^•PLOSIONE Check for updates fi OPEN ACCESS Citation: Jarreau PB, Cancellare IA, Carmichael BJ, Porter L, Toker D, Yammine SZ (2019) Using self ies to challenge public stereotypes of scientists. PLoS ONE 14(5): e0216625. https://doi.org/ 10.1371/journal.pone.0216625 Editor: Mellissa H Withers, USC Keck School of Medicine, Institute for Global Health, UNITED STATES Received: January 31,2019 Accepted: April 24,2019 Published: May 10,2019 Copyright: © 2019 Jarreau et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Data Availability Statement: All relevant data are within the manuscript and its Supporting Information files. Funding: This study was funded through the Experiment.com crowdfunding campaign "To Selfie or Not to Selfie - How Can Scientists Foster Public Trust on Instagram?" that raised $10,704.00 for a study of the perceptions of scientists on Instagram. This funder/funding provided support for survey participant incentives and viewing devices for lab experiments, but did not have any RESEARCH ARTICLE Using selfies to challenge public stereotypes of scientists Paige Brown Jarreau 1'2*, Imogene A. Cancellare3, Becky J. Carmichael2, Lance Porter4, Daniel Toker5, Samantha Z. Yammine6 1 LifeOmic, Indianapolis, Indiana, United States of America, 2 College of Science, Communication across the Curriculum, LSU, Baton Rouge, Louisiana, United States of America, 3 Department of Entomology and Wildlife Ecology, University of Delaware, Newark, Delaware, United States of America, 4 Manship School of Mass Communication, LSU, Baton Rouge, Louisiana, United States of America, 5 Helen Wills Neuroscience Institute, UC Berkeley, Berkeley, California, United States of America, 6 Donnelly Centre for Cellular and Biomolecular Research, University of Toronto, Toronto, Ontario, Canada * paiqebjarreau@qmail.com Abstract In an online Qualtrics panel survey experiment (N = 1620), we found that scientists posting self-portraits ("selfies") to Instagram from the science lab/field were perceived as significantly warmer and more trustworthy, and no less competent, than scientists posting photos of only their work. Participants who viewed scientist selfies, especially posts containing the face of a female scientist, perceived scientists as significantly warmer than did participants who saw science-only images or control images. Participants who viewed selfies also perceived less symbolic threat from scientists. Most encouragingly, participants viewing selfies, either of male or female scientists, did not perceive scientists as any less competent than did participants viewing science-only or control images. Subjects who viewed female scientist selfies also perceived science as less exclusively male. Our findings suggest that self-portraiture by STEM professionals on social media can mitigate negative attitudes toward scientists. Introduction There is growing awareness within the scientific community that researchers and their institutions can and should make public engagement and communication more of a priority [1-3]. But while scientists are answering calls to leverage new media tools [4], it is important for them to consider that not all science communication efforts are created equal. The impacts of these efforts depend on a variety of factors including real or perceived characteristics of the message (content, tone, frame, etc.), the audience(s), and the communicator(s) [5]. People tend to be cognitive misers in evaluating information and forming attitudes [6]. They often use a communicator's perceived trustworthiness as a heuristic for how credible, personally relevant, reliable, pertinent, and persuasive a message is [7, 8]. People are more likely to pay attention to and act upon "quality information communicated by sources they see as trustworthy in terms of expertise, honesty, and shared identity" [5] and PLOS ONE I https://doi.orq/10.1371/iournal.pone.0216625 May 10, 2019 1 /23 ^•PLOSIONE Scientists who selfie additional role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing interests: We declare a commercial affiliation with LifeOmic. Author PBJ has a current affiliation with LifeOmic (salary). This affiliation started in March 2018, after all data had been collected for this study. LifeOmic did not play a role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript, nor provided any direct financial support connected to this study. Since March 2018, LifeOmic provided support in the form of salaries for authors [PBJ] (which was not in any way connected to the work or time spent writing up the results of this study and submitting the study for publication), but did not have any additional role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript. All authors listed contributed equally to this study. All of the authors of this manuscript are Instagram users and several could be called influencers within the IG science community (IAC, DT, SZY). We also declare a commercial affiliation to LifeOmic. This does not alter our adherence to PLOS ONE policies on sharing data and materials. from sources they "like" [9]. Trust, which involves mutual understanding between communicator and audience, is vital for scientists speaking on issues of pressing public concern such as public health or climate change [10, 111. The audience member should feel that scientists "get me". Scientists and their institutions need to be sincerely friendly, demonstrate a shared value system, and engage people's emotions [12]. High quality interactions with likeable scientists encourage positive beliefs about science and scientists' messages [3]. Unfortunately, scientists are generally perceived by Americans as highly competent but only moderately warm; they "earn respect but not trust" [12]. Studies investigating scientists' prioritization of public communication objectives have also found very little attention to this issue of warmth or focus on building trust [3,13]. Scientists' isolation in "the ivory tower" [2, 14] and focus on defending science rather than building relationships with the public [ ], along with stereotype-reinforcing media portrayals, have likely contributed to both ambivalent public perceptions and more negative stereotypes of scientists. Scientists' use of direct-to-public communication tools such as social media [15] may be spurring concern over their public image. However, social media could also be an avenue for relationship building between scientists and citizens. These relationships could help challenge stereotypes of scientists as being competent but socially inept and eccentric [16], cold, and capable of immoral conduct [17]. To accomplish this, scientists must use social media in ways that communicate their warmth and not just their scientific competence. Background and literature review Social cognition, warmth and competence We use social cognition and the stereotype content model [18] as the framework to understand how perceptions of scientist Instagrammers may impact trust in scientists. Social cognition consists of the idea that "people everywhere differentiate each other by liking (warmth, trustworthiness) and by respecting (competence, efficiency)" [19]. The stereotype content model postulates that group stereotypes can be captured by the two universal dimensions of social cognition: warmth and competence [18]. Warmth is based on one's perceptions of another individual's or group's intentions, while competence is based on one's confidence that the individual or group can act on their intentions. Traits that add to perceived warmth include friendliness, tolerance, modesty, openness, sociability, honesty, happiness, helpfulness, sincerity, and morality. Traits that add to perceived competence include intelligence, skill, creativity, and persistence [19]. Both competence and warmth are critical constructs of trust. However, perceived warmth carries more weight in terms of people's attitudes and behaviors toward an individual or stereotyped group [19]. Morality or warmth judgments are primary— they are made more quickly, they determine approach or avoidance behavior, and they carry more weight in attitude formation than competence judgments do [19-21]. Outgroups (people unlike oneself) are often perceived as either competent but not warm and associated with feelings of threat and envy, neither competent nor warm and associated with feelings of contempt, or warm but not competent and thus pitied [19]. How one perceives outgroups can be predicted by the perceived status of the group and competition with one's in-group. High status and competitive outgroups such as scientists are often perceived as competent but not very warm [12]. In-group individuals, close allies, and other groups like nurses and teachers are admired and perceived as both highly warm and competent [12]. Instead of threatening, they are "like you" in their underlying interests and values [19, 22, 23]. They have high potential to be compelling communicators. By sharing their values and humanizing themselves, we think scientists can get there too. PLOS ONE | https://doi.orq/10.1371/iournal.pone.0216625 May 10, 2019 2/23 ^•PLOSIONE Scientists who selfie Perceptions of scientists Longitudinal results from the General Social Survey (GSS) since the 1970s suggest relatively stable and high levels of U.S. public confidence in scientists. Men, younger individuals, and individuals with higher levels of education and income express the highest levels of confidence in scientists. However, an increasing percentage of the U.S. public also finds scientists to be "odd and peculiar" (52% strongly agreed or agreed, up from 36% in 2012) [24]. This stereotype reflects longtime fiction media portrayals of scientists as eccentric, strange and detached from everyday life [25, 26]. Based on this and other stereotypes, scientists as a group don't immediately inspire friendliness and warmth, vital characteristics of trusted communicators. We should note that media portrayals of scientists have improved in the last decade toward characters that can generally be categorized as intelligent and "good" [27]. However, portrayals of scientists in both entertainment and news media still lack diversity, particularly in terms of gender and race according to the recent "Portray Her" report from The Lyda Hill Foundation and the Geena Davis's Institute on Gender in Media. Popular images of the scientist as an elderly, white-haired man and people's general unfamiliarity with real-life scientists and what they do tend to maintain negative stereotypes or at least perceptions that scientists are among an elite outgroup that doesn't share the values, interests, and identity of most people. Gender science stereotypes, female scientists in the media, and warmth stereotypes Warmth and competence stereotypes of scientists are likely exacerbated by gender stereotypes and gender science stereotypes that associate STEM with being male [28]. Women are generally perceived as more warm, likeable, and friendly, but slightly less competent, than men [7, 181. Positive perceptions of female scientists as not only competent but also warm could be a positive force for scientist stereotypes, except that gender science stereotypes limit opportunities for women to succeed, become leaders in their fields, and positively influence the image of science. Pervasive stereotypes associate science and scientific work with men more than women [29]. These stereotypes both aggravate and are aggravated by gender gaps in retention, achievement, visibility, and leadership in STEM [30-37]. The Catch-22 for scientists' public image problem is that women scientists who could improve perceptions of scientists' relatabil-ity and warmth are both underrepresented in prestigious scientific media outlets [38] and in media that nonscientists consume [39]. They are also judged for taking on stereotypically male roles and traits associated with scientific work [35,40, 41]. The clashing of gender roles in STEM is exemplified by the fact that female scientists are often told that they "don't look like scientists" [42]. Independent, self-confident, competitive, senior women in male-dominated fields are harshly critiqued [43]. This makes it difficult for women to achieve the equal representation in STEM and the media attention that could help shift stereotypes of scientists and help young girls see themselves in STEM. Scientists from other marginalized groups face similar challenges. The scientist as we've known "him" from literature is a white male, a social outsider who is unscrupulous or even dangerous in his pursuit of scientific discovery [44] • More modern media portrayals of scientists also tend to be male. Female scientists are often presented in media programs as more of an exception to the rule (a lonely heroine, an unlikely clever beauty, a rare superstar) than as positive counter-stereotypes [45]. Media portrayals of female scientists are likely to show them in supporting roles to male scientists and to focus on their appearance, sexuality, and personal lives as opposed to their work successes [46]. A fetishizing of the appearance of female scientists in the media reinforces gender role beliefs and promotes the idea that women in STEM are eye candy rather than experts. This was exemplified by the PLOSONE | https://doi.orq/10.1371/iournal.pone.0216625 May 10, 2019 3/23 ^•PLOSIONE Scientists who selfie #distractinglysexy social media campaign that became viral in 2015 [47]. Media misrepresentation undermines women scientists' ability to positively influence the stereotypical image of scientists as aloof, nerdy, socially inept, and male. Social media platforms may offer women and minority scientists an opportunity to create their own narratives through self-images and new takes on what scientists look like, what they do, what they care about, and how both warm and competent they are. A diversity of real-life scientist faces in our media environments could help others begin to individuate scientists, help break down stereotypes, and foster more inclusion in STEM. Using visuals to put a friendly face on science and change stereotypes Visual social networking platforms cater to real-time self-disclosure, which is critical to the relationship building that can promote trust between individuals and break down group stereotypes [48, 49]. One way to operationalize how a scientist could visually communicate their warmth is through today's cultural artifact for online self-disclosure—the "selfie". Selfies close the distance between photographer and viewer—even astronauts take selfies [50]. There is an emerging trend within the online STEM community to use portraits to help change people's perceptions of what a scientist looks like and does on a daily basis. Hashtag movements such as #ThisIsWhatAScientistLooksLike, #BlackandSTEM, #DistractinglySexy, and #Scientists-WhoSelfie have attempted to combat stereotypes of STEM professionals through visual imagery [51]. While these movements may promote feelings of empowerment within the group seeking to change its public image, it is unknown whether such movements can actually improve stereotyped perceptions of scientists. The idea of using portraits to change perceptions has a basis in social cognition research. Counterstereotypical exemplars have been shown to change perceptions of and bias toward outgroups, as well as reduce or revise stereotypes [52]. Individuation processes, in which viewers pay attention to a person's individual attributes and see that person as human with ideally shared beliefs and good intentions [53], can help counter negative stereotypes about that person and their group. But individuating takes effort [54]. People often don't put this effort toward stereotyped outgroups or groups they aren't familiar with. However, individuation can take place when people are asked or motivated to take notice of individual attributes or pay extra attention to individual faces [53]. Individual processing of scientists' faces, especially when they display warmth attributes, might help challenge stereotypes. Positive facial expressions [55], images of friendly faces [56], and images of attractive faces [57] can influence social cognition. Smiling faces vs. neutral faces score higher on attributes including friendliness [58], generosity, agreeableness, sincerity [59], extroversion [55], and warmth [60]. Smiling can also combine with gender to influence perceptions of warmth. In one study, female smiling faces were evaluated as warmer than male smiling faces [56]. But while facial expressions can influence interpersonal trait inferences, it is unclear whether they can, when displayed by members of outgroups, move the needle on warmth stereotypes. There is evidence that positive images of individuals can suppress automatic attitudes and stereotypes [61], that social exposure to counterstereotypes (like women leaders) can reduce automatic gender stereotyping [62], and that counterstereotypic mental imagery (e.g. imagining a strong woman) can reduce implicit stereotypes [63]. In the study investigating exposure to counterstereotypic women leaders [62], exposure to a greater number of women faculty and deans over time had long-term impacts (on the range of at least a year, with ongoing exposure) on automatic gender stereotypes. Short-term exposures to counterstereotypic scientists via media stories or social media are likely to have less long-term impacts on scientist and science gender stereotypes. However, these exposures might PLOS ONE I https://doi.orq/10.1371/iournal.pone.0216625 May 10, 2019 4/23 ^•PLOSIONE Scientists who selfie have greater impact if viewers discovered and then followed counterstereotypic (e.g. female) scientists on Instagram, for example, thus seeing more from these scientists over time and even developing relationships with them. The relative impacts of repeated exposure to counterstereotypic scientists on Instagram as opposed to one-time exposure are out of the scope of this study, but should be investigated in future research, especially if a one-time exposure can produce significant changes in social cognitions. Why study Instagram as a tool for building trust in science Instagram (IG), released in 2010, is a photo and video social media sharing platform that encourages viewers to enter Instagrammers' worlds and get to know them through selfies and self-videos. Scientists, including several authors of this paper, are using IG to open up their lab and fieldwork for broader audiences. An interconnected group of thousands of scientists at various career stages use IG to share their lives and work as scientists, for example via #weareS-TEMsquad. A number of popular "Rotation Curation" accounts feature different scientists through images and captions about their work, career paths, hobbies, etc. Through the use of hashtags, users can identify and follow trending topics. Scientists can also reach outside of their scientific circles through use of popular hashtags. Hashtags enhance the visibility of an Instagram post as it then becomes searchable and may even be featured in Instagram's dynamic "top" or featured lists of posts for that hashtag. Strategic use of hashtags that resonate with key target demographics can help an Instagrammer get their topic seen by broader audiences who follow that hashtag but not the Instagrammer him-or herself [64]. For example, a mix of science-specific hashtags and less science-focused hash-tags will allow an Instagram post to perform well but also show up in the feeds of those who may not regularly follow science content. This is a common technique used in digital marketing that depends on some level of prior "audience listening" to anticipate popular trends and key words. In summary, the growth and visibility of a diverse, humanized scientific community on IG begs a greater understanding of how viewers evaluate scientist Instagrammers and whether connecting with scientists on IG fosters trust. This study sought to fill a gap in understanding how humanized content posted by scientists on Instagram impacts perceptions of scientists among viewers. We were interested to see whether such humanized content could improve perceptions of scientists' warmth, as suggested by prior studies on the impacts of positive facial expressions on social cognition, and change stereotypes about what a scientist typically looks like, or the typical characteristics and demographics of scientists. Research questions and hypotheses Our primary research question for this study was whether scientist selfies on IG posted by scientist Instagrammers (IGers) change stereotypes of scientists' warmth vs. competence. We expected that individuating a multitude of scientists in selfies displaying warmth traits (broad smiles, eye contact, friendly gestures) would help change group warmth stereotypes for scientists in general. Based on our literature review, we hypothesized the following: HI. Scientist IGers who post selfies will be perceived as warmer than scientist IGers who post science-only photos. H2. Scientist selfies on IG will positively influence perceptions of scientists' warmth in general, as compared to science-only IG posts. We also explored the impact of stimulus condition on competence, although we did not expect scientists posting selfies to be perceived as more or less competent than scientists PLOSONE | https://doi.orq/10.1371/iournal.pone.0216625 May 10, 2019 5/23 ^•PLOSIONE Scientists who selfie posting science-only photos. We did not expect changes in perceived competence for our IGers or scientists in general given the presence of prominent scientific competence cues (lab coats, scientific equipment, etc.) and scientific captions used consistently across stimulus groups. We predicted that female scientists would be perceived as warmer than male scientists. We also expected that a strong representation of women scientists in IG posts would positively influence perceptions of gender science stereotypes as well as perceptions of scientists' warmth. H3a. Female scientist IGers will be perceived as warmer than male scientist IGers. H3b. Female scientist IG posts (especially selfies) will result in positive changes in gender science stereotypes, as compared to male scientist IG posts, especially for selfies. H3c. Female scientist IG posts (especially selfies) will result in more positive perceptions of scientists' warmth in general, as compared to male scientist IG posts. Materials and methods Online survey experiment This study was approved by the Louisiana State University Institutional Review Board: IRB# E10449, Expiration 4/12/2020. Participants agreed to participate in the study consent by viewing an electronic consent form—written consent was not required (data analyzed anonymously). Survey respondents were recruited and screened through a paid Qualtrics Panel to meet U.S. representative quotas on gender, education and age (see SI). We worked with Qualtrics to eliminate and replace 200 "trash" or unintelligible survey responses based on an open-ended stimulus recall question. Respondents viewed IG content embedded in a private Square-space website designed to look like an aggregation site for posts from a "Scientists of Insta-gram" Rocur IG account (see www.scientistsofig.com). Participants were instructed to return to our survey to complete the study after browsing the content to answer questions about the IGers they encountered. Data processing. A total of 1,620 survey responses were recorded and included in our final analyses. These included complete and partial but substantially complete responses. We removed responses where the stimulus materials had been viewed for less than 15 seconds. Where we detected that respondents had entered "5" or "1" across the board on warmth/ competence scales, regardless of survey item direction, we removed these responses. Following this removal, no extreme outliers for warmth/competence were detected in SPSS (interquartile range test outlier test). Our final analyses were based on 339 respondents viewing control IG posts, 319 respondents viewing science-only male posts, 313 respondents viewing science-only female posts, 318 respondents viewing male selfies, and 329 respondents viewing female selfies. Our SPSS dataset (SI Dataset) is available online with a filter variable labeled "Outlier does not equal 1 (FILTER)" provided to remove the inappropriate responses as described above. Experimental design and stimulus content All research participants were randomly assigned to one of five different experimental conditions, in all of which they saw a series of Instagram posts (photographs and captions): (1) a control IG account of male and female "Humans of Southeastern" (@humansofselu, lab experiment) or "Humans of Broadway" (@humansofbroadway, online survey experiment); (2) a "Scientists of IG" account depicting seven photos of scientific objects / equipment in lab or PLOSONE | https://doi.orq/10.1371/iournal.pone.0216625 May 10, 2019 6/23 ^•PLOSIONE Scientists who selfie field settings, posted by seven different male scientists as identified by first name / pseudonym in the caption (Sam, Liam, Deboki, Martin, Kyle, Luis, Ian); (3) same as (2) but posted by seven different female scientists (Samantha, Laura, Deboki, Martha, Kylla, Luisa, Imogene); (4) a "Scientists of IG" account depicting seven photos of the same scientific objects / equipment but with a male scientist's smiling face present; (5) same as (4) but with all female faces. We refer to these stimulus conditions throughout this manuscript as: (1) control; (2) science-only male; (3) science-only female; (4) male selfies; (5) female selfies. Selfies were recreations of science-only photos (same objects/equipment and same overall scene) but containing a scientist's smiling face. For example, one science-only photo depicted a colorful bioreactor set up on a lab bench, while the selfie equivalent depicted the bioreactor from the same angle but with a male/female scientist in the shot pointing to the equipment and smiling at the camera. See SI Appendix for a detailed description of stimulus content. Following stimulus exposure, participants were asked questions about their enjoyment of the content and their impressions of the IGers they had encountered. They were then asked 1) to evaluate the warmth and competence of individual IGers as represented by a select sample of visual-only screenshots from the stimulus they had just browsed (embedded in an online questionnaire), and 2) to answer questions about their perceptions of scientists in general. We randomized the order of these question sections (IGer evaluations and scientist stereotypes) for our online survey experiment. We complied with the terms of service for the sites and social media platforms that we used in this study for stimulus exposure, including Instagram and Squarespace. We provided accurate information about the scientific work of the scientists featured in our stimulus images, with their permission, using their first names / pseudonyms in the IG post captions (pseudonyms created so that the male and female names for corresponding male and female selfies were complimentary, e.g. Sam and Samantha). All stimulus content (images and captions) was created originally for this study and used with the permission of the creators. All individuals who helped create content and who were featured in the stimulus images for our study, some of which are featured in Fig 1, have given written informed consent (as outlined in the PLOS consent form) to publish these case details, their real names and the photographs/self-portraits they produced for our study. Measures See SI Appendix for all survey question wording and more information on scale items. Enjoyment. We asked respondents to rate their enjoyment of the stimulus content in a series of questions about how interesting, visually appealing, credible, useful, and educational the IG posts were. These questions were combined into a 9-item enjoyment scale (Cronbach's alpha = .93, M = 3.57, SD = .82). Perceptions of scientist IGers. We measured how trustworthy respondents perceived the IGers portrayed in stimulus content to be on 7-point Likert scale. For the selfie conditions we also measured the extent to which respondents thought the IGers looked like scientists (5-point Likert-rype scale). We measured the extent to which respondents saw themselves as similar or dissimilar to the IGers in terms of values, goals, opinions, thoughts, and beliefs (7-point bipolar scales based on opposite word pairs). These questions were combined into a 5-item value similarity scale (Cronbach's alpha = .89, M = 4.17, SD = 1.36). Warmth and competence of scientist IGers. We used a selection of four photos from our experimental IG posts to ask respondents to evaluate the warmth, competence, and physical attractiveness of individual scientist IGers. Based on photos without captions, respondents were asked how warm/competent/attractive the person who took each photo (science - PLOSONE | https://doi.orq/10.1371/iournal.pone.0216625 May 10, 2019 7/23 ^•PLOSIONE Scientists who selfie #ScientistsWhoSelfie _N = 1620 Total Survey Respondents_ Male/Female Scientists: Photos of scientific objects with caption attributed to male or female scientists Male/Female Selfies: Photos of scientists posing with the same scientific objects Female Scientist • n = 313 • No significant differences in primary outcomes| as compared to science posts by male IGers Female Selfie{S] a 1 • n = 329 • attractiveness > stereotypes (Effect = .03, 95% CI [.01, .04]); stimulus > warmth of IGers > stereotypes (Effect = .18, 95%CI [.14, .23]); stimulus > attractiveness > warmth of IGers > stereotypes (Effect = .04, 95% CI [.02, .05]). The total indirect effect of stimulus on warmth stereotypes (Effect = .25, 95%CI [.20, .29]) is greater than the total effect {J3 = .13, £(1236) = 3.81,p < .001). The stimulus condition combines selfies of both genders. However, female scientists in selfies are evaluated as significantly more attractive than male scientists (p < .001), and male scientists in selfies are not evaluated as any more attractive than the scientists who take science-only photos are guessed to be (see SI). We also found that attractiveness of IGers is a mediator of stimulus effect on competence stereotypes in a similar PROCESS (Model 4) analysis with covariates as described above. Without attractiveness, stimulus had a significant and positive total effect on competence stereotypes 08 = .10, t(1236) = 2.69,p < .01; Model R2 = .17,p < .001). However, attractiveness was also a significant predictor of competence stereotypes (fi = .14, £(1235) = 4.54, p < .001; Model PLOSONE I https://doi.orq/10.1371/iournal.pone.0216625 May 10, 2019 14/23 ^•PLOSIONE Scientists who selfie R2 = .19, p < .001) and suppressed the direct impact of stimulus on competence when included in the regression model (/> = .07). Gender science stereotypes and female scientists' warmth and competence We hypothesized that seeing female scientists in IG posts would counteract gender science stereotypes and promote warmth (H3b and H3c). We ran ANCOVA analyses and linear regression models using dichotomous factors for stimulus condition (science-only vs. selfies) and scientist gender (male vs. female). We also accounted for participant age, gender, education, interest in science, importance of religion, IG use, political affiliation, and knowing a scientist personally. We found that only female scientist selfies resulted in a significant shift in stereotypes that associate STEM fields with being male (see Table 2). In a linear regression model (F(ll, 1236) = 5.36, p < .001; R2 = .04), the presence of female scientists in IG posts was a significant predictor {fi = .08,/) < .01) of female-shifted science stereotypes, while overall stimulus condition (science-only vs. selfies) was not a significant predictor (fi = .04,/) = .21). Other significant predictors of female-shifted stereotypes were younger age ifi = -.08,/) < .05), lower education level ifi = -.15,p < .001), and less IG use (fi = -.07,p < .05) among participants. Thus, we find support for H3b. To investigate warmth stereotypes by scientist gender, we ran a linear regression model (F (11,1236) = 10.23,/) < .001; R2 = .08) predicting warmth using two variables: a collapsed stimulus variable (science-only vs. selfies) and a scientist gender variable (male vs. female) (see Table 3, Model 2). All other predictors included in the model were the same as those used to investigate warmth in the Predicting Science Stereotypes section above. We found that across all stimulus conditions, scientist gender was not a significant predictor of warmth. However, when we used dummy coded variables for all stimulus conditions with the control group as the reference (F(13, 1567) = 11.38, p < .001; R2 = .09), we found that only female scientist selfies resulted in significantly greater perceptions of scientists' warmth ifi = .12, p < .001). Male scientist selfies resulted in marginally greater perceptions of scientists' warmth compared to control images (fi = .06,p < .06). Female scientist selfies resulted in significantly greater perceptions of scientists' warmth compared to every other stimulus condition individually (vs. scimale,j8= .12,/) < .001; vs. sci female,/J = .11,p < .001; vs. selfie male.^S = .06,/) < .04). This increase in perceived warmth was accompanied by an increase in perceived competence. Only female scientist selfies resulted in significantly greater perceptions of scientists' competence in general (F(13,1567) = 7.37,/) < .001; R2 = .06;0 = .06,/) < .05), with the control group as reference. Discussion Scientists' less-than-warm public image is problematic because warmth is a key component of trust [19]. To build trust, scientists might consider humanizing science through friendly self-portraiture in scientific settings. Scientists are generally perceived as competent but often stereotyped as nerdy and socially inept, or more generally not warm. We found that humanizing scientists via Instagram selfies can counter these stereotypes. Viewers of scientist selfies, both from male and female scientists, perceived both scientist Instagrammers and scientists in general to be warmer and at least as competent or more competent than viewers of science-only or control Instagram posts did. Seeing scientist selfies—as opposed to images of scientific objects posted by scientists online—thus appears to selectively reduce negative stereotypes, likely through a process of individuating scientists based on their faces. Here, this process involved seeing a series of smiling faces of particularly PLOSONE | https://doi.orq/10.1371/iournal.pone.0216625 May 10, 2019 15/23 ^•PLOSIONE Scientists who selfie counterstereotypic scientists and identifying them as "warm" individuals as opposed to unfamiliar members of an envied and relatively "cold" outgroup. In terms of preserving perceived competence, competence cues (e.g. a scientific setting, scientific equipment, recounting of scientific work being conducted visually and through photo captions), may be an important feature of scientist selfies. We found gender to be a major driver of changes in stereotypes of scientists' warmth and gender science stereotypes. We found that seeing a series of female scientist selfies on Instagram can shift stereotypes that associate STEM fields with being male, as well as positively shift stereotypes about scientists' warmth. Seeing and evaluating individual smiling/friendly faces contributed to this warmth shift. These findings again suggest that people viewing scientists' Instagram posts begin to individuate scientists as opposed to categorizing them into an outgroup that is competent but cold. This individuating process could also help explain why female scientist IGers posting selfies were not evaluated as stereotypically less competent than male scientist IGers, even if their gender elicited greater perceived warmth. Their identity as competent scientists may have held a prominent sway over these evaluations. Counter to our initial concerns that female scientist selfies could trigger gender stereotypes causing viewers to perceive these scientists as warmer but also less competent and more narcissistic, for example, we found that female scientist selfies actually led to more positive competence evaluations for scientists overall and significantly reduced perceptions that scientists in general are "vain" (see SI). Although previous research has shown that female smiling faces may influence men's social judgments to a larger extent than women's social judgments [55], we found that our female scientist selfies had a similar effect for both male and female viewers. On the other hand, science-only photos posted by female scientists did not significantly shift gender science stereotypes. This could be because the female names in the captions were not attention-grabbing and/or memorable, or because participants did not read the captions. We did not evaluate whether participants properly attributed our science-only posts to scientists of the intended gender. However, it is likely that female scientist selfies were most successful in moving gender science stereotypes due to the greater accessibility of gender in these posts. In summary, we suggest that humanizing, visual science communication may be one way for broader audiences to "get to know" scientists. However, we believe that scientists should be genuine and transparent in humanizing themselves. By sharing selfies and stories from their daily lives in the lab/field, talking about their motivations and struggles, inviting viewer participation, and opening up the scientific process, scientists could foster trust by individually and collectively embodying warmth as well as competence. Social media platforms like Instagram are unique in the opportunities they offer researchers to reach public audiences. Social media platforms give diverse researchers a platform to express themselves freely—an opportunity traditional media outlets historically have not given women and minorities [66]. Social media platforms could represent a turning of the tide for perceptions of scientists if leveraged by diverse individuals to interact with broad audiences in ways that communicate warmth and build mutual trust. We acknowledge that reaching truly broad audiences can be a challenge for the scientist communicator. Training and planning can help. For example, the Instagram algorithm relies on engagement to evaluate which posts users see in their feed. The strategic use of relevant hashtags and interactive options (e.g. quizzes and fill in the blank prompts unique to the Instagram Story) are likely to increase overall engagement rates and thus increase scientists' exposure to non-scientist audiences. PLOSONE | https://doi.orq/10.1371/iournal.pone.0216625 May 10, 2019 16/23 ^•PLOSIONE Scientists who selfie Limitations We cannot pinpoint the exact mechanism by which selfies increased participants' perceived warmth of scientist Instagrammers and changed their stereotypes of scientists. We prioritized externally valid stimulus content (created for us by scientists in the IG community) and guided participants to browse a series of Instagram posts that combined multiple warmth-related characteristics. For example, selfie conditions consisted of seven smiling scientist faces of different ethnicities and styles, together with captions written from the perspective of each featured scientist. The presence of smiling expressions, the diversity in scientist appearance (including ethnicity), and the combination of selfies with captions describing scientists' motivations to do their work—"I have been developing new and exciting technologies using genetic engineering to make our immune system better at fighting cancer"—might all have contributed to changes in stereotypes. The effects of these stimulus elements should be teased apart in future work. The physical attractiveness of our scientist Instagrammers also played a role in their perceived warmth and competence. We observed that attractiveness is a mediator of individual competence and warmth evaluations. Our female scientist Instagrammers were consistently evaluated as more attractive than male scientist Instagrammers, although this could be due to stereotypes that focus on women's physical appearance and lead to female smiling faces being more positively evaluated than male smiling faces. We did not manipulate but rather measured attractiveness in our experiments. That said, we note that we tried to match the physical appearance of male and female scientists in selfies (similar clothing, only light makeup, etc.). Perceived attractiveness has been associated with assessments of moral goodness, aka the "Beauty-Is-Good" stereotype [67], trustworthiness [68], and intelligence [69]. This may help explain the boosts in competence and warmth that our female scientist Instagrammers in selfies received. The Beauty-Is-Good stereotype particularly influences judgments of social competence and interpersonal ease [70]. We note that all of the faces we used in our experiment displayed broad smiles and friendly expressions, which likely contributed not only to their perceived warmth but also their physical attractiveness [71]. We suggest that future studies investigate the effects of smiling vs. neutral scientist faces on perceptions of scientists' warmth and competence, and the extent to which gender stereotypes might interact with those effects. The positive effect of selfies on warmth stereotypes of scientists may also be dependent on viewers being motivated to individuate the scientists in these images and see them as human. We prompted such motivations by asking participants to evaluate individual scientist Instagrammers—their warmth, competence and attractiveness, how likely viewers would be to follow them in real life on Instagram, etc.—before asking participants to reflect on perceptions of scientists in general. When participants were asked questions about scientist stereotypes before evaluating individual Instagrammers, our stimulus had a much weaker effect on warmth stereotypes (see Table C in SI Appendix). This is a limitation that should be investigated in future studies. Another limitation of this study is that the results do not reveal whether or not scientists' humanizing efforts on social media could have long-term impacts on stereotypes or even real-world impact on behavior. While one-time exposure to a series of humanized science images on IG reduced scientist and science gender stereotypes in our study, it remains to be seen whether this effect could extend beyond the experimental setting and persist over time. Likely important factors for this would be long-term or repeated exposure, two-way communication (e.g. scientist Instagrammers spending time replying to comments and engaging with PLOSONE | https://doi.orq/10.1371/iournal.pone.0216625 May 10, 2019 17/23 ^•PLOSIONE Scientists who selfie followers), and meaningful relationship building between viewers and counterstereotypic scientists. These factors are theoretically supported by Instagram (through hashtag browsing, following, comments and replies, IGTV live broadcasts from the lab/field, etc.), but we did not incorporate these factors into this experimental design. Long-term, frequent, and meaningful exposure to counterstereotypic examples appear to be key to building trust and changing stereotypes long-term [62]. Juxtaposed to the concern about whether impacts of humanized science IG posts can be long-lasting is the concern that those spreading pseudoscience could leverage our "scientist selfie" approach to boost their own trustworthiness. There is no regulation as to who can use Instagram as a "scientist"; many have exploited scientific jargon and paraphernalia online to gain undue credibility and to spread pseudoscientific information. Some scientists have called for Twitter and Instagram to "verify" them (with a blue Verified badge) as it does other media figures and celebrities. However, a top-down approach to identifying "scientists" for verification could have the side effect of reinforcing traditional conceptions of who counts as a scientist. Considering that minorities in science may also be more likely to be boundary spanners, that women and minority scientists tend to be much less represented in the media, and that women scientist communicators accrue fewer social media followers [72], platform-initiated verification of scientists would be inherently problematic. We can instead support and amplify, as a community, grassroots humanized content from self-identified scientists on Instagram and other social networks to challenge science stereotypes, and train scientists in social media marketing to increase their reach. Self-identified scientists can use other source credibility cues (credentials, links to academic websites, references to peer-reviewed literature, etc.) to establish their credibility and raise awareness of how scientific credibility should be evaluated. In summary, we believe that the power of selfies lies in humanizing and individuating scientists of all appearances who display warmth traits (smiling, eye contact, self-disclosure, sincerity, etc.). Any real-life scientist's smiling face is likely to be perceived as warmer than common perceptions of scientists as peculiar, unfamiliar, and unfriendly. When it comes to smiling, however, we acknowledge gender bias that leads to women being asked to "smile" for the benefit of others, a request not often made of men. We do not endorse any use of our findings to bolster such bias. Finally, there are many ways for scientist communicators to display warmth, honesty, and shared values with audiences. The more diverse the approaches, the greater the potential reach. To connect on a human level through a smiling "selfie" is just one promising approach among many. Implications By humanizing themselves on social media, scientists may be able to increase trust, public support, and interest in science. Scientists might help viewers individuate them by creating personally curated content with detailed captions, regularly engaging with audiences (asking questions, etc.), and sharing content that highlights "real life" as much as the scientific process. Selfies on Instagram posted by scientists from underrepresented groups may help others better "see" themselves in science and thereby help increase diversity and inclusion in STEM. Schins-kie and colleagues [73] found that college students' stereotypes of scientists are informed by their exposure to minority scientists in real life. During the course of this study, we heard from several educators who were using public #ScientistsWhoSelfie posts in their classrooms. For example, 8th graders at Woodstock Union Middle School in Woodstock, Vermont were challenged to reflect on their existing perceptions of scientists, through a Draw-A-Scientist test, and then explore photos of real scientists posted with the #ScientistsWhoSelfie hashtag on Instagram (R. Becker, personal communication). The students reflected on how their original PLOSONE | https://doi.orq/10.1371/iournal.pone.0216625 May 10, 2019 18/23 ^•PLOSIONE Scientists who selfie drawings compared to the #ScientistsWhoSelfie photos. Many observed that scientists they saw in the photos, versus the ones they had drawn, did not fit a particular mold (e.g. white lab coats) but rather looked and dressed like "normal people" [74]. These anecdotal observations correspond with our own findings that selfies in scientific settings can help shape more positive perceptions of scientists. This is perhaps partly a result of exposure to a diverse set of faces that don't fit traditional stereotypes. This study is among the first to realistically address the usefulness of selfies for science communication. Previous research has explored various obstacles and training needs associated with scientists' use of social media [751. We suggest that scientists' use of Instagram and selfies can positively impact how they are perceived. Our results add to a growing body of evidence that scientists can benefit from using social media. Scientists' online presence can increase the visibility of their research [76], promote collaboration opportunities, improve research efficiency, enhance professional networking, and, in light of our findings, help promote public trust. STEM communities on social media might be able to meaningfully change science stereotypes by taking control of online representations of scientists, humanizing them and bolstering their online visibility. Institutional support for such efforts will be key to making what is good for public perceptions of scientists be good for scientists, too. This is especially true for women and minorities in STEM, who are often subjected to harsh online criticism and yet are pushed into public engagement responsibilities, without any formal reward or recognition. Where public engagement is forced, unwelcome or unrewarded, scientists may also have little motivation to communicate with warmth. Supporting information SI Dataset. Full dataset for Qualtries panel survey experiment. (ZIP) SI Appendix. Supporting information for "Using selfies to challenge public stereotypes of scientists, including Tables A-D. (PDF) Acknowledgments We would like to acknowledge the help of the many scientist Instagrammers who produced stimulus images for this study, including Deboki Chakravarti, Marthe Ludtmann, Melissa Mathers, Diego Balbuena, Nelle Couret, Laura Skates, Charlyn Partridge, Nichole Holm, Nicole Wood, Ev Sotnikova, Raquel Nuno, Kylla Benes, Fernando Mateos-Gonzalez, Luis Miguel Montilla, Juan Swanepoel, Mallory Ladd, Martina Hestericova, Jodie Rummer, Joel Ostblom and colleagues. These individuals were not study participants but were instrumental in helping us design and conduct this study, and were kept informed of results and manuscript preparation. We acknowledge Susan Fiske for informing how we measured stereotypes and for inspiring this research. Author Contributions Conceptualization: Paige Brown Jarreau, Imogene A. Cancellare, Becky J. Carmichael, Lance Porter, Daniel Toker, Samantha Z. Yammine. Data curation: Paige Brown Jarreau. Formal analysis: Paige Brown Jarreau, Samantha Z. Yammine. PLOSONE | https://doi.orq/10.1371/iournal.pone.0216625 May 10, 2019 19/23 ^•PLOSIONE Scientists who selfie Funding acquisition: Paige Brown Jarreau, Imogene A. Cancellare, Becky J. Carmichael, Daniel Toker, Samantha Z. Yammine. Investigation: Paige Brown Jarreau, Imogene A. Cancellare, Becky J. Carmichael, Daniel Toker, Samantha Z. Yammine. Methodology: Paige Brown Jarreau, Imogene A. Cancellare, Becky J. Carmichael, Lance Porter, Daniel Toker, Samantha Z. Yammine. Project administration: Paige Brown Jarreau. Resources: Paige Brown Jarreau, Becky J. Carmichael, Lance Porter. Software: Daniel Toker. Supervision: Paige Brown Jarreau. Validation: Paige Brown Jarreau, Daniel Toker. Visualization: Paige Brown Jarreau. Writing - original draft: Paige Brown Jarreau, Imogene A. Cancellare, Becky J. Carmichael, Lance Porter, Daniel Toker, Samantha Z. Yammine. Writing - review & editing: Paige Brown Jarreau, Imogene A. Cancellare, Becky J. Carmichael, Lance Porter, Daniel Toker, Samantha Z. Yammine. References 1. Leshner Al. Public engagement with science; 2003. 2. Bauer MW, Jensen P. The mobilization of scientists for public engagement. Public Understanding of Science. 2011; 20(1 ):3-11. 3. Dudo A, Besley JC. Scientists' prioritization of communication objectives for public engagement. PloS one. 2016; 11 (2):e0148867. https://doi.orq/10.1371 /journal.pone.0148867 PMID: 26913869 4. Peters HP. Gap between science and media revisited: Scientists as public communicators. Proceedings of the National Academy of Sciences. 2013; 110(Supplement 3): 14102-14109. https://doi.org/10. 1073/pnas. 1212745110 5. NASEM. Communicating science effectively: a research agenda. National Academies Press; 2017. 6. Fiske ST, Taylor SE. Social cognition: From brains to culture. Sage; 2013. 7. Eagly AH, Mladinic A. Are people prejudiced against women? Some answers from research on attitudes, gender stereotypes, and judgments of competence. European review of social psychology. 1994;5(1):1-35. 8. Renn O, Levine D. Credibility and trust in risk communication. In: Communicating risks to the public. Springer; 1991. p. 175-217. 9. Roskos-Ewoldsen DR, Bichsel J, Hoffman K. The influence of accessibility of source likability on persuasion. Journal of Experimental Social Psychology. 2002; 38(2): 137-143. https://doi.org/10.1006/ jesp.2001.1492 10. Brassard D, Nisbet MC. Deference to scientific authority among a low information public: Understanding US opinion on agricultural biotechnology. International Journal of Public Opinion Research. 2007; 19 (1V24-52. https://doi.org/10.1093/ijpor/edl003 11. Rabinovich A, Morton TA, Birney ME. Communicating climate science: The role of perceived communicator's motives. Journal of Environmental Psychology. 2012; 32(1):11—18. https://doi.orq/10.1016/i. jenvp.2011.09.002 12. Fiske ST, Dupree C. Gaining trust as well as respect in communicating to motivated audiences about science topics. Proceedings of the National Academy of Sciences. 2014; 111 (Supplement 4): 13593-13597. https://doi.org/10.1073/pnas.1317505111 13. Besley JC, Dudo A, Yuan S. Scientists' views about communication objectives. Public Understanding of Science. 2018; 27(6):708-730. https://doi.orq/10.1177/0963662517728478 PMID: 28841818 PLOS ONE I https://doi.org/10.1371/iournal.pone.0216625 May 10, 2019 20/23 ^•PLOSIONE Scientists who selfie 14. Bond R, Paterson L. Coming down from the ivory tower? Academics' civic and economic engagement with the community. Oxford Review of Education. 2005; 31 (3):331-351. https://doi.org/10.1080/ 03054980500221934 15. Van Eperen L, Marincola FM. How scientists use social media to communicate their research. Journal of Translational Medicine. 2011; 9(199). https://doi.orq/10.1186/1479-5876-9-199 PMID: 22085450 16. Losh SC. Stereotypes about scientists over time among US adults: 1983 and 2001. Public Understanding of Science. 2010; 19(3):372-382. https://doi.orq/10.1177/0963662508098576 17. Rutjens BT, Heine SJ. The immoral landscape? Scientists are associated with violations of morality. PloS one. 2016; 11 (4):e0152798. https://doi.org/10.1371/iournal.pone.0152798 PMID: 27045849 18. Fiske ST, Cuddy AJ, Glick P, Xu J. A model of (often mixed) stereotype content: Competence and warmth respectively follow from perceived status and competition (2002). Journal of personality and social psychology. 2002; 82(6):878-902. https://doi.Org/10.1037/0022-3514.82.6.878 PMID: 12051578 19. Fiske ST, Cuddy AJ, Glick P. Universal dimensions of social cognition: Warmth and competence. Trends in cognitive sciences. 2007; 11(2):77-83. https://doi.Org/10.1016/i.tics.2006.11.005 PMID: 17188552 20. Wojciszke B, Bazinska R, Jaworski M. On the dominance of moral categories in impression formation. Personality and Social Psychology Bulletin. 1998; 24(12): 1251 -1263. https://doi.orq/10.1177/ 01461672982412001 21. Brambilla M, Rusconi P, Sacchi S, Cherubini P. Looking for honesty: The primary role of morality (vs. sociability and competence) in information gathering. European journal of social psychology. 2011; 41 (2): 135-143. https://doi.orq/10.1002/ejsp.744 22. Siegrist M, Cvetkovich G, Roth C. Salient value similarity, social trust, and risk/benefit perception. Risk analysis. 2000; 20(3):353-362. https://doi.orq/10.1111/0272-4332.203034 PMID: 10949414 23. Kervyn N, Fiske S, Yzerbyt V. Forecasting the primary dimension of social perception. Social Psychology; 2015; 46(1):36-45. https://doi.orq/10.1027/1864-9335/a000219 PMID: 30555596 24. Science and Engineering Indicators 2018. NSB-2018-1. Available from: https://www.nsf.gov/statistics/ indicators/. 25. Weingart P, Muhl C, Pansegrau P. Of power maniacs and unethical geniuses: Science and scientists in fiction film. Public Understanding of Science. 2003; 12(3):279-287. https://doi.Org/10.1177/ 0963662503123006 26. Pansegrau P. 12 Stereotypes and Images of Scientists in Fiction Films. Science images and popular images of the sciences. 2008; 8:257. 27. Dudo A, Brossard D, Shanahan J, Scheufele DA, Morgan M, Signorielli N. Science on television in the 21st century: Recent trends in portrayals and their contributions to public attitudes toward science. Communication Research. 2011; 38(6):754-777. https://doi.Org/10.1177/0093650210384988 28. Moss-Racusin CA, Dovidio JF, Brescoll VL, Graham MJ, Handelsman J. Science faculty's subtle gender biases favor male students. Proceedings of the National Academy of Sciences. 2012; 109 (41): 16474-16479. https://doi.Org/10.1073/pnas.1211286109 29. Chambers DW. Stereotypic images of the scientist: The Draw-a-Scientist Test. Science education. 1983; 67(2):255-265. https://doi.Org/10.1002/sce.3730670213 30. Mau WC. Factors that influence persistence in science and engineering career aspirations. The Career Development Quarterly. 2003; 51 (3):234-243. https://doi.Org/10.1002/J.2161 -0045.2003.tb00604.x 31. Nosek BA, Smyth FL, Sriram N, Lindner NM, DevosT, Ayala A, et al. National differences in gender-science stereotypes predict national sex differences in science and math achievement. Proceedings of the National Academy of Sciences. 2009; 106(26): 10593-10597. https://doi.org/10.1073/pnas. 0809921106 32. Steffens MC, Jelenec P, Noack P. On the leaky math pipeline: Comparing implicit math-gender stereotypes and math withdrawal in female and male children and adolescents. Journal of Educational Psychology. 2010; 102(4):947. https://doi.org/10.1037/a0019920 33. Griffith AL. Persistence of women and minorities in STEM field majors: Is it the school that matters? Economics of Education Review. 2010; 29(6):911-922. https://doi.Org/10.1016/i.econedurev.2010.06. 010 34. Koenig AM, Eagly AH, Mitchell AA, Ristikari T. Are leader stereotypes masculine? A meta-analysis of three research paradigms. Psychological bulletin. 2011; 137(4):616. https://doi.org/10.1037/a0023557 PMID: 21639606 35. Reuben E, Sapienza P, Zingales L. How stereotypes impair women's careers in science. Proceedings of the National Academy of Sciences. 2014; 111 (12):4403-4408. https://doi.Org/10.1073/pnas. 1314788111 PLOS ONE | https://doi.org/10.1371/iournal.pone.0216625 May 10, 2019 21 /23 ^•PLOSIONE Scientists who selfie 36. Galdi S, Cadinu M, Tomasetto C. The roots of stereotype threat: When automatic associations disrupt girls' math performance. Child development. 2014; 85(1):250-263. https://doi.Org/10.1111/cdev. 12128 PMID: 23713580 37. Miller Dl, Eagly AH, Linn MC. Women's representation in science predicts national gender-science stereotypes: Evidence from 66 nations. Journal of Educational Psychology. 2015; 107(3):631. https://doi. orq/10.1037/edu0000005 38. Loverock B, Hart MM. What a scientist looks like: Portraying gender in the scientific media. FACETS. 2018; 3(1):754-763. https://doi.orq/10.1139/facets-2017-0110 39. Long M, Steinke J, Applegate B, Knight Lapinski M, Johnson MJ, Ghosh S. Portrayals of male and female scientists in television programs popular among middle school-age children. Science Communication. 2010; 32(3):356-382. https://doi.Org/10.1177/1075547009357779 40. Eagly AH, Karau SJ. Role congruity theory of prejudice toward female leaders. Psychological review. 2002; 109(3):573. https://doi.Org/10.1037/0033-295X.109.3.573 PMID: 12088246 41. Koenig AM, Eagly AH. Evidence for the social role theory of stereotype content: Observations of groups' roles shape stereotypes. Journal of Personality and Social Psychology. 2014; 107(3):371. https://doi. orq/10.1037/a0037215 PMID: 25133722 42. Banchefsky S, Westfall J, Park B, Judd CM. But you don't look like a scientist!: Women scientists with feminine appearance are deemed less likely to be scientists. Sex Roles. 2016; 75(3-4):95-109. https:// doi.orq/10.1007/s11199-016-0586-1 43. Heilman ME. Description and prescription: How gender stereotypes prevent women's ascent up the organizational ladder. Journal of social issues. 2001; 57(4):657-674. https://doi.Org/10.1111/0022-4537.00234 44. Haynes R. From alchemy to artificial intelligence: Stereotypes of the scientist in Western literature. Public Understanding of Science. 2003; 12(3):243-253. https://doi.Org/10.1177/0963662503123003 45. Long M, Boiarsky G, Thayer G. Gender and racial counter-stereotypes in science education television: A content analysis. Public Understanding of Science. 2001; 10(3):255-269. 46. Chimba M, Kitzinger J. Bimbo or boffin? Women in science: An analysis of media representations and how female scientists negotiate cultural contradictions. Public Understanding of Science. 2010; 19 (5):609-624. https://doi.Org/10.1177/0963662508098580 PMID: 21553601 47. Tetzlaff E, Jago E, Pegoraro A, EgerT. # DistractinglySexy: How Social Media was used as a Counter Narrative on Gender in STEM. In: Proceedings of the 8th International Conference on Social Media & Society. ACM; 2017. 48. Tajfel H. Social psychology of intergroup relations. Annual review of psychology. 1982; 33(1): 1-39. https://doi.orq/10.1146/annurev.ps.33.020182.000245 49. Scherz Z, Oren M. How to change students' images of science and technology. Science Education. 2006; 90(6):965-985. https://doi.Org/10.1002/sce.20159 50. Best Ever Astronaut Selfies; 2013. Available from: https://www.popsci.com/article/science/best-ever-astronaut-selfies. 51. Liu F, Ford D, Parnin C, Dabbish L. Selfies as social movements: Influences on participation and perceived impact on stereotypes. Proceedings of the ACM on Human-Computer Interaction 1. CSCW; 2017:72. 52. Mclntyre K, Paolini S, Hewstone M. Changing people's views of outgroups through individual-to-group generalisation: meta-analytic reviews and theoretical considerations. European Review of Social Psychology. 2016; 27(1 ):63-115. https://doi.Org/10.1080/10463283.2016.1201893 53. Swencionis JK, Fiske ST. More human: Individuation in the 21st century. Humanness and dehumaniza-tion.2014;276-293. 54. Fiske ST, Taylor SE. Social cognition reading. MA: Addison-Wesley; 1984. 55. Mehu M, Little AC, Dunbar Rl. Sex differences in the effect of smiling on social judgments: an evolutionary approach. Journal of Social, Evolutionary, and Cultural Psychology. 2008; 2(3): 103. https://doi.org/ 10.1037/h0099351 56. Hack T. Forming impressions: effects of facial expression and gender stereotypes. Psychological reports. 2014; 114(2):557-571. https://doi.org/10.2466/07.17.PR0.114k17w6 PMID: 24897907 57. Griffin AM, Langlois JH. Stereotype directionality and attractiveness stereotyping: Is beauty good or is ugly bad? Social Cognition. 2006; 24(2): 187-206. https://doi.Org/10.1521/soco.2006.24.2.187 PMID: 17016544 58. Knutson B. Facial expressions of emotion influence interpersonal trait inferences. Journal of Nonverbal Behavior. 1996; 20(3): 165-182. https://doi.org/10.1007/BF02281954 PLOSONE | https://doi.org/10.1371/iournal.pone.0216625 May 10, 2019 22/23 ^•PLOSIONE Scientists who selfie 59. Otta E, Lira BBP, Delevati NM, Cesar OP, Pires CSG. The effect of smiling and of head tilting on person perception. The Journal of psychology. 1994; 128(3):323-331. https://doi.org/10.1080/00223980.1994. 9712736 PMID: 8046666 60. Paulus A, Rohr M, Dotsch R, Wentura D. Positive feeling, negative meaning: Visualizing the mental representations of in-groupand out-group smiles. PloS one. 2016; 11(3):e0151230. https://doi.org/10. 1371 /journal.pone.0151230 PMID: 26963621 61. Dasgupta N, Greenwald AG. On the malleability of automatic attitudes: combating automatic prejudice with images of admired and disliked individuals. Journal of personality and social psychology. 2001; 81 (5):800. https://doi.Org/10.1037/0022-3514.81.5.800 PMID: 11708558 62. Dasgupta N, Asgari S. Seeing is believing: Exposure to counterstereotypic women leaders and its effect on the malleability of automatic gender stereotyping. Journal of Experimental Social Psychology. 2004; 40(5):642-658. https://doi.orq/10.1016/i.iesp.2004.02.003 63. Blair IV, Ma JE, Lenton AP. Imagining stereotypes away: the moderation of implicit stereotypes through mental imagery. Journal of personality and social psychology. 2001; 81 (5):828. https://doi.orq/10.1037/ 0022-3514.81.5.828 PMID: 11708560 64. Ma Z, Sun A, Cong G. On predicting the popularity of newly emerging hashtags in Twitter. Journal of the American Society for Information Science and Technology. 2013; 64(7): 1399-1410. https://doi.org/10. 1002/asi.22844 65. Hayes A.F. Introduction to mediation, moderation, and conditional process analysis: A regression-based approach. Guilford Publications; 2017. 66. Yammine SZ, Liu C, Jarreau PB, Coe IR. Social media for social change in science. Science. 2018; 360 (6385): 162-163. https://doi.org/10.1126/science.aat7303 PMID: 29650665 67. TsukiuraT, Cabeza R. Shared brain activity for aesthetic and moral judgments: implications for the Beauty-is-Good stereotype. Social cognitive and affective neuroscience. 2010; 6(1):138-148. https:// doi.orq/10.1093/scan/nsq025 PMID: 20231177 68. Ma F, Xu F, Luo X. Children's and adults' judgments of facial trustworthiness: the relationship to facial attractiveness. Perceptual and Motor Skills. 2015; 121(1): 179-198. https://doi.org/10.2466/27.22.PMS. 121c10x1 PMID: 26108060 69. Langlois JH, Kalakanis L, Rubenstein AJ, Larson A, Hallam M, Smoot M. Maxims or myths of beauty? A meta-analytic and theoretical review. Psychological bulletin. 2000; 126(3):390. https://doi.org/10.1037/ 0033-2909.126.3.390 PMID: 10825783 70. Eagly AH, Ashmore RD, Makhijani MG, Longo LC. What is beautiful is good, but...: A meta-analytic review of research on the physical attractiveness stereotype. Psychological bulletin. 1991; 110(1): 109. https://doi.orq/10.1037/0033-2909.110.1.109 71. Reis HT, Wilson IM, Monestere C, Bernstein S, Clark K, Seidl E, et al. What is smiling is beautiful and good. European Journal of Social Psychology. 1990; 20(3):259-267. https://doi.org/10.1002/eisp. 2420200307 72. Meraz S. The blogosphere's gender gap: Differences in visibility, popularity, and authority. In Women, men, and news. Ed. by Poindexter P. Routledge; 2008. 73. Schinske JN, Perkins H, Snyder A, Wyer M. Scientist spotlight homework assignments shift students' stereotypes of scientists and enhance science identity in a diverse introductory science class. CBE— Life Sciences Education. 2016; 15(3):ar47. https://doi.orq/10.1187/cbe. 16-01 -0002 PMID: 27587856 74. Jarreau P. Scientists Who Selfie in the Classroom! 2017. Available from: https://experiment.eom/u/ AeHaLA. 75. Bik HM, Goldstein MC. An introduction to social media for scientists. PLoS biology. 2013; 11(4): e1001535. https://doi.org/10.1371/journal.pbio.1001535 PMID: 23630451 76. Liang X, Su LYF, Yeo SK, Scheufele DA, Brassard D, Xenos M, et al. Building Buzz: (Scientists) Communicating Science in New Media Environments. Journalism & Mass Communication Quarterly. 2014; 91 (4):772-791. https://doi.Org/10.1177/1077699014550092 PLOS ONE | https://doi.org/10.1371/iournal.pone.0216625 May 10, 2019 23/23