Abstract
Based on role congruity theory, we investigated how gender bias may influence public attitudes toward the vaccine in Turkey. Using a between-subjects design, we tested whether an emphasis on the female versus the male scientist as the vaccine’s inventor in a news story influenced attitudes about the BioNTech vaccine and vaccination intentions. Partly confirming role congruity theory, three-way interaction results from 665 participants demonstrated that among male participants with a stronger belief in traditional gender roles (compared to males with lower belief), the presence of the female inventor, either by herself or together with the male inventor, decreased the perceived efficacy and safety of the vaccine and reduced intentions to be vaccinated by the BioNTech vaccine. We did not observe such differences for women. These findings highlight how gender bias may influence individuals’ information processing and decision making in a way that may have negative consequences for public health.
How does gender bias influence perceptions of scientific developments and scientists? This is a crucial yet understudied question at a time of increasing recognition of women in science (e.g. Carli et al., 2016; Knobloch-Westerwick and Glynn, 2013). It is essential to understand the effects of media coverage about scientists because such acceptance of scientific developments may have important implications for public health and wellbeing (e.g. Long et al., 2010; Mitchell and McKinnon, 2019). In this study, we focus on the press coverage of the first vaccine against COVID-19 to test whether an emphasis on the gender of the inventors affects the readers’ evaluations of the new vaccine.
Since COVID-19 was declared a pandemic by the World Health Organization (WHO) in March 2020, scientists accelerated their work on vaccines against the disease. In November 2020, the promising third phase results of the COVID-19 vaccine developed by BioNTech and Pfizer (hereafter called the BioNTech vaccine) were announced and received considerable attention in Turkey, partly due to the Turkish origin of the inventors. Immediately, however, critiques noted that the reports in Turkish media outlets published about the BioNTech vaccine overemphasized the role of the male inventor (Dr Uğur Şahin) over the female inventor (Dr Özlem Türeci; Medya Günlüğü, 2020). Given that the Turkish population still subscribes to traditional gender roles, this controversy, while not surprising, gives us a unique opportunity to test how media portrayals of female scientists may have repercussions for public health.
Considering the role that mass media play in the communication of information about COVID-19 related developments to the public, this study relied on the role congruity theory (Eagly and Karau, 2002) to test how such biases in media coverage may influence individuals’ vaccine-related perceptions and behavior. We conducted an experiment to test the impact of the gender of the inventor that a news article emphasizes (female inventor, male inventor, both inventors) on the perceived safety and efficacy of the vaccine and vaccination intentions. Specifically, we examined the influence of gender role beliefs on how individuals reacted to gender-conditioned coverage. Overall, our findings suggest that among men who subscribe to traditional gender roles, exposure to media coverage that emphasizes the presence of a female inventor may lead to a decrease in acceptance of the vaccine.
1. Role congruity theory
As frequently discussed in the literature, women are underrepresented in scientific engagements and academic publishing, especially in science, technology, engineering, and mathematics (STEM) fields (Holman et al., 2018; Wang and Degol, 2017). This underrepresentation has been observed in various media, including newspapers (Shor et al., 2015) and social media (Amarasekara and Grant, 2019). Furthermore, a wealth of research has illustrated that when they are portrayed in media, female scientists are misrepresented either as exceptions or superheroes or by emphasizing their appearances and roles as wives or mothers rather than their expertise (Chimba and Kitzinger, 2010; LaFollette, 1988; Mitchell and McKinnon, 2019; Niemi and Pitkänen, 2017; Shachar, 2000). Such framing on the successful careers of female scientists leads to what has been dubbed as the Matilda effect (Rossiter, 1993), which can be defined as a general tendency to underestimate the achievement of female scientists and attribute their success to their male colleagues (Knobloch-Westerwick et al., 2013).
According to the role congruity theory, a critical mechanism that may explain this effect is the incongruity between functions attributed to women and characteristics attributed to scientists (Eagly and Karau, 2002). Role congruity theory is built on the social roles theory (Eagly, 1987) to explain prejudicial behavior toward female leaders. Accordingly, social roles can be conceptualized as reflections regarding men and women’s respective characteristics and behavioral tendencies in a hierarchically structured society (Eagly, 1987; Eagly and Wood, 1999). Specifically, women are associated with communal characteristics such as warmth, kindness, and being nurturing. In contrast, men are related to agentic qualities such as being assertive, confident, and competent. Role congruity theory broadens social roles theory by focusing on ascriptions about occupation-related roles (Eagly and Karau, 2002; Koenig and Eagly, 2014). For example, communal gender roles are perceived to be incompatible with leadership qualities, which results in prejudicial treatment of female leaders (e.g. negative evaluation of their work). Furthermore, the theory proposes that belief in traditional gender roles would amplify such prejudicial treatment.
Recent studies on public perceptions about scientists suggest that, like leadership, being a scientist is associated with agentic qualities predominantly attributed to men (Banchefsky and Park, 2018; Carli et al., 2016). In a cross-national study conducted in 34 countries (Nosek et al., 2009), the existence of preconceptions associating science with men has been observed to have a mutually reinforcing relationship with gender inequalities in scientific achievement and engagement. For example, in a controlled experiment within which the same manuscript was evaluated by students who were told that the author was either female or male, female authors were evaluated as being less credible. Similarly, the quality of the paper was rated to be lower for women authors (Krawczyk and Smyk, 2016). A similar pattern emerges when citation numbers of scientists and the peer review process are considered. Not only are female scientists less likely to receive favorable peer reviews (Fox and Paine, 2019), but they receive fewer citations to their publications than males (Knobloch-Westerwick and Glynn, 2013). Furthermore, in line with the premise of the role congruity theory, not only are male academics more likely to subscribe to science-gender stereotypes (Banchefsky and Park, 2018), but they also are less likely to cite female scientists (Knobloch-Westerwick and Glynn, 2013).
2. The current study
In this study, we utilize the role congruity theory to test the effects of the news coverage of the development of BioNTech’s COVID-19 vaccine. Because the inventors of the BioNTech vaccine were a married couple of Turkish descent living in Germany, the news of the invention was of particular importance in Turkey. When Turkish scholars and scientists get significant achievements and international recognitions, there is immense media coverage and public attention, as seen with Nobel Prize winners Orhan Pamuk (2006, Literature) and Aziz Sancar (2015, Chemistry). This coverage sets a tone in public perceptions of scientists that are not driven by objective and scientific factors or criteria but rather specific cultural and political dynamics and biases. Gender is one, and probably the most important, of such prejudices. The case of Turkey is particularly pertinent for examining the extent and consequences of gender bias because it is ranked 133rd out of 156 countries in terms of gender gap (World Economic Forum, 2021) and where male-science stereotypes persist (Türkmen, 2008). A more recent survey evaluating the changes in Turkish public’s opinions about violence against women indicates that while progressive views about gender equality are increasing, it is still a minority view in Turkey (Konda, 2020). For instance, the percentage of participants who wanted Turkey to remain in the Istanbul Convention, an international treaty to prevent violence against women, was only 36%. Indeed, the Turkish government has withdrawn from the treaty following overwhelming support from the conservative constituency of the ruling party.
However, progressive segments of the population increasingly work to call out and offset these biases. Shortly after the news about BioNTech’s invention was published, social media users started a campaign criticizing the Turkish mass media for emphasizing the role of the male inventor (Dr Uğur Şahin) over the role of the female inventor (Dr Özlem Türeci), who was often portrayed as “the wife” of Dr Şahin (Medya Günlüğü, 2020). This bias reflects a more general trend both in online searches about the BioNTech vaccine inventors and on posts on Twitter. First, Google Trends data in Turkey between 26 November and 5 December in 2020 (corresponding to the dates when data for this experiment were collected) shows a large difference in the public interest for the two scientists’ names, with a much stronger focus on Dr Şahin (see Supplemental Figure S1). For the same period, there is no such gap between Dr Şahin and Dr Türeci in Germany (see Supplemental Figure S2). Second, analysis of tweets we collected in Turkey as part of a larger project 1 about COVİD-19 also indicates a higher interest in Dr Şahin. Between 26 November and 5 December 2021, Dr Özlem Türeci was mentioned in 455 tweets, while Dr Uğur Şahin was mentioned in 896 tweets (see Supplemental Figure S3).
Specifically, in this study, conducted during the intense media coverage of the couple, we tested whether an emphasis on the female versus the male scientist as the vaccine’s inventor influenced attitudes about the BioNTech vaccine and vaccination intentions. In this respect, this study contributes to the growing body of literature on the consequences of prejudice and stereotypes against women in science.
For this purpose, 1 week after the results of the Phase 3 trials of the BioNTech vaccine were announced on 18 November 2020, we fielded an experiment with three versions of a news report about the BioNTech vaccine based on real news reports. Versions differed from each other in terms of whether they emphasized (1) only the female inventor, (2) only the male inventor, or (3) both of the inventors. We tested the effects of these three versions on the perceived efficacy and safety of the BioNTech vaccine and intentions to be vaccinated with the BioNTech vaccine.
In line with the premise of the role congruity theory, we also examined the moderating role of news readers’ gender and beliefs in traditional gender roles (Knobloch-Westerwick and Glynn, 2013). The research questions and hypotheses listed below were pre-registered before the data collection at Aspredicted.org. 2
RQ1. How does the gender of the vaccine inventor emphasized in the news reports influence the perceived safety and efficacy of the vaccine and vaccination intentions?
H1. Participants who read the news report that emphasize the female inventor will perceive lower safety and efficacy of the vaccine, and express lower vaccination intention than participants who read either the news report emphasizing the male inventor or both inventors.
RQ2. How does the gender of the vaccine inventor emphasized in the news reports interact with news readers’ gender and their beliefs in traditional gender roles in predicting the perceived safety and efficacy of the vaccine and vaccination intentions?
H2. The effects predicted in H1 will be stronger for male participants than female participants.
H3. The effects predicted in H1 will be stronger for the participants who have higher scores for belief in traditional gender roles.
In addition, for reasons that will be discussed in detail in the “Results” section, we also conducted an exploratory analysis of the effects of the three-way interaction between experimental condition, gender of the participants, and beliefs in traditional gender role on the outcome variables.
3. Method
Participants
We conducted an online cross-sectional non-probability survey of the general adult population in Turkey between 26 November and 5 December 2020, about a week after Pfizer’s press release regarding the results of the Phase 3 study of the BioNTech vaccine candidate (Pfizer, 2020). This survey shared time with a larger project on COVID-19 risk perceptions and protective behaviors, which consisted of five independent surveys within a rolling cross-sectional framework. 3 The current experiment was placed in the fifth and final wave. Each new sample was a fresh cross-section of respondents, and there was no other experiment; no carryover effects or attrition bias is expected. We used a quota sample to reflect the demographic characteristics of the general Turkish population in terms of age, education, and sex. In total, 746 participants completed the experiment (51.2% male, Mean Age = 34(SD = 11.2)). Not surprisingly, our online sample underrepresents groups of lower education. Close to half of the respondents (48.9%) had high school education, followed by those who had at least some college education (40.5%). The remaining (9.9%) had less than high school education.
Design
We used a between-subject design within which respondents were randomly exposed to one of three versions of a faux news report we constructed based on contemporary news stories. We created the faux news reports using verbatim sentences from several news articles published in Turkish media about the BioNTech vaccine development. The English translation of the stimulus, which was administered in Turkish, is presented in Supplemental Table S4. The three stories differed in terms of the gender of the inventor of the BioNTech vaccine that was emphasized in the report: (1) only the female inventor (i.e. Dr Özlem Türeci), (2) only the male inventor (i.e. Dr Uğur Şahin), or (3) both of the inventors (i.e. Dr Özlem Türeci and Dr Uğur Şahin). For the manipulation, we only changed the name and the photograph of the inventor across conditions.
After reading the news report, participants were asked to rate the safety of the vaccine (0 = not at all to 10 = extremely safe), efficacy of the vaccine (0 = not at all effective to 10 = very effective), and report their intentions to be vaccinated by the vaccine (0 = never to 4 = definitely) as the outcome variables (this item was adopted from other modules of the larger project, hence we used a five-point scale to maintain comparability with the other modules). We also asked whether they were exposed to similar news for the last few days or not (70% reported reading similar news about the BioNTech vaccine). Finally, as a manipulation check, respondents were asked to report the gender of the inventor that the news report they were exposed to mentioned. The manipulation was effective for most of the participants, χ2 (4, N = 745) = 1055.385, p < .001. We ran all the analyses after eliminating 81 respondents who failed to correctly report the gender of the inventor in the news report they saw, resulting in a sample size of 665. The resulting sample had a mean age of 33.826 (SD = 11.08), 52.9% were male, with a slightly lower proportion of participants having less than high school education (7.8%).
Participants’ traditional gender role beliefs were assessed in the main study before the experimental module with five items (e.g. “due to their nature women cannot be good leaders/managers,” “a man’s job is to earn money; a woman’s job is to look after the home and family”) using a five-point scale (0 = strongly disagree to 4 = strongly agree; Cronbach’s α = .819). Higher scores indicated lower beliefs in traditional gender roles. There was no significant difference on participants’ belief in traditional gender roles (as measured before the experimental manipulation) across the experimental conditions, F(2, 660) = 1.401, p = .247,
4. Results
All the research questions and hypotheses were tested using analysis of covariance (ANCOVA). See Table 1 for the descriptive summaries and the correlations between the study variables.
Descriptive statistics and correlations (n = 665).
The numbers in the diagonal are the means and the standard deviations of the variables.
p < .05; **p < .01; ***p < .01.
Overall, when compared to male participants, female participants provided higher ratings for the safety of the BioNTech vaccine (MFemale = 6.88, MMale = 6.538), and had higher vaccination intentions (MFemale = 2.429, MMale = 2.253), F(1, 631) = 3.723, p = .054,
Similarly, when compared to the participants who were not exposed to similar news before, the ones who were exposed provided higher ratings for the safety (MExposed = 7.161, MNot-exposed = 6.257) and efficacy of the BioNTech vaccine (MExposed = 7.29, MNot-exposed = 6.326), and had higher vaccination intentions (MExposed = 2.508, MNot-exposed = 2.173), F(1, 631) = 21.776, p < .001,
Research question 1: Main effect of inventor’s gender
We tested the effect of the gender of the inventor emphasized in the news report while controlling for participant gender, age, education level, belief in traditional gender roles, prior exposure to similar news. Hypothesis 1 was not confirmed; there was no significant main effect of inventor’s gender on perceived efficacy and safety of the vaccine, and vaccination intention, F(2, 631) = 1.142, p = .32,
Research question 2: Interaction of inventor’s gender with participants’ gender and beliefs in traditional gender roles
We ran two-way ANCOVAs separately for the interactions between the gender of the inventor and gender of the participant and the interactions between the gender of the inventor and beliefs in traditional gender roles. We controlled for age, education, and prior exposure to similar news for both ANCOVAs and added gender as a covariate for the second one.
With respect to the interaction between gender of the inventor and gender of the participant (H2), results showed that there was no significant interaction effect on perceived efficacy and safety of vaccine, and vaccination intention, F(2, 632) = 1.55, p = .213,
To assess the interaction between gender of the inventor and beliefs in traditional gender roles (H3), we first divided participants’ traditional gender role beliefs scores by median (=3) as strong (<3) versus weak belief (⩾3) in traditional gender roles. There was no significant interaction effect of inventor’s gender and participants’ beliefs in traditional gender roles on perceived efficacy and safety of vaccine, and vaccination intention, F(2, 629) = 1.331, p = .265,
Exploratory analysis
Given that male participants reported significantly stronger beliefs in traditional gender roles than female participants (see the “Methods” section), we conducted an additional exploratory analysis. In the light of the possibility that this difference between female and male participants may also influence how they process the reports emphasizing male versus female inventors, we tested the three-way interaction between inventor gender, participant gender, and belief in traditional gender roles. We kept the covariates the same as the aforementioned analyses.
There was a significant three-way interaction predicting perceived vaccine efficacy and vaccination intention, F(2,624) = 2.677, p = .07,
Post hoc simple effects showed that, in the female-only inventor condition, males with strong beliefs in traditional gender roles reported significantly lower perceived vaccine efficacy, safety, and vaccination intentions compared to males with weak beliefs in traditional gender roles, F(1,624) = 6.433, p = .011,
Simple effects of (strong vs weak) beliefs in traditional gender roles for male participants in the female-only inventor and both inventor’s conditions.
SE: standard error.
Boxplots showing the simple effects of (strong vs weak) beliefs in traditional gender roles on (a) perceived vaccine efficacy, (b) perceived vaccine safety, (c) vaccination intention for male participants across three experimental conditions. *p < .05.
5. Discussion
This study aimed to explore how media coverage of vaccine news affects vaccine attitudes and behavioral intentions. Specifically, we investigated if the gender of the inventor(s) emphasized in the news reports affects the perceived safety and efficacy of the vaccine and vaccination intentions. We conducted a survey experiment by manipulating the gender of the emphasized vaccine inventor(s) with faux news reports based on actual news reports about the results of the Phase 3 trial of the BioNTech vaccine. We did not find a significant main effect of the inventor’s gender on outcome variables. We did not find a significant two-way interaction of the inventor’s gender with participants’ gender and beliefs in traditional gender roles.
These findings diverge from the premise of the role congruity theory (Eagly and Karau, 2002) and extant literature (e.g. Knobloch-Westerwick et al., 2013; Krawczyk and Smyk, 2016). The results are in line with a recent experiment conducted in Denmark finding no gender biased perceptions of female experts by male and female audience (Greve-Poulsen et al., 2021). However, given that Denmark and Turkey are at the opposite ends of country-level gender equality index (see World Economic Forum, 2021), we did not expect a similar finding. One reason for this may be the touted impact of the BioNTech vaccine, which might have made it harder for us to capture gender biases in this context. More importantly, prior research indicates that Turkey’s relatively strong cultural norms about women’s role in the family may display more conflicted results than samples from more egalitarian cultures (Caner et al., 2016). In our sample, we observed that male participants were more likely to endorse stronger traditional gender roles than female participants. This difference reflects a more general proclivity in Turkey, whereby men have been shown to have hostile sexism tendencies against women (Glick et al., 2016).
Hence, our follow-up exploratory analyses focused on the combined influence of the experimental condition, gender of the participant, and belief in traditional gender roles. We found that strong versus weak belief in traditional gender roles among male participants had an impact on the outcome variables. Specifically, among male participants with a stronger belief in traditional gender roles (i.e. lower beliefs in gender equality), the presence of the female inventor, either by herself or together with the male inventor, decreased the perceived efficacy and safety of the vaccine and reduced intentions to be vaccinated by the BioNTech vaccine. However, we did not see such differences for women.
These findings support role congruity theory’s premise that women’s work is perceived as less credible due to stereotypes among those who subscribe to traditional gender role beliefs, especially men. Moreover, it is essential to highlight that these findings were observed while a very life-threatening pandemic unfolded. Given that the science-male stereotypes exist even in elementary school students in Turkey (Türkmen, 2008), interventions targeting fundamental changes in gender stereotypes should be addressed to avoid threatening public health consequences. Insofar, as a vital decision about whether to vaccinate may be affected by how individuals react to how science—and more specifically, female scientists—are covered, media institutions will need to assume more responsibility to foster an environment that can challenge the dominant ways of thinking about gender roles. While our results come from a culture where the gender gap is still very high (World Economic Forum, 2021), it is important to note that endorsement of traditional gender norms endures even in the countries that hold the highest ranks in gender equality, such as Finland (Niemi and Pitkänen, 2017). Hence, these findings underscore the importance of further studying how coverage of female scientists may impact routine health-related decisions.
This study was limited in terms of the representativeness of the sample due to the online data collection. Therefore, the generalizability of the findings to the Turkish population is limited. Also, since this study was a part of a larger research project, participants may have experienced fatigue while answering other survey questions before the experiment. As a result, participants’ attention and engagement with the materials may have been low, which may have affected our findings in turn.
To conclude, this study highlighted that gender bias in a society can threaten public health during the COVID-19 pandemic. It enhances our understanding of how the public makes sense of the given scientific information from the media and how the coverage of media affects our attitudes and behavioral intentions. Even though it does not confirm the basic premise of role congruity theory, it partially supported the theory by showing the gender bias of male participants with strong beliefs in traditional gender roles in evaluating vaccines. Considering the media’s role to represent and disseminate gender stereotypes, it is important to be attentive to not discriminate or misrepresent women scientists.
Supplemental Material
sj-pdf-1-pus-10.1177_09636625211060472 – Supplemental material for Biased perceptions against female scientists affect intentions to get vaccinated for COVID-19
Supplemental material, sj-pdf-1-pus-10.1177_09636625211060472 for Biased perceptions against female scientists affect intentions to get vaccinated for COVID-19 by İsminaz Doğan, Lemi Baruh, Zeynep Cemalcilar, Ozan Kuru, Kerem Yıldırım and Ali Çarkoğlu in Public Understanding of Science
Footnotes
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was conducted as a part of a larger research project supported by The Scientific and Technological Research Council of Turkey (TÜBİTAK) (Project number: 120K438).
Supplemental material
Supplemental material for this article is available online.
Notes
Author biographies
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
