they
encountered two open-ended questions about climate scientists
motivations, in response to this prompt: Many climate scientists
(climatologists) argue that human activity is largely responsible
for climate change worldwide. (i) Why do you think climate
scientists make such an argument? (ii) What do you think
motivates climate scientists to spread this argument to the pub-
lic? (These responses have not been analyzed.)
Participants then read: Climate scientists (climatologists)
argue that human activity is largely responsible for climate change
worldwide. Please rate the extent to which you believe the following
reasons apply to climate scientists who make this argument, using
a five-point scale from extremely unlikely to extremely likely, fol-
lowed by items generated from pilot work: Climate scientists wish
to lie with statistics; . . . educate the public; . . . make a simple story
into something quite complicated; . . . save the earth; . . . save hu-
manity; . . . show the public how intelligent and superior to others
they are; . . . gain more money from the government for their re-
search; . . . are simply pursuing a liberal agenda; . . . wish to save
the environment; . . . just wish to be provocative; . . . wish to
hurt big corporations.
Results and discussion. Principal component factor analysis with
varimax rotation yielded two factors described in the main text as
trust (three items) and distrust (seven items). Individual items
were averaged to yield the respective scales. See Climate Scientists
Presumed Agendas for the list of items.
1. Fiske ST, Cuddy AJC, Glick P (2007) Universal dimensions of social cognition: Warmth
and competence. Trends Cogn Sci 11(2):7783.
2. Fiske ST, Cuddy AJ, Glick P, Xu J (2002) A model of (often mixed) stereotype content:
Competence and warmth respectively follow from perceived status and competition. J
Pers Soc Psychol 82(6):878902.
3. Cuddy AJC, Fiske ST, Glick P (2007) The BIAS map: Behaviors from intergroup affect and
stereotypes. J Pers Soc Psychol 92(4):631648.
4. Cuddy AJC, et al. (2009) Stereotype content model across cultures: Towards universal
similarities and some differences. Br J Soc Psychol 48(Pt 1):133.
5. Durante F, et al. (2013) Nations income inequality predicts ambivalence in stereotype
content: How societies mind the gap. Br J Soc Psychol 52(4):726746.
6. Fiske ST, Cuddy AJ, Glick P, Xu J (2002) A model of (often mixed) stereotype content:
Competence and warmth respectively follow from perceived status and competition.
J Pers Soc Psychol 82(6):878902.
7. Cuddy AJC, Fiske ST, Glick P (2007) The BIAS map: Behaviors from intergroup affect and
stereotypes. J Pers Soc Psychol 92(4):631648.
Table S1. Distinct emotions reportedly directed toward distinct warm competence clusters of job holders
Cluster Envy + jealousy Disgust + contempt Pride + admiration Pity + sympathy
Lowlow 1.28 2.07 1.52 2.79
Highhigh 2.46 1.55 3.26 2.28
High competencelow
warmth
3.13 2.08 3.16 1.49
Middle 2.03 1.96 2.21 2.20
Means predicted to be higher than others in their columns are presented in bold.
Fiske and Dupree www.pnas.org/cgi/content/short/1317505111 2 of 2