Authors’ academic but not personal expertise affects message credibility in science communication: A series of experiments on text perception

Authors’ academic but not personal expertise affects message credibility in science communication: A series of experiments on text perception | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Authors’ academic but not personal expertise affects message credibility in science communication: A series of experiments on text perception Annalena Ulsperger, Lara Pfannenschwarz, Clara Schetla, Karen Poletilo, and 3 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9010835/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 11 You are reading this latest preprint version Abstract Evaluating message credibility is critical for understanding scientific information and informed decision making. People’s ability to assess the credibility of science communication texts is influenced by several factors. In three experiments, we examined the impact of authors’ gender as well as different types of expertise on readers’ credibility perceptions. In Experiment 1 (n = 203), we varied authors’ alleged gender and expertise. We found that a text allegedly written by a high-expertise author was rated as more credible than the same text by a low-expertise author. There was no effect of authors’ gender. Experiment 2 (n = 182) was a replication of the first study with a different sample. We did not find any effects of gender or expertise on perceived message credibility. In Experiment 3 (n = 206), we differentiated between academic and personal expertise and manipulated these types of expertise independently. The data indicated that academic expertise was a significant positive predictor of perceived message credibility, while personal expertise was not. We discuss these findings in terms of their significance for the trustworthiness of science communication and examine the relevance of differentiating the concept of expertise in this context. credibility expertise science communication academic expertise personal expertise Figures Figure 1 1. Introduction Science communication plays an important role in everyday information transfer and can be a relevant basis for laypeople’s decision making 1 . However, this accessibility of information also entails significant risks, such as the spread of misinformation, since the verification of the content being disseminated is removed from the control of traditional gatekeepers 2 , 3 . Instead, the responsibility for verifying the credibility of the offered content increasingly falls onto the consumers themselves. Therefore, it is essential that people understand current science-related topics and are able to critically assess the wide range of information available 4 – 6 . Since laypeople often lack the specific subject-matter knowledge to reliably judge the factual accuracy of every content they encounter, they tend to consult other factors that can be used to assess the credibility of information 5 , 6 . These can be characteristics of the content itself like its fluency, channel characteristics like the media modality, source characteristics like the expertise of an author, or even characteristics of the recipients 7 , as credibility is subjectively perceived by message receivers 8 . In the research presented here, we focus on the potential influence of author characteristics on laypeople’s subjective perception of message credibility. In the following paragraphs, we first introduce the concept of message credibility and its role in science communication before we present expertise and gender as two author characteristics that may influence message credibility perceptions. 1.1 Message credibility Message credibility refers to people’s subjective perceptions of the integrity, reliability, and trustworthiness of a given content, such as a text or other content presentation 9 , 10 . The credibility of a message describes the belief of receivers that the provided information is true, approvable, and trustworthy 11 , 12 . This belief strongly impacts the acceptance of the information provided and its effect on the receivers beliefs and attitudes, with more credible content being more influential than less- and non-credible content 13 – 16 . Various subfactors have been used to conceptualize and quantify the perceived credibility of a message 9 . One important facet of message credibility is its trustworthiness 17 . Trustworthiness means that a message is perceived to be reliable and believable, and that its recipients presume a high level of veracity of that message 18 . Other relevant dimensions of credibility include the perceived truth of and the agreement with a message. Truth comprises the perceived accuracy of the provided statements 19 . A greement reflects the overlap of a receiver's opinion with the message, which would result in approving the message and consenting to it 20 . 1.2 Expertise Source expertise has been shown to influence message credibility in various contexts, like in online marketing communications 21 , 22 , online health information 23 – 25 , or political information on social media platforms 26 , 27 . Expertise refers to the domain-specific knowledge and ability of a source, for example the author of a text, to provide accurate information in said domain developed over time 8 , 28 , 29 . Established expertise indicators are, for example, academic qualifications and credentials, as seen in health practitioners 18 , 30 . In general, the literature shows that experts are considered to be more credible and persuasive than non-experts 25 , 26 : For example, Eastin 23 found that study participants were more likely to believe experts when it came to health information if they themselves had no prior knowledge of the subject. Jung et al. 24 found similar results for participants with low prior knowledge in the field of online diet and nutrition information credibility. Zimmermann et al. 27 showed that political news on Instagram were rated higher in credibility when higher source expertise was indicated. And Meinert and Krämer 26 showed that decision times between two sources were decreased when expertise cues were present, with expert sources being ascribed more credibility and chosen more often. 1.3 Gender Another less obvious factor influencing the credibility of a text is the gender of its author. While historically female experts have been perceived as less competent than their male counterparts in science communication 31 , 32 , contemporary literature indicates a more equalized perception 33 – 35 , or even an advantage of female experts compared to males 33 , 36 . For example, female sports reporters covering women's sports were rated as more likable and credible than male reporters, especially by female study participants 37 . Bundi et al. 33 found similar preferences by their female participants for female experts in climate change communication. Jones and Mitchell 38 even found that female experts were rated as slightly more credible in traditionally male-dominated fields such as security policy. Taken together, prior research suggests that expertise and gender as author characteristics may have an impact on readers’ credibility perceptions regarding science communication texts. While the expertise effect is rather established, the effect of an author’s gender is subject to change. However, empirical evidence on how these factors interact in the current science communication climate remains limited. To address this research gap, we conducted a series of three studies examining the role of the authors’ characteristics in perceived credibility of a science communication article. 2. Study 1 In the first study, we investigated whether the alleged expertise and gender of an author affected the perceived credibility of a science communication text. Based on the literature review and the considerations presented above, we formulated the following hypotheses: H1: Scientific texts by authors with high expertise are attributed greater credibility compared to texts by authors with low expertise. H2: Scientific texts by female authors are attributed greater credibility compared to texts by male authors. Furthermore, we investigated as an exploratory research question based on Bundi et al. (2025) whether there are empirical differences based on the gender of the participants (ERQ1). We also explored whether there is an interaction effect between the variables gender and expertise (ERQ2). 2.1. Methods All three studies reported in this paper were preregistered on AsPredicted prior to data collection (Study 1: https://aspredicted.org/cjcj-rc3b.pdf, Study 2: https://aspredicted.org/nyrs-tm7s.pdf, Study3: https://aspredicted.org/tbdk-9mx8.pdf) and approved by the Ethics Committee of the Leibniz-Institut für Wissensmedien (LEK 2024/054). Participants provided written informed consent before and after participation. Participation was anonymous and participants were informed beforehand that they had the right to discontinue the study and withdraw their data at any time during the participation. All three studies were conducted in German. The experiments were performed in accordance with relevant guidelines and regulations. 2.1.1. Participants We conducted Study 1 as an online experiment using SoSci Survey (www.soscisurvey.de; Version 3.6.10) 39 . In a 2x2 between-groups design, we varied the source factors expertise (high vs. low) and gender (male vs. female) of the alleged author of a science communication text by means of an author description. Based on an a priori power analysis (GPower, version 3.1.9.7), we aimed to recruit at least 199 participants to achieve 80% power for detecting a small effect (f=0.2, α =.05). N=234 participants completed the study, of which 31 had to be excluded due to the predefined exclusion criteria of high prior knowledge in the topic of the text or a failed attention check. Participants were recruited via the e-mail distribution list of a German university and given the option to participate in a gift-card raffle as compensation. They had to be at least 16 years old and have at least a C1 level of proficiency in German. The demographic data of the samples of all three studies are shown in Table 1. 2.1.2. Procedure First, participants filled in the informed consent form. They were told that they would be asked about their assessment of a text. Afterwards, the participants were presented with one of four texts introducing an author who allegedly had written a science communication text about natural disaster preparedness, which the participants would read afterwards. Participants were distributed equally among the four conditions ( n high exp., fem. =52, n high exp., male =50, n low exp., fem. =50, n low exp., male =51). After ten seconds the participants could move on to the next page, where they were presented with a text concerning natural disaster preparedness. The text was the same across all four conditions and was shown for at least 15 seconds before participants could move on. Thereafter, they were asked to judge the credibility of the text. Participants were then asked to determine the gender of the alleged author (attention check), the expertise level of the author (manipulation check), and their prior knowledge about natural disaster preparedness. Additionally, participants were directly asked whether they perceived men, women, or both as more credible in the topic. Then the participants were asked about their demographics (gender, age, and educational background) and given the option to participate in the raffle. Finally, they were debriefed and dismissed. 2.1.3. Materials and Measures The independent variables author expertise and author gender were manipulated via the author description. Expertise was presented as either high (longstanding academic career and international consulting experience with respect to the topic) or low (no academic background, only some personal interest due to previous flood experience in the author’s hometown). Gender was indicated through title, name, and pronouns as either male or female. All four versions of these stimuli are provided in the Supplementary Material (S1). The science communication text was introduced as describing the author’s views on how sustainable and ecological methods can provide optimal protection against natural disasters. After briefly outlining the topic, the text claimed that nature-based measures alone were sufficient to prevent natural disasters. We created the text using specific prompts in ChatGPT-4 (OpenAI; Version 4.0) 40 . See Supplementary Material for the full text (S2). We assessed message credibility using 11 items (α=.94) answered on a 7-point Likert scale (1 being “not at all” and 7 being “fully applies”). The scale comprised three dimensions: trustworthiness (4 items), truth (3 items), and agreement (4 items). The items for both the trustworthiness dimension and two of the items of the agreement dimension were adapted to the present context from Klebolte 20 . The other two items of the agreement dimension were added by the authors. The truth dimension was extracted from Mehlis 19 . The message credibility score was calculated as the mean of all items. The complete scale is provided in the Supplementary Material (S3). The question “Does the author have expertise in the subject area?” answered on a 7-point Likert scale (1=“not at all” and 7=“fully applies”) was used as a manipulation check. To measure prior knowledge, we asked the following question: “Did you already have prior knowledge in the field of natural disaster preparedness?” Participants answered on a 7-point Likert scale (1=“no prior knowledge at all” and 7=“extremely high prior knowledge”). Participants indicating 6 or 7 on the scale were excluded from the analysis As an attention check, we asked “Was the author a man or a woman?” with the options: “male”, “female”, or “I don’t remember”. Participants who did not indicate the correct answer were excluded. To ensure that the topic of the text itself was not tied to a specific gender, we asked “Do you consider men or women to be more credible in the field of preventive natural disaster preparedness?” with the answering options “male”, “female”, or “both equally”. 2.2. Results We used a Welch t-test to check the manipulation. It showed that the perceived expertise differed significantly between the high and the low-expertise conditions, t (162.73)=-13.19, p <.001, d =1.86. As the authors in the high-expertise conditions were perceived as having significantly more expertise ( M =5.97, SD =0.96) than the authors in the low-expertise conditions ( M =3.51, SD =1.61), the manipulation was considered as successful. Furthermore, the topic of the science communication text was successfully considered gender-neutral, since most participants answered the corresponding question with “both equally” ( n both =172, n women =18, n men =13). To test the hypotheses that authors with high expertise (H1) and female authors (H2) would be attributed greater credibility, we conducted a two-way analysis of variance (ANOVA) with the between-groups factors expertise (high/low) and gender (female/male) of the author as independent variables and message credibility as the dependent variable. The ANOVA revealed a significant main effect of author expertise, F (1, 199)=5.38, p =.021, partial η ² =.03. The results indicate that the text by authors with alleged high expertise was rated as more credible ( M =4.85, SD =1.09) than the identical text by authors with alleged low expertise ( M =4.47, SD =1.27), supporting H1. There was no main effect of gender, F (1, 199)=0.10, p =.757, therefore rejecting H2. Regarding ERQ2, there was no significant interaction effect between the factors author expertise and author gender, F (1, 199)=0.06, p =.809. To examine whether there was an empirical difference based on the gender of the participants (ERQ1), the analysis was expanded to a three-way ANOVA with participant gender as an additional between-groups factor. The main effect of author expertise remained significant, F (1, 193)=5.38, p =.021, partial η² =.03, while the main effects of author gender and participant gender were not significant, F s.331. Neither of the two-way interaction, F s.694, nor the three-way interaction, F (1, 193)=2.75, p =.099, reached statistical significance. 2.3. Discussion Our findings support a positive effect of expertise on perceived message credibility in line with previous literature 23,27 . However, the observed effect was small and even the low-expertise condition received rather high credibility evaluations. Contrary to our hypotheses and contemporary studies reporting a preference for female over male experts 33,37 , we did not find differences in credibility perceptions between male and female authors. We also observed no differences in credibility ratings based on the participants’ own gender. Instead, the results align with more recent literature suggesting a shift toward gender-independent credibility evaluations in science 33–35 , indicating a societal shift toward gender neutrality when evaluating experts. This might have been further promoted by our choice of a gender-neutral topic, which possibly minimized the salience of the author’s gender. Moreover, the lack of a gender effect might also reflect the specific characteristics of our sample. Egalitarian views and social desirability bias might be more prominent in a sample of university students. 3. Study 2 To assess the generalizability of the findings of Study 1, we conducted a replication study using a sample that was more representative of the general population in terms of age and education. Since we did not observe an effect of the authors’ gender in Study 1, which is in accordance with more recent literature 33–35 , we assumed to find only a main effect of expertise in the second study: H1. Scientific texts allegedly written by authors with high expertise are attributed greater credibility compared to texts by authors with low expertise. We were, again, interested in whether there were empirical differences based on the gender of the participants (ERQ1) and whether there was an interaction effect between expertise and gender of the author (ERQ2). Whether there would be a difference in credibility perceptions of the text based on the alleged gender of the author with this broader sample was also only investigated as an exploratory research question (ERQ3). 3.1. Methods 3.1.1. Participants The same between-groups design and exclusion criteria were used as in Study 1. However, the online experiment was hosted on Qualtrics (Qualtrics, Provo, UT). To achieve a sample more representative of the general population, participants were recruited via the online crowdsourcing platform Clickworker , each receiving monetary compensation of 2,14€ for completing the study. 219 participants took part in the study of which 37 had to be excluded due to the predefined exclusion criteria. In contrast to Study 1, this sample included more male than female participants, the participants were, on average, older than those in Study 1, and the educational background was more diverse with more participants indicating a completed apprenticeship, secondary school, and technical college as their highest degree. The demographic data of Study 2 is presented in Table 1. 3.1.2. Procedure The procedure was identical to Study 1. The participants were distributed among the four conditions as follows: n high exp., fem. =41, n high exp., male =45, n low exp., fem. =47, n low exp., male =49. 3.1.3. Materials and Measures The same materials and measures were used as in Study 1. The internal consistency of the message credibility scale was, again, excellent with α=.95. 3.2. Results Again, a Welch t-test confirmed that the manipulation was successful, t (138.10)=-13.05, p <.001, d =1.88. Perceived expertise differed significantly between the conditions, with authors in the high-expertise conditions being perceived as having significantly more expertise ( M =6.13, SD =0.76) than the authors in the low-expertise conditions ( M =3.71, SD =1.63). Participants also indicated that they again did not prefer one gender over the other in the context of preventive natural disaster preparedness ( n both =166, n women =9, n men =7). As in Study 1, we used an ANOVA with the between-groups factors expertise (high/low) and gender (female/male) of the author as independent variables and message credibility score as the dependent variable to test for the hypothesized main effect of expertise (H1), as well as to answer ERQ2 and ERQ3. The two-way ANOVA showed neither a main effect of author expertise, F (1, 178)=0.24, p =.626, nor a main effect of gender, F (1, 178)=0.19, p =.662, or an interaction between the factors, F (1, 178)=3.39, p =.067. Therefore, H1 was not supported. We also found no differences regarding ERQ2 and ERQ3. To examine whether participants’ own gender influenced credibility ratings (ERQ1), an additional three-way ANOVA with participant gender as an additional between-groups factor was conducted. None of the main effects, F s.434, neither of the two-way interactions, F s.074, nor the three-way interaction, F (1, 172)=0.42, p =.520, reached significance. 3.3. Discussion Study 2 was conducted to evaluate whether the effects observed in Study 1 would generalize to a more heterogeneous sample. For the gender of the author and the gender of the participants, we found again no significant effect. This is consistent with Study 1 and strengthens the conclusion that author and participant gender do not influence participants’ credibility perception under our design conditions. Thus, neither author nor participant gender were further investigated in the subsequent study. The main effect of the authors’ expertise could not be replicated. Whereas Study 1 revealed a significant main effect of author expertise on perceived message credibility, Study 2 showed no such effect in the sample with larger demographic variability. This suggests a dependence of the effect observed in Study 1 on characteristics of the student sample. However, expertise is widely regarded as an established indicator of perceived credibility across various samples in the literature 23–27 , wherefore this explanation seems unlikely. A closer look reveals the study material as a plausible source for the inconsistent findings: While the low-expertise condition introduced the alleged author as being new to the topic at hand, they were still ascribed previous real-life experience and personal interest. Expertise is not a homogeneous construct, but can encompass various forms of knowledge, skills, and abilities all perceived as expertise in real-life communication practice 28,41 . Specifically, it can be differentiated based on what grounds people are deemed to be experts: Weinstein 42 differentiates between epistemic expertise, which is based on theoretical knowledge, and performative expertise, which is based on practical skills. Similarly, Wagemans 43 also proposes a dichotomy between expert opinion based on professional knowledge acquired through professional training and experiential knowledge gained through personal experience. Following this breakdown of the concept of expertise, the author descriptions used in Study 1 and 2 possibly entailed an unintended dichotomy between professional academic expertise, and experience-based expertise. As different types of expertise may affect credibility perceptions to differing extents, this blend in our study material might have led to inconsistent results. 4. Study 3 Therefore, we investigated whether there were differences in the effect of expertise on credibility judgments of science communication texts based on the type of expertise. This provides an important conceptual contribution to expertise research and may disentangle the conflicting results of Study 1 and 2. Previous research on source expertise often combined these expertise indicators 18 and experimental research systematically differentiating between different forms of expertise and their impact on perceived message credibility is sparse. Building on the differing effects of normatively strong vs normatively weak evidence on persuasion 44 , Burgers et al. 45 empirically demonstrated the importance of this distinction. Their study found that the perceived expertise of technical experts (based on systematic professional knowledge) was significantly higher than that of experiential experts (based on practical experience), even though the perceived trustworthiness was roughly the same. In addition, the statements made by authors with professional knowledge were significantly more persuasive than those made by authors with experiential knowledge. Ferreira and Wingrove 46 made a similar distinction between expert training and expert experience, showing that both factors had an independent impact on perceived credibility of forensic experts. We deduced that there might have been similar effects of the two forms of author expertise on perceived message credibility in our studies. The author descriptions in Studies 1 and 2 contained professional academic as well as personal experiential expertise as described in Wagemans 43 and Burgers et al. 45 . Academic expertise was based on formalized knowledge, academic training, and institutional recognition. Personal expertise was based on practical knowledge acquired through direct engagement with the topic. In Study 3, we explicitly differentiated between the two types of expertise and investigated whether there are differences in credibility judgments. We formulated the following hypotheses: H3. Main effect of academic expertise: A scientific text is perceived as more credible when allegedly written by an author with high academic expertise compared to an author with no academic expertise. H4. Main effect of personal expertise: A scientific text is perceived as more credible when allegedly written by an author with high personal expertise compared to an author with no personal expertise. H5. Combined effect: A scientific text is perceived as significantly more credible when allegedly written by an author with both forms of expertise (high academic and high personal expertise) compared to authors with only one form of expertise (either academic or personal). Furthermore, we investigated in two exploratory research questions whether there is an interaction effect between the alleged academic expertise and personal expertise of an author of a scientific text (ERQ4), as well as whether there is a difference in the perception of credibility of scientific texts allegedly written by an author with exclusively academic expertise compared to an author with exclusively personal expertise (ERQ5). 4.1. Methods 4.1.1. Participants Study 3 was conducted as an online experiment using SoSci Survey (www.soscisurvey.de; Version 3.6.10) 39 . It contained a 2x2 between-groups design with academic expertise (high vs. none) and personal expertise (high vs. none) of the alleged author of a science communication text being varied. Of the 268 participants who had completed the study, 62 had to be excluded due to the predefined exclusion criteria. Like in Studies 1 and 2, participants were excluded based on high prior knowledge in the topic (indicating 6 or 7 on a 7-point Likert scale) or a failed attention check. Participants were recruited via Prolific (www.prolific.com), an online participant recruitment platform. They received monetary compensation of 1,20€ for completing the study. They had to confirm that they were at least 16 years of age and had at least a C1 level of proficiency in German. For further demographic information see Table 1. 4.1.2. Procedure Participants were informed that the goal of the study was to investigate the perception of scientific texts. After completing the informed consent, the participants were instructed to first read one of four descriptions of an author of a scientific text and then the text itself. They had a minimal reading time of 15 seconds for the author description and minimally 60 seconds for the scientific text. Participants were distributed equally among the four conditions ( n dual exp. =46, n acad. exp. =48, n pers. exp. =51, n no exp.= 61). On the following page, the manipulation check was conducted using two questions regarding the degree of academic and personal experience of the author in the subject area. Then the participants were presented with the same science communication text used in Studies 1 and 2. Afterwards, the participants indicated their perceived credibility of the text. They were also asked to indicate their prior knowledge about natural disaster preparedness. The study concluded with demographic questions. 4.1.3. Materials and Measures Participants were presented with author descriptions reflecting four expertise conditions: (1) dual expertise (high academic and high personal expertise), (2) high academic expertise only, (3) high personal expertise only, and (4) no expertise. The descriptions were based on those of Study 1 and 2 but adapted to more concisely separate the types of expertise in the different conditions. Academic expertise was signaled by a doctorate, an extensive academic career, and international consulting experience. Personal expertise was conveyed through firsthand experience with a major flood and substantial practical engagement in flood prevention. In the no-expertise condition, the author was described as a newcomer to the topic with minimal prior involvement. We created the descriptions using ChatGPT-4 (OpenAI; Version 4.0) 40 . The stimuli are provided in the Supplementary Material (S4). The same science communication text was used as in Studies 1 and 2. Message credibility was measured using the same 11-item scale as in Studies 1 and 2 (α=.96). As a manipulation check, the participants independently answered the two questions “How much academic experience do you think the author has in the subject area?" and "How much personal experience do you think the author has in the subject area?” on 7-point Likert scales (1=“no experience” and 7=“extensive experience”). To measure prior knowledge, we asked the same question as in the previous studies, also answered on a 7-point Likert scale. However, the labeling of the ends was switched to 1=“not at all” and 7=“fully applies”. Participants indicating 6 or 7 on the scale were, again, excluded from the analysis 4.2. Results The statistical analysis was conducted using R (Version 4.3.0) 47 . The successful manipulation of both the academic and personal expertise was confirmed using two Welch’s t-tests. For academic expertise, t (200.12)=21.56, p <.001, d =3.02, the conditions with high academic expertise were regarded significantly higher in academic experience ( M =5.85, SD =1.27) than the conditions without academic expertise ( M =1.96, SD =1.32). For personal expertise, t (160.56)=16.92, p <.001, d =2.29, the conditions with high personal expertise were regarded significantly higher in personal experience ( M =6.27, SD =0.90) than the conditions without personal expertise ( M =2.95, SD =1.86). To understand whether there was a main effect of academic expertise (H3), a main effect of personal expertise (H4), or an interaction between the two (ERQ4), a linear model was used. The overall model was significant, F (3, 202)=3.65, p =.013, R ²=.05. Academic expertise was a significant positive predictor of perceived message credibility (β=0.52, p =.002), whereas personal expertise (β=-0.12, p =.459) and the interaction (β=0.21, p =.524) were not significant. This supports H3, while H4 could not be supported and ERQ4 was negated. The credibility values of all four conditions are depicted in Figure 1. To investigate whether there was a combined effect of the types of expertise (H5) and a difference based on the types of expertise presented exclusively (ERQ5), we fitted a linear model with contrast coding (CH5, CE5, R1). Contrast coding was applied as follows: CH5=[0, -0.5, -0.5, 1], CE5=[0, -0.5, 0.5, 0], R1=[-0.75, 0.25, 0.25, 0.25], with [1, 2, 3, 4] referring to the experimental conditions 1) no expertise, 2) no academic / high personal expertise, 3) high academic / no personal expertise, and 4) high academic / high personal expertise. The overall model was significant, F (3, 202)=3.65, p =.013, R ²=.05. The contrast CH5, representing the contrast between dual and the average of the singular types of expertise (H5), was not significant (β=0.20, p =.144). Therefore, H5 could not be supported. However, the contrast CE5, which directly compared exclusively academic with exclusively personal expertise (ERQ5), was a significant positive predictor of credibility (β=0.64, p =.007), answering ERQ5 affirmatively. The residual contrast R1, which compared all three expertise conditions with the control condition of no expertise, was not significant (β=0.20, p =.266). 4.3. Discussion Study 3 showed a significant main effect of academic expertise, indicating higher perceived credibility for texts written by sources with high academic expertise than sources with no academic expertise. Academic-only sources were also rated more credible than personal-only sources when only one type of expertise was present. The significance of academic expertise is in accordance with literature indicating effects of expertise on credibility perception 23,27 . Furthermore, it is partially in accordance with the findings by Burgers et al. 45 and Ferreira and Wingrove 46 , who showed increased persuasiveness of people with professional compared to personal expertise and independent effects of professional expertise on credibility assessments, respectively. 5. General discussion Correctly judging the credibility of science communication texts without proper expertise in the field is an essential skill for laypeople. Besides the content, other properties of the communication environment, like information about the author, can be used to approximate message credibility. Therefore, it is important to critically regard the impact of author features on the subjective perception of messages in science communication. In three experimental studies, we investigated whether the alleged expertise and gender of an author impact the perceived message credibility of a gender-neutral science communication text. In terms of author gender, Study 1 and 2 showed no significant effects of author gender on perceived message credibility. The participants’ own gender was also no moderator of the credibility judgments. We interpreted the results of Study 1 and 2 as supporting the current trend toward gender-independent credibility assessments of science communication 33 – 35 . While Study 1 found an effect of author expertise on message credibility with alleged high author expertise leading to increased credibility ratings in a student sample, Study 2 failed to replicate this effect with a more diverse sample. While labelling the conditions in Studies 1 and 2 as high or low in expertise, the conditions comprised a mixture of academic and personal expertise. Strictly differentiating between the two concepts, we investigated their specific impact on message credibility in Study 3. We found that only academic, but not personal expertise had a significant effect on perceived message credibility. We also did not find additive or interactive effects of both types of expertise. These results are only partially in accordance with the literature, which indicates independent effects for both academic and personal expertise 45 , 46 . In accordance with the Elaboration Likelihood Model 48 , 49 and the Unifying Framework of Credibility Assessment 50 , we assume that the presented science communication text was judged based on heuristics activated by the author descriptions, which are more favorable to established cues like academic credentials and thereby reduce processing effort 51 – 53 . This is supported by the consideration that the text itself was probably not personally relevant to the participants and was thus processed with low effort 51 , 54 , 55 . The symbolic power of academic titles probably carries specific weight in relation to the results of Study 1. Academic titles represent institutional validation by the scientific system 56 , 57 and the symbolic power of academic titles activates culturally embedded associations with authority 20 , 58 . Therefore, academic expertise specifically acts as an established signal of credibility through institutional legitimation and formalized knowledge structures 25 , 56 , 59 . In contrast, experience-based expertise represents an alternative form of knowledge legitimation through practical provenance 43 . The concept of a recognition heuristic 51 suggests that the familiarity of academic titles can act as a signal of credibility. University students in particular are embedded in the academic realm and potentially used to accepting academic credentials as a sign of epistemic authority 60 , 61 . The incomplete separation of academic and personal expertise may have been processed more variably by the sample of Study 2, which might be more removed from academia. This could have potentially led to the non-significant findings in Study 2, while the student sample probably focused more on the academic credentials as a signal of epistemic authority, leading to the significant findings in Study 1. However, the results have to be considered in the context of the study environment. The present studies used isolated online settings with controlled text presentations, whereas science communication often takes place in more complex, interactive media environments. Zimmermann et al. 27 , for example, demonstrated that on Instagram, political expertise significantly increased perceived credibility, while traditional source types had little influence. This highlights how much the context of the medium can influence expertise effects. Therefore, future studies should examine the robustness of the effects of different types of expertise in different digital media and communication environments. 6. Conclusion Our findings support the notion that an author’s academic credentials provide a decisive advantage regarding the perceived credibility of a science communication text, while personal expertise does not add credibility in this context. The clear hierarchy between different forms of expertise confirms that not all expertise signals are equally effective. This clarifies previous assumptions about the effects of expertise on credibility perceptions and shows that the type of expertise, and not its presence in general, is crucial in science communication. Declarations Acknowledgements / Funding statement This research is part of the Metavorhaben “Digitalisierung im Bildungsbereich“ (Digi-EBF ii) and was funded by the BMBFSFJ. Author contributions In accordance with the CRediT taxonomy, the authors contributed to this paper as follows: A.U.: conceptualization, methodology, validation, formal analysis, investigation, data curation, writing (original draft), visualization, project administration. L.P.: conceptualization, formal analysis, investigation, data curation, writing (review and editing ) . C.S.: conceptualization, methodology, formal analysis, investigation, data curation, writing (review and editing). K.P.: conceptualization, methodology, formal analysis, investigation, data curation, writing (review and editing). L.S.: conceptualization, methodology, formal analysis, investigation, data curation, writing (review and editing). M.E.: conceptualization, methodology, formal analysis, investigation, data curation, writing (review and editing). J.K.: conceptualization, methodology, resources, writing (review and editing), supervision, project administration, funding acquisition. The authors all read and approved this manuscript. Additional Information Competing Interests Statement: The author(s) declare no competing interests. Data Availability The datasets generated and analyzed during the current study are available in the Open Science Framework, https://osf.io/qjts4/overview?view_only=508b142e9ea54a08b9791056df455cf7 References Fischhoff, B. Evaluating science communication. Proc. Natl. Acad. Sci. USA . 116 , 7670–7675 (2019). Höttecke, D. & Allchin, D. Reconceptualizing nature-of-science education in the age of social media. Sci. Educ. 104 , 641–666 (2020). Pantic, M. & Ziek, P. Gatekeeping in a Digital Media Habitat: The Role of Secondary Gatekeepers. Electron. News . 19 , 94–110 (2025). Krause, N. M., Freiling, I. & Scheufele, D. A. Our changing information ecosystem for science and why it matters for effective science communication. Proc. Natl. Acad. Sci. 122, e2400928121 (2025). Osborne, J. & Pimentel, D. Science education in an age of misinformation. Sci. Educ. 107 , 553–571 (2023). Scharrer, L., Thomm, E., Stadtler, M. & Bromme, R. What Makes Sources Credible? How Source Features Shape Evaluation of Scientific Information. J. Exp. Educ. 0 , 1–24 (2025). Ou, M. & Ho, S. S. Factors Associated With Information Credibility Perceptions: A Meta-Analysis. J. Mass. Commun. Q. 101 , 346–372 (2024). Tseng, S. & Fogg, B. J. Credibility and computing technology. Commun. ACM . 42 , 39–44 (1999). Hanimann, A., Heimann, A., Hellmueller, L. & Trilling, D. Believing in Credibility Measures: Reviewing Credibility Measures in Media Research From 1951 to 2018. Int. J. Commun. 17 , 214–235 (2023). Metzger, M. J., Flanagin, A. J., Eyal, K., Lemus, D. R. & Mccann, R. M. Credibility for the 21st Century: Integrating Perspectives on Source, Message, and Media Credibility in the Contemporary Media Environment. Ann. Int. Commun. Assoc. 27 , 293–335 (2003). Appelman, A. & Sundar, S. S. Measuring Message Credibility: Construction and Validation of an Exclusive Scale. J. Mass. Commun. Q. 93 , 59–79 (2016). Flanagin, A. J. & Metzger, M. J. The role of site features, user attributes, and information verification behaviors on the perceived credibility of web-based information. New. Media Soc. 9 , 319–342 (2007). Anna) Kim, E., Shoenberger, H., Kwon, E. & Ratneshwar, S. (eds) (,(Penny) A narrative approach for overcoming the message credibility problem in green advertising. J. Bus. Res. 147, 449–461 (2022). Clark, J. K., Wegener, D. T., Sawicki, V., Petty, R. E. & Briñol, P. Evaluating the Message or the Messenger? Implications for Self-Validation in Persuasion. Pers. Soc. Psychol. Bull. 39 , 1571–1584 (2013). Shamim, K. & Islam, T. Digital influencer marketing: How message credibility and media credibility affect trust and impulsive buying. J. Glob Sch. Mark. Sci. 32 , 601–626 (2022). Slater, M. D. & Rouner, D. How Message Evaluation and Source Attributes May Influence Credibility Assessment and Belief Change. J. Mass. Commun. Q. 73 , 974–991 (1996). Winter, S. & Krämer, N. C. A question of credibility – Effects of source cues and recommendations on information selection on news sites and blogs. Communications 39 , (2014). Gottschling, S., Kammerer, Y., Thomm, E. & Gerjets, P. How laypersons consider differences in sources’ trustworthiness and expertise in their regulation and resolution of scientific conflicts. Int. J. Sci. Educ. Part. B . 10 , 335–354 (2020). Mehlis, K. Von der Sender- zur Nutzerqualität. in Journalismus und (sein) Publikum: Schnittstellen zwischen Journalismusforschung und Rezeptions- und Wirkungsforschung (eds (eds Loosen, W. & Dohle, M.) 253–271 (Springer Fachmedien, Wiesbaden, doi: 10.1007/978-3-531-19821-7_14 . (2014). Klebolte, K. Aspekte der Glaubwürdigkeitsbeurteilung – Eine empirische Untersuchung zum Einfluss von Attraktivität, Expertise und Prominenz. Jr Manag Sci. 4 , 524–552 (2019). Ismagilova, E., Slade, E., Rana, N. P. & Dwivedi, Y. K. The effect of characteristics of source credibility on consumer behaviour: A meta-analysis. J. Retail Consum. Serv. 53 , 101736 (2020). Lis, B. eWOM We Trust WIRTSCHAFTSINFORMATIK 55 , 121–134 (2013). Eastin, M. S. Credibility Assessments of Online Health Information: the Effects of Source Expertise and Knowledge of Content. J. Comput. -Mediat Commun. 6 , JCMC643 (2001). Jung, E. H., Walsh-Childers, K. & Kim, H. S. Factors influencing the perceived credibility of diet-nutrition information web sites. Comput. Hum. Behav. 58 , 37–47 (2016). Thon, F. M. & Jucks, R. Believing in Expertise: How Authors’ Credentials and Language Use Influence the Credibility of Online Health Information. Health Commun. 32 , 828–836 (2017). Meinert, J. & Krämer, N. C. How the expertise heuristic accelerates decision-making and credibility judgments in social media by means of effort reduction. PLOS ONE . 17 , e0264428 (2022). Zimmermann, D., Klee, A. & Kaspar, K. Political news on Instagram: influencer versus traditional magazine and the role of their expertise in consumers’ credibility perceptions and news engagement. Front Psychol 14 , (2023). Anders Ericsson, K., Towne, T. J. & Expertise WIREs Cogn. Sci. 1 , 404–416 (2010). van Antwerpen, N., Green, E. B., Sturman, D. & Searston, R. A. The impacts of expertise, conflict, and scientific literacy on trust and belief in scientific disagreements. Sci. Rep. 15 , 11869 (2025). Hocevar, K. P., Metzger, M. & Flanagin, A. J. Source Credibility, Expertise, and Trust in Health and Risk Messaging. in Oxford Research Encyclopedia of Communication (Oxford University Press, doi: 10.1093/acrefore/9780190228613.013.287 . (2017). Knobloch-Westerwick, S., Glynn, C. J. & Huge, M. The Matilda Effect in Science Communication: An Experiment on Gender Bias in Publication Quality Perceptions and Collaboration Interest. Sci. Commun. 35 , 603–625 (2013). Rossiter, M. W. The Matthew Matilda Effect in Science. Soc. Stud. Sci. 23 , 325–341 (1993). Bundi, P. & Varone, F. Hanimann,Anina, Portmann, Leaand The future might be female: how does the public perceive experts? J. Eur. Public Policy 32, 843–869 (2025). Gorisek, C. Zusammenhang von Expertise, Reputation und Glaubwürdigkeit (Hochschule Rhein-Waal, 2022). Greve-Poulsen, K., Larsen, F. K., Pedersen, R. T. & Albæk, E. No Gender Bias in Audience Perceptions of Male and Female Experts in the News: Equally Competent and Persuasive. Int. J. Press. 28 , 116–137 (2023). Bigham, A. The Effect of Source Credibility and Gender on Message Persuasiveness. (2017). http://hdl.handle.net/2346/73515 Pratt, A. N., Tadlock, M. E., Watts, L. L., Wilson, T. C. & Denham, B. E. Perceptions of Credibility and Likeability in Broadcast Commentators of Women’s Sports. J. Sports Media . 13 , 75–97 (2018). Jones, C. W. & Mitchell, J. S. Women reporters as experts on security affairs in Jordan? Rethinking gender and issue competency stereotypes. Mediterr. Polit . 28 , 434–462 (2023). Leiner, D. J. SoSci Survey. (2025). Available at https://www.soscisurvey.de OpenAI. ChatGPT. https://chat.openai.com/ Collins, H. Studies of Expertise and Experience. Topoi 37 , 67–77 (2018). Weinstein, B. D. What is an expert? Theor. Med. 14 , 57–73 (1993). Wagemans, J. H. M. The Assessment of Argumentation from Expert Opinion. Argumentation 25 , 329–339 (2011). Hornikx, J. & Hoeken, H. Cultural Differences in the Persuasiveness of Evidence Types and Evidence Quality. Commun. Monogr. 74 , 443–463 (2007). Burgers, C., de Graaf, A. & Callaars, S. Differences in actual persuasiveness between experiential and professional expert evidence. J. Argum Context . 1 , 194–208 (2012). Ferreira, P. A., Wingrove, T. & and Expert Witness Training History and Professional Experience Exert Separable Impacts on Expert Credibility Perceptions. J. Forensic Psychol. Res. Pract. 24 , 524–540 (2024). R Core Team. R: A Language and Environment for Statistical Computing (R Foundation for Statistical Computing, 2024). Petty, R. E. & Cacioppo, J. T. The Elaboration Likelihood Model of Persuasion. in Advances in Experimental Social Psychology vol. 19 123–205Academic Press, (1986). Pornpitakpan, C. The Persuasiveness of Source Credibility: A Critical Review of Five Decades’ Evidence. J. Appl. Soc. Psychol. 34 , 243–281 (2004). Hilligoss, B. & Rieh, S. Y. Developing a unifying framework of credibility assessment: Construct, heuristics, and interaction in context. Inf. Process. Manag . 44 , 1467–1484 (2008). Metzger, M. J. & Flanagin, A. J. Credibility and trust of information in online environments: The use of cognitive heuristics. J. Pragmat. 59 , 210–220 (2013). Sundar, S. S. The MAIN Model: A Heuristic Approach to Understanding Technology Effects on Credibility. (2008). Yalch, R. F. & Elmore-Yalch, R. The Effect of Numbers on the Route to Persuasion. J. Consum. Res. 11 , 522 (1984). Reinhard, M. A. & Sporer, S. L. Content versus source cue information as a basis for credibility judgments: The impact of task involvement. Soc. Psychol. 41 , 93–104 (2010). Sundar, S. S., Knobloch-Westerwick, S. & Hastall, M. R. News cues: Information scent and cognitive heuristics. J. Am. Soc. Inf. Sci. Technol. 58 , 366–378 (2007). Lee, J. Y. & Sundar, S. S. To Tweet or to Retweet? That Is the Question for Health Professionals on Twitter. Health Commun. 28 , 509–524 (2013). Wathen, C. N. & Burkell, J. Believe it or not: Factors influencing credibility on the Web. J. Am. Soc. Inf. Sci. Technol. 53 , 134–144 (2002). Little, D. & Green, D. A. Credibility in educational development: trustworthiness, expertise, and identification. High. Educ. Res. Dev. 41 , 804–819 (2022). Brown, D. K. The Social Sources of Educational Credentialism: Status Cultures, Labor Markets, and Organizations. Sociol. Educ. 74 , 19–34 (2001). Bokros, S. E. A deference model of epistemic authority. Synthese 198 , 12041–12069 (2021). Pierson, R. The Epistemic Authority of Expertise. PSA Proc. Bienn. Meet. Philos. Sci. Assoc. 398–405 (1994). (1994). Bigham, A., Meyers, C., Li, N. & Irlbeck, E. The Effect of Emphasizing Credibility Elements and the Role of Source Gender on Perceptions of Source Credibility. J Appl. Commun 103 , (2019). Tables Table 1 Sample size, gender, age and educational distribution for Studies 1-3 Study 1 Study 2 Study 3 N total 203 182 206 Gender Female 139 (68.5%) 70 (38.5%) 50 (24.3%) Male 61 (30.1%) 110 (60.5%) 154 (74.8%) Non-binary 3 (1.5%) 1 (0.6%) 2 (1.0%) Other - - 0 (0.0%) Not indicated 0 (0.0%) 1 (0.6%) 0 (0.0%) Age (years) Mean (SD) 30.8 (13.2) 43.5 (12.2) 33.1 (10.2) Range 18-72 19-77 18-72 Educational level University degree 98 (48.3%) 86 (47.3%) 92 (44.7%) Technical college 2 (1.0%) 19 (10.5%) 18 (8.7%) High school diploma 97 (47.8%) 37 (20.3%) 58 (28.2%) Completed apprenticeship 5 (2.5%) 26 (14.3%) 23 (11.2%) Secondary school 1 (0.5%) 11 (6.0%) 12 (5.8%) Secondary modern school 0 (0.0 %) 1 (0.6%) 3 (1.5%) Additional Declarations No competing interests reported. Supplementary Files ExpertiseSupplement.pdf Cite Share Download PDF Status: Under Review Version 1 posted Editorial decision: Revision requested 23 Apr, 2026 Reviews received at journal 21 Apr, 2026 Reviewers agreed at journal 16 Apr, 2026 Reviews received at journal 31 Mar, 2026 Reviewers agreed at journal 13 Mar, 2026 Reviewers agreed at journal 10 Mar, 2026 Reviewers invited by journal 10 Mar, 2026 Editor assigned by journal 10 Mar, 2026 Editor invited by journal 06 Mar, 2026 Submission checks completed at journal 05 Mar, 2026 First submitted to journal 05 Mar, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9010835","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":609585820,"identity":"79b28fd9-16ce-4a44-b30d-78479c8d708c","order_by":0,"name":"Annalena Ulsperger","email":"data:image/png;base64,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","orcid":"","institution":"Leibniz-Institut für Wissensmedien","correspondingAuthor":true,"prefix":"","firstName":"Annalena","middleName":"","lastName":"Ulsperger","suffix":""},{"id":609585821,"identity":"c719bfc1-20a6-453c-a292-e6ee7c669a2a","order_by":1,"name":"Lara Pfannenschwarz","email":"","orcid":"","institution":"University of Tübingen","correspondingAuthor":false,"prefix":"","firstName":"Lara","middleName":"","lastName":"Pfannenschwarz","suffix":""},{"id":609585822,"identity":"92e79ee2-fa0b-41e6-9ff0-317ddabbe804","order_by":2,"name":"Clara Schetla","email":"","orcid":"","institution":"University of Tübingen","correspondingAuthor":false,"prefix":"","firstName":"Clara","middleName":"","lastName":"Schetla","suffix":""},{"id":609585823,"identity":"7515e156-fc4c-4c4e-967b-5e371f950989","order_by":3,"name":"Karen Poletilo","email":"","orcid":"","institution":"University of Tübingen","correspondingAuthor":false,"prefix":"","firstName":"Karen","middleName":"","lastName":"Poletilo","suffix":""},{"id":609585824,"identity":"1b40963a-4641-4aaf-99dd-5644c6362e73","order_by":4,"name":"Lina Strecker","email":"","orcid":"","institution":"University of Tübingen","correspondingAuthor":false,"prefix":"","firstName":"Lina","middleName":"","lastName":"Strecker","suffix":""},{"id":609585825,"identity":"51e5d740-7569-4249-b97a-62f888368bd9","order_by":5,"name":"Marlene Egner","email":"","orcid":"","institution":"University of Tübingen","correspondingAuthor":false,"prefix":"","firstName":"Marlene","middleName":"","lastName":"Egner","suffix":""},{"id":609585826,"identity":"b3b933dd-759e-4244-a688-29b1d976b359","order_by":6,"name":"Joachim Kimmerle","email":"","orcid":"","institution":"Leibniz-Institut für Wissensmedien","correspondingAuthor":false,"prefix":"","firstName":"Joachim","middleName":"","lastName":"Kimmerle","suffix":""}],"badges":[],"createdAt":"2026-03-02 13:39:31","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9010835/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9010835/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":105371515,"identity":"7d04f417-8380-4c21-b402-8c4237d07e95","added_by":"auto","created_at":"2026-03-25 09:28:22","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":32635,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eBoxplot of the perceived message credibility values for the four experimental conditions in Study 3. Mean credibility values are indicated with a lilac diamond.\u003c/em\u003e\u003c/p\u003e","description":"","filename":"Figure1.png","url":"https://assets-eu.researchsquare.com/files/rs-9010835/v1/64419a4842b206c1835f9d9a.png"},{"id":105371590,"identity":"4a82a767-cf63-48a9-8e9a-c70459e17c72","added_by":"auto","created_at":"2026-03-25 09:28:38","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":920616,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9010835/v1/c75f2706-5efa-4caa-bb91-091d8d8abce2.pdf"},{"id":105371475,"identity":"d0a5173c-62b6-48b6-9f94-859297ce6667","added_by":"auto","created_at":"2026-03-25 09:28:15","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":154373,"visible":true,"origin":"","legend":"","description":"","filename":"ExpertiseSupplement.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9010835/v1/e1ceec6ab14a52ce78ea4767.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Authors’ academic but not personal expertise affects message credibility in science communication: A series of experiments on text perception","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eScience communication plays an important role in everyday information transfer and can be a relevant basis for laypeople\u0026rsquo;s decision making\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e. However, this accessibility of information also entails significant risks, such as the spread of misinformation, since the verification of the content being disseminated is removed from the control of traditional gatekeepers\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e2\u003c/span\u003e,\u003cspan class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e. Instead, the responsibility for verifying the credibility of the offered content increasingly falls onto the consumers themselves. Therefore, it is essential that people understand current science-related topics and are able to critically assess the wide range of information available\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e4\u003c/span\u003e\u0026ndash;\u003cspan class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e. Since laypeople often lack the specific subject-matter knowledge to reliably judge the factual accuracy of every content they encounter, they tend to consult other factors that can be used to assess the credibility of information\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e5\u003c/span\u003e,\u003cspan class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e. These can be characteristics of the content itself like its fluency, channel characteristics like the media modality, source characteristics like the expertise of an author, or even characteristics of the recipients\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e, as credibility is subjectively perceived by message receivers\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e. In the research presented here, we focus on the potential influence of author characteristics on laypeople\u0026rsquo;s subjective perception of message credibility. In the following paragraphs, we first introduce the concept of message credibility and its role in science communication before we present expertise and gender as two author characteristics that may influence message credibility perceptions.\u003c/p\u003e\n\u003ch3\u003e1.1 Message credibility\u003c/h3\u003e\n\u003cp\u003eMessage credibility refers to people\u0026rsquo;s subjective perceptions of the integrity, reliability, and trustworthiness of a given content, such as a text or other content presentation\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e9\u003c/span\u003e,\u003cspan class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e. The credibility of a message describes the belief of receivers that the provided information is true, approvable, and trustworthy\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e11\u003c/span\u003e,\u003cspan class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e. This belief strongly impacts the acceptance of the information provided and its effect on the receivers beliefs and attitudes, with more credible content being more influential than less- and non-credible content\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e13\u003c/span\u003e\u0026ndash;\u003cspan class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eVarious subfactors have been used to conceptualize and quantify the perceived credibility of a message\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e. One important facet of message credibility is its trustworthiness\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e. \u003cem\u003eTrustworthiness\u003c/em\u003e means that a message is perceived to be reliable and believable, and that its recipients presume a high level of veracity of that message\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e. Other relevant dimensions of credibility include the perceived truth of and the agreement with a message. \u003cem\u003eTruth\u003c/em\u003e comprises the perceived accuracy of the provided statements\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e. A\u003cem\u003egreement\u003c/em\u003e reflects the overlap of a receiver's opinion with the message, which would result in approving the message and consenting to it\u003csup\u003e20\u003c/sup\u003e.\u003c/p\u003e\n\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\n\u003ch2\u003e1.2 Expertise\u003c/h2\u003e\n\u003cp\u003eSource expertise has been shown to influence message credibility in various contexts, like in online marketing communications\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e21\u003c/span\u003e,\u003cspan class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e, online health information\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e23\u003c/span\u003e\u0026ndash;\u003cspan class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e, or political information on social media platforms\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e26\u003c/span\u003e,\u003cspan class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e. Expertise refers to the domain-specific knowledge and ability of a source, for example the author of a text, to provide accurate information in said domain developed over time\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e8\u003c/span\u003e,\u003cspan class=\"CitationRef\"\u003e28\u003c/span\u003e,\u003cspan class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e. Established expertise indicators are, for example, academic qualifications and credentials, as seen in health practitioners\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e18\u003c/span\u003e,\u003cspan class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eIn general, the literature shows that experts are considered to be more credible and persuasive than non-experts\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e25\u003c/span\u003e,\u003cspan class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e: For example, Eastin\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e found that study participants were more likely to believe experts when it came to health information if they themselves had no prior knowledge of the subject. Jung et al.\u003csup\u003e24\u003c/sup\u003e found similar results for participants with low prior knowledge in the field of online diet and nutrition information credibility. Zimmermann et al.\u003csup\u003e27\u003c/sup\u003e showed that political news on Instagram were rated higher in credibility when higher source expertise was indicated. And Meinert and Kr\u0026auml;mer\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e showed that decision times between two sources were decreased when expertise cues were present, with expert sources being ascribed more credibility and chosen more often.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec4\" class=\"Section2\"\u003e\n\u003ch2\u003e1.3 Gender\u003c/h2\u003e\n\u003cp\u003eAnother less obvious factor influencing the credibility of a text is the gender of its author. While historically female experts have been perceived as less competent than their male counterparts in science communication\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e31\u003c/span\u003e,\u003cspan class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e, contemporary literature indicates a more equalized perception\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e33\u003c/span\u003e\u0026ndash;\u003cspan class=\"CitationRef\"\u003e35\u003c/span\u003e\u003c/sup\u003e, or even an advantage of female experts compared to males\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e33\u003c/span\u003e,\u003cspan class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/sup\u003e. For example, female sports reporters covering women's sports were rated as more likable and credible than male reporters, especially by female study participants\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e37\u003c/span\u003e\u003c/sup\u003e. Bundi et al.\u003csup\u003e33\u003c/sup\u003e found similar preferences by their female participants for female experts in climate change communication. Jones and Mitchell\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e even found that female experts were rated as slightly more credible in traditionally male-dominated fields such as security policy.\u003c/p\u003e\n\u003cp\u003eTaken together, prior research suggests that expertise and gender as author characteristics may have an impact on readers\u0026rsquo; credibility perceptions regarding science communication texts. While the expertise effect is rather established, the effect of an author\u0026rsquo;s gender is subject to change. However, empirical evidence on how these factors interact in the current science communication climate remains limited. To address this research gap, we conducted a series of three studies examining the role of the authors\u0026rsquo; characteristics in perceived credibility of a science communication article.\u003c/p\u003e\n\u003c/div\u003e\n\u003ch3\u003e\u0026nbsp;\u003c/h3\u003e"},{"header":"2.\tStudy 1","content":"\u003cp\u003eIn the first study, we investigated whether the alleged expertise and gender of an author affected the perceived credibility of a science communication text. Based on the literature review and the considerations presented above, we formulated the following hypotheses:\u003c/p\u003e\n\u003cp\u003eH1: Scientific texts by authors with high expertise are attributed greater credibility compared to texts by authors with low expertise.\u003c/p\u003e\n\u003cp\u003eH2: Scientific texts by female authors are attributed greater credibility compared to texts by male authors.\u003c/p\u003e\n\u003cp\u003eFurthermore, we investigated as an exploratory research question based on Bundi et al. (2025) whether there are empirical differences based on the gender of the participants (ERQ1). We also explored whether there is an interaction effect between the variables gender and expertise (ERQ2).\u003c/p\u003e\n\u003ch2\u003e2.1. Methods\u003c/h2\u003e\n\u003cp\u003eAll three studies reported in this paper were preregistered on AsPredicted prior to data collection (Study 1: https://aspredicted.org/cjcj-rc3b.pdf, Study 2: https://aspredicted.org/nyrs-tm7s.pdf, Study3: https://aspredicted.org/tbdk-9mx8.pdf) and approved by the Ethics Committee of the Leibniz-Institut für Wissensmedien (LEK 2024/054). Participants provided written informed consent before and after participation. Participation was anonymous and participants were informed beforehand that they had the right to discontinue the study and withdraw their data at any time during the participation. All three studies were conducted in German. The experiments were performed in accordance with relevant guidelines and regulations.\u003c/p\u003e\n\u003ch2\u003e2.1.1.\u0026nbsp; Participants\u003c/h2\u003e\n\u003cp\u003eWe conducted Study 1 as an online experiment using SoSci Survey (www.soscisurvey.de; Version 3.6.10)\u003csup\u003e39\u003c/sup\u003e. In a 2x2 between-groups design, we varied the source factors expertise (high vs. low) and gender (male vs. female) of the alleged author of a science communication text by means of an author description.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eBased on an a priori power analysis (GPower, version 3.1.9.7), we aimed to recruit at least 199 participants to achieve 80% power for detecting a small effect (f=0.2, α =.05). N=234 participants completed the study, of which 31 had to be excluded due to the predefined exclusion criteria of high prior knowledge in the topic of the text or a failed attention check. Participants were recruited via the e-mail distribution list of a German university and given the option to participate in a gift-card raffle as compensation. They had to be at least 16 years old and have at least a C1 level of proficiency in German. The demographic data of the samples of all three studies are shown in Table 1.\u003c/p\u003e\n\u003ch2\u003e2.1.2.\u0026nbsp; Procedure\u003c/h2\u003e\n\u003cp\u003eFirst, participants filled in the informed consent form. They were told that they would be asked about their assessment of a text. Afterwards, the participants were presented with one of four texts introducing an author who allegedly had written a science communication text about natural disaster preparedness, which the participants would read afterwards. Participants were distributed equally among the four conditions (\u003cem\u003en\u003c/em\u003e\u003csub\u003ehigh exp., fem.\u003c/sub\u003e=52, \u003cem\u003en\u003c/em\u003e\u003csub\u003ehigh exp., male\u003c/sub\u003e=50, \u003cem\u003en\u003c/em\u003e\u003csub\u003elow exp., fem.\u003c/sub\u003e=50, \u003cem\u003en\u003c/em\u003e\u003csub\u003elow exp., male\u003c/sub\u003e=51). After ten seconds the participants could move on to the next page, where they were presented with a text concerning natural disaster preparedness. The text was the same across all four conditions and was shown for at least 15 seconds before participants could move on. Thereafter, they were asked to judge the credibility of the text. Participants were then asked to determine the gender of the alleged author (attention check), the expertise level of the author (manipulation check), and their prior knowledge about natural disaster preparedness. Additionally, participants were directly asked whether they perceived men, women, or both as more credible in the topic. Then the participants were asked about their demographics (gender, age, and educational background) and given the option to participate in the raffle. Finally, they were debriefed and dismissed.\u0026nbsp;\u003c/p\u003e\n\u003ch2\u003e2.1.3.\u0026nbsp; Materials and Measures\u003c/h2\u003e\n\u003cp\u003eThe independent variables \u003cem\u003eauthor expertise\u003c/em\u003e and \u003cem\u003eauthor gender\u003c/em\u003e were manipulated via the author description. Expertise was presented as either high (longstanding academic career and international consulting experience with respect to the topic) or low (no academic background, only some personal interest due to previous flood experience in the author’s hometown). Gender was indicated through title, name, and pronouns as either male or female. All four versions of these stimuli are provided in the Supplementary Material (S1).\u003c/p\u003e\n\u003cp\u003eThe science communication text was introduced as describing the author’s views on how sustainable and ecological methods can provide optimal protection against natural disasters. After briefly outlining the topic, the text claimed that nature-based measures alone were sufficient to prevent natural disasters. We created the text using specific prompts in ChatGPT-4 (OpenAI; Version 4.0)\u003csup\u003e40\u003c/sup\u003e. See Supplementary Material for the full text (S2).\u003c/p\u003e\n\u003cp\u003eWe assessed message credibility using 11 items (α=.94) answered on a 7-point Likert scale (1 being “not at all” and 7 being “fully applies”). The scale comprised three dimensions: trustworthiness (4 items), truth (3 items), and agreement (4 items). The items for both the trustworthiness dimension and two of the items of the agreement dimension were adapted to the present context from Klebolte\u003csup\u003e20\u003c/sup\u003e. The other two items of the agreement dimension were added by the authors. The truth dimension was extracted from Mehlis\u003csup\u003e19\u003c/sup\u003e. The message credibility score was calculated as the mean of all items. The complete scale is provided in the Supplementary Material (S3).\u003c/p\u003e\n\u003cp\u003eThe question “Does the author have expertise in the subject area?” answered on a 7-point Likert scale (1=“not at all” and 7=“fully applies”) was used as a manipulation check.\u003c/p\u003e\n\u003cp\u003eTo measure prior knowledge, we asked the following question: “Did you already have prior knowledge in the field of natural disaster preparedness?” Participants answered on a 7-point Likert scale (1=“no prior knowledge at all” and 7=“extremely high prior knowledge”). Participants indicating 6 or 7 on the scale were excluded from the analysis\u003c/p\u003e\n\u003cp\u003eAs an attention check, we asked “Was the author a man or a woman?” with the options: “male”, “female”, or “I don’t remember”. Participants who did not indicate the correct answer were excluded.\u003c/p\u003e\n\u003cp\u003eTo ensure that the topic of the text itself was not tied to a specific gender, we asked “Do you consider men or women to be more credible in the field of preventive natural disaster preparedness?” with the answering options “male”, “female”, or “both equally”.\u0026nbsp;\u003c/p\u003e\n\u003ch2\u003e2.2. Results\u003c/h2\u003e\n\u003cp\u003eWe used a Welch t-test to check the manipulation. It showed that the perceived expertise differed significantly between the high and the low-expertise conditions, \u003cem\u003et\u003c/em\u003e(162.73)=-13.19, \u003cem\u003ep\u003c/em\u003e\u0026lt;.001,\u003cem\u003e\u0026nbsp;d\u003c/em\u003e=1.86. As the authors in the high-expertise conditions were perceived as having significantly more expertise (\u003cem\u003eM\u003c/em\u003e=5.97, \u003cem\u003eSD\u003c/em\u003e=0.96) than the authors in the low-expertise conditions (\u003cem\u003eM\u003c/em\u003e=3.51, \u003cem\u003eSD\u003c/em\u003e=1.61), the manipulation was considered as successful. Furthermore, the topic of the science communication text was successfully considered gender-neutral, since most participants answered the corresponding question with “both equally” (\u003cem\u003en\u003c/em\u003e\u003csub\u003eboth\u003c/sub\u003e=172, \u003cem\u003en\u003c/em\u003e\u003csub\u003ewomen\u003c/sub\u003e=18, \u003cem\u003en\u003c/em\u003e\u003csub\u003emen\u003c/sub\u003e=13).\u003c/p\u003e\n\u003cp\u003eTo test the hypotheses that authors with high expertise (H1) and female authors (H2) would be attributed greater credibility, we conducted a two-way analysis of variance (ANOVA) with the between-groups factors expertise (high/low) and gender (female/male) of the author as independent variables and message credibility as the dependent variable. The ANOVA revealed a significant main effect of author expertise, \u003cem\u003eF\u003c/em\u003e(1, 199)=5.38, \u003cem\u003ep\u003c/em\u003e=.021, partial η\u003cem\u003e²\u003c/em\u003e=.03. The results indicate that the text by authors with alleged high expertise was rated as more credible (\u003cem\u003eM\u003c/em\u003e=4.85, \u003cem\u003eSD\u003c/em\u003e=1.09) than the identical text by authors with alleged low expertise (\u003cem\u003eM\u003c/em\u003e=4.47, \u003cem\u003eSD\u003c/em\u003e=1.27), supporting H1. There was no main effect of gender, \u003cem\u003eF\u003c/em\u003e(1, 199)=0.10, \u003cem\u003ep\u003c/em\u003e=.757, therefore rejecting H2. Regarding ERQ2, there was no significant interaction effect between the factors author expertise and author gender, \u003cem\u003eF\u003c/em\u003e(1, 199)=0.06, \u003cem\u003ep\u003c/em\u003e=.809.\u003c/p\u003e\n\u003cp\u003eTo examine whether there was an empirical difference based on the gender of the participants (ERQ1), the analysis was expanded to a three-way ANOVA with participant gender as an additional between-groups factor. The main effect of author expertise remained significant,\u003cem\u003e\u0026nbsp;F\u003c/em\u003e(1, 193)=5.38,\u003cem\u003e\u0026nbsp;p\u003c/em\u003e=.021, partial \u003cem\u003eη²\u003c/em\u003e=.03, while the main effects of author gender and participant gender were not significant, \u003cem\u003eF\u003c/em\u003es\u0026lt;1.12, \u003cem\u003ep\u003c/em\u003es\u0026gt;.331. Neither of the two-way interaction, \u003cem\u003eF\u003c/em\u003es\u0026lt;0.37, \u003cem\u003ep\u003c/em\u003es\u0026gt;.694, nor the three-way interaction, \u003cem\u003eF\u003c/em\u003e(1, 193)=2.75, \u003cem\u003ep\u003c/em\u003e=.099, reached statistical significance.\u0026nbsp;\u003c/p\u003e\n\u003ch2\u003e2.3. Discussion\u003c/h2\u003e\n\u003cp\u003eOur findings support a positive effect of expertise on perceived message credibility in line with previous literature\u003csup\u003e23,27\u003c/sup\u003e. However, the observed effect was small and even the low-expertise condition received rather high credibility evaluations.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eContrary to our hypotheses and contemporary studies reporting a preference for female over male experts\u003csup\u003e33,37\u003c/sup\u003e, we did not find differences in credibility perceptions between male and female authors. We also observed no differences in credibility ratings based on the participants’ own gender. Instead, the results align with more recent literature suggesting a shift toward gender-independent credibility evaluations in science\u003csup\u003e33–35\u003c/sup\u003e, indicating a societal shift toward gender neutrality when evaluating experts. This might have been further promoted by our choice of a gender-neutral topic, which possibly minimized the salience of the author’s gender. Moreover, the lack of a gender effect might also reflect the specific characteristics of our sample. Egalitarian views and social desirability bias might be more prominent in a sample of university students.\u0026nbsp;\u003c/p\u003e"},{"header":"3.\tStudy 2","content":"\u003cp\u003eTo assess the generalizability of the findings of Study 1, we conducted a replication study using a sample that was more representative of the general population in terms of age and education. Since we did not observe an effect of the authors’ gender in Study 1, which is in accordance with more recent literature\u003csup\u003e33–35\u003c/sup\u003e, we assumed to find only a main effect of expertise in the second study:\u003c/p\u003e\n\u003cp\u003eH1.\u0026nbsp;Scientific texts allegedly written by authors with high expertise are attributed greater credibility compared to texts by authors with low expertise.\u003c/p\u003e\n\u003cp\u003eWe were, again, interested in whether there were empirical differences based on the gender of the participants (ERQ1) and whether there was an interaction effect between expertise and gender of the author (ERQ2). Whether there would be a difference in credibility perceptions of the text based on the alleged gender of the author with this broader sample was also only investigated as an exploratory research question (ERQ3).\u0026nbsp;\u003c/p\u003e\n\u003ch2\u003e3.1. Methods\u003c/h2\u003e\n\u003ch2\u003e3.1.1.\u0026nbsp; Participants\u003c/h2\u003e\n\u003cp\u003eThe same between-groups design and exclusion criteria were used as in Study 1. However, the online experiment was hosted on Qualtrics (Qualtrics, Provo, UT). To achieve a sample more representative of the general population, participants were recruited via the online crowdsourcing platform \u003cem\u003eClickworker\u003c/em\u003e, each receiving monetary compensation of 2,14€ for completing the study. 219 participants took part in the study of which 37 had to be excluded due to the predefined exclusion criteria. In contrast to Study 1, this sample included more male than female participants, the participants were, on average, older than those in Study 1, and the educational background was more diverse with more participants indicating a completed apprenticeship, secondary school, and technical college as their highest degree. The demographic data of Study 2 is presented in Table 1.\u003c/p\u003e\n\u003ch2\u003e3.1.2.\u0026nbsp; Procedure\u003c/h2\u003e\n\u003cp\u003eThe procedure was identical to Study 1. The participants were distributed among the four conditions as follows: \u003cem\u003en\u003c/em\u003e\u003csub\u003ehigh exp., fem.\u003c/sub\u003e=41, \u003cem\u003en\u003c/em\u003e\u003csub\u003ehigh exp., male\u003c/sub\u003e=45, \u003cem\u003en\u003c/em\u003e\u003csub\u003elow exp., fem.\u003c/sub\u003e=47, \u003cem\u003en\u003c/em\u003e\u003csub\u003elow exp., male\u003c/sub\u003e=49.\u003c/p\u003e\n\u003ch2\u003e3.1.3.\u0026nbsp; Materials and Measures\u003c/h2\u003e\n\u003cp\u003eThe same materials and measures were used as in Study 1. The internal consistency of the message credibility scale was, again, excellent with α=.95.\u003c/p\u003e\n\u003ch2\u003e3.2. Results\u003c/h2\u003e\n\u003cp\u003eAgain, a Welch t-test confirmed that the manipulation was successful, \u003cem\u003et\u003c/em\u003e(138.10)=-13.05, \u003cem\u003ep\u003c/em\u003e\u0026lt;.001, \u003cem\u003ed\u003c/em\u003e=1.88. Perceived expertise differed significantly between the conditions, with authors in the high-expertise conditions being perceived as having significantly more expertise (\u003cem\u003eM\u003c/em\u003e=6.13, \u003cem\u003eSD\u003c/em\u003e=0.76) than the authors in the low-expertise conditions (\u003cem\u003eM\u003c/em\u003e=3.71, \u003cem\u003eSD\u003c/em\u003e=1.63). Participants also indicated that they again did not prefer one gender over the other in the context of preventive natural disaster preparedness (\u003cem\u003en\u003c/em\u003e\u003csub\u003eboth\u003c/sub\u003e=166, \u003cem\u003en\u003c/em\u003e\u003csub\u003ewomen\u003c/sub\u003e=9, \u003cem\u003en\u003c/em\u003e\u003csub\u003emen\u003c/sub\u003e=7).\u003c/p\u003e\n\u003cp\u003eAs in Study 1, we used an ANOVA with the between-groups factors expertise (high/low) and gender (female/male) of the author as independent variables and message credibility score as the dependent variable to test for the hypothesized main effect of expertise (H1), as well as to answer ERQ2 and ERQ3. The two-way ANOVA showed neither a main effect of author expertise, \u003cem\u003eF\u003c/em\u003e(1, 178)=0.24, \u003cem\u003ep\u003c/em\u003e=.626, nor a main effect of gender, \u003cem\u003eF\u003c/em\u003e(1, 178)=0.19, \u003cem\u003ep\u003c/em\u003e=.662, or an interaction between the factors, \u003cem\u003eF\u003c/em\u003e(1, 178)=3.39, \u003cem\u003ep\u003c/em\u003e=.067. Therefore, H1 was not supported. We also found no differences regarding ERQ2 and ERQ3.\u003c/p\u003e\n\u003cp\u003eTo examine whether participants’ own gender influenced credibility ratings (ERQ1), an additional three-way ANOVA with participant gender as an additional between-groups factor was conducted. None of the main effects, \u003cem\u003eF\u003c/em\u003es\u0026lt;0.92,\u003cem\u003e\u0026nbsp;p\u003c/em\u003es\u0026gt;.434, neither of the two-way interactions, \u003cem\u003eF\u003c/em\u003es\u0026lt;3.23, \u003cem\u003ep\u003c/em\u003es\u0026gt;.074, nor the three-way interaction,\u003cem\u003e\u0026nbsp;F\u003c/em\u003e(1, 172)=0.42, \u003cem\u003ep\u003c/em\u003e=.520, reached significance.\u0026nbsp;\u003c/p\u003e\n\u003ch2\u003e3.3. Discussion\u003c/h2\u003e\n\u003cp\u003eStudy 2 was conducted to evaluate whether the effects observed in Study 1 would generalize to a more heterogeneous sample. For the gender of the author and the gender of the participants, we found again no significant effect. This is consistent with Study 1 and strengthens the conclusion that author and participant gender do not influence participants’ credibility perception under our design conditions. Thus, neither author nor participant gender were further investigated in the subsequent study.\u003c/p\u003e\n\u003cp\u003eThe main effect of the authors’ expertise could not be replicated. Whereas Study 1 revealed a significant main effect of author expertise on perceived message credibility, Study 2 showed no such effect in the sample with larger demographic variability. This suggests a dependence of the effect observed in Study 1 on characteristics of the student sample. However, expertise is widely regarded as an established indicator of perceived credibility across various samples in the literature\u003csup\u003e23–27\u003c/sup\u003e, wherefore this explanation seems unlikely. A closer look reveals the study material as a plausible source for the inconsistent findings: While the low-expertise condition introduced the alleged author as being new to the topic at hand, they were still ascribed previous real-life experience and personal interest.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eExpertise is not a homogeneous construct, but can encompass various forms of knowledge, skills, and abilities all perceived as expertise in real-life communication practice\u003csup\u003e28,41\u003c/sup\u003e. Specifically, it can be differentiated based on what grounds people are deemed to be experts: Weinstein\u003csup\u003e42\u003c/sup\u003e differentiates between epistemic expertise, which is based on theoretical knowledge, and performative expertise, which is based on practical skills. Similarly, Wagemans\u003csup\u003e43\u003c/sup\u003e also proposes a dichotomy between expert opinion based on professional knowledge acquired through professional training and experiential knowledge gained through personal experience. Following this breakdown of the concept of expertise, the author descriptions used in Study 1 and 2 possibly entailed an unintended dichotomy between professional academic expertise, and experience-based expertise. As different types of expertise may affect credibility perceptions to differing extents, this blend in our study material might have led to inconsistent results.\u0026nbsp;\u003c/p\u003e"},{"header":"4.\tStudy 3","content":"\u003cp\u003eTherefore, we investigated whether there were differences in the effect of expertise on credibility judgments of science communication texts based on the type of expertise. This provides an important conceptual contribution to expertise research and may disentangle the conflicting results of Study 1 and 2. Previous research on source expertise often combined these expertise indicators\u003csup\u003e18\u003c/sup\u003e and experimental research systematically differentiating between different forms of expertise and their impact on perceived message credibility is sparse.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eBuilding on the differing effects of normatively strong vs normatively weak evidence on persuasion\u003csup\u003e44\u003c/sup\u003e, Burgers et al.\u003csup\u003e45\u003c/sup\u003e empirically demonstrated the importance of this distinction. Their study found that the perceived expertise of technical experts (based on systematic professional knowledge) was significantly higher than that of experiential experts (based on practical experience), even though the perceived trustworthiness was roughly the same. In addition, the statements made by authors with professional knowledge were significantly more persuasive than those made by authors with experiential knowledge. Ferreira and Wingrove\u003csup\u003e46\u003c/sup\u003e made a similar distinction between expert training and expert experience, showing that both factors had an independent impact on perceived credibility of forensic experts. We deduced that there might have been similar effects of the two forms of author expertise on perceived message credibility in our studies.\u003c/p\u003e\n\u003cp\u003eThe author descriptions in Studies 1 and 2 contained professional academic as well as personal experiential expertise as described in Wagemans\u003csup\u003e43\u003c/sup\u003e and Burgers et al.\u003csup\u003e45\u003c/sup\u003e. Academic expertise was based on formalized knowledge, academic training, and institutional recognition. Personal expertise was based on practical knowledge acquired through direct engagement with the topic. In Study 3, we explicitly differentiated between the two types of expertise and investigated whether there are differences in credibility judgments. We formulated the following hypotheses:\u003c/p\u003e\n\u003cp\u003eH3.\u0026nbsp;Main effect of academic expertise: A scientific text is perceived as more credible when allegedly written by an author with high academic expertise compared to an author with no academic expertise.\u003c/p\u003e\n\u003cp\u003eH4.\u0026nbsp;Main effect of personal expertise: A scientific text is perceived as more credible when allegedly written by an author with high personal expertise compared to an author with no personal expertise.\u003c/p\u003e\n\u003cp\u003eH5.\u0026nbsp;Combined effect: A scientific text is perceived as significantly more credible when allegedly written by an author with both forms of expertise (high academic and high personal expertise) compared to authors with only one form of expertise (either academic or personal).\u003c/p\u003e\n\u003cp\u003eFurthermore, we investigated in two exploratory research questions whether there is an interaction effect between the alleged academic expertise and personal expertise of an author of a scientific text (ERQ4), as well as whether there is a difference in the perception of credibility of scientific texts allegedly written by an author with exclusively academic expertise compared to an author with exclusively personal expertise (ERQ5).\u003c/p\u003e\n\u003ch2\u003e4.1. Methods\u003c/h2\u003e\n\u003ch2\u003e4.1.1.\u0026nbsp; Participants\u003c/h2\u003e\n\u003cp\u003eStudy 3 was conducted as an online experiment using SoSci Survey (www.soscisurvey.de; Version 3.6.10)\u003csup\u003e39\u003c/sup\u003e. It contained a 2x2 between-groups design with academic expertise (high vs. none) and personal expertise (high vs. none) of the alleged author of a science communication text being varied.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eOf the 268 participants who had completed the study, 62 had to be excluded due to the predefined exclusion criteria. Like in Studies 1 and 2, participants were excluded based on high prior knowledge in the topic (indicating 6 or 7 on a 7-point Likert scale) or a failed attention check. Participants were recruited via Prolific (www.prolific.com), an online participant recruitment platform. They received monetary compensation of 1,20\u0026euro; for completing the study. They had to confirm that they were at least 16 years of age and had at least a C1 level of proficiency in German. For further demographic information see Table 1.\u003c/p\u003e\n\u003ch2\u003e4.1.2.\u0026nbsp; Procedure\u003c/h2\u003e\n\u003cp\u003eParticipants were informed that the goal of the study was to investigate the perception of scientific texts. After completing the informed consent, the participants were instructed to first read one of four descriptions of an author of a scientific text and then the text itself. They had a minimal reading time of 15 seconds for the author description and minimally 60 seconds for the scientific text. Participants were distributed equally among the four conditions (\u003cem\u003en\u003c/em\u003e\u003csub\u003edual exp.\u003c/sub\u003e=46, \u003cem\u003en\u003c/em\u003e\u003csub\u003eacad. exp.\u003c/sub\u003e=48, \u003cem\u003en\u003c/em\u003e\u003csub\u003epers. exp.\u003c/sub\u003e=51, \u003cem\u003en\u003c/em\u003e\u003csub\u003eno exp.=\u003c/sub\u003e61). On the following page, the manipulation check was conducted using two questions regarding the degree of academic and personal experience of the author in the subject area. Then the participants were presented with the same science communication text used in Studies 1 and 2. Afterwards, the participants indicated their perceived credibility of the text. They were also asked to indicate their prior knowledge about natural disaster preparedness. The study concluded with demographic questions.\u003c/p\u003e\n\u003ch2\u003e4.1.3.\u0026nbsp; Materials and Measures\u003c/h2\u003e\n\u003cp\u003eParticipants were presented with author descriptions reflecting four expertise conditions: (1) dual expertise (high academic and high personal expertise), (2) high academic expertise only, (3) high personal expertise only, and (4) no expertise. The descriptions were based on those of Study 1 and 2 but adapted to more concisely separate the types of expertise in the different conditions. Academic expertise was signaled by a doctorate, an extensive academic career, and international consulting experience. Personal expertise was conveyed through firsthand experience with a major flood and substantial practical engagement in flood prevention. In the no-expertise condition, the author was described as a newcomer to the topic with minimal prior involvement. We created the descriptions using ChatGPT-4 (OpenAI; Version 4.0)\u003csup\u003e40\u003c/sup\u003e. The stimuli are provided in the Supplementary Material (S4). The same science communication text was used as in Studies 1 and 2.\u003c/p\u003e\n\u003cp\u003eMessage credibility was measured using the same 11-item scale as in Studies 1 and 2 (\u0026alpha;=.96).\u003c/p\u003e\n\u003cp\u003eAs a manipulation check, the participants independently answered the two questions \u0026ldquo;How much academic experience do you think the author has in the subject area?\u0026quot; and \u0026quot;How much personal experience do you think the author has in the subject area?\u0026rdquo; on 7-point Likert scales (1=\u0026ldquo;no experience\u0026rdquo; and 7=\u0026ldquo;extensive experience\u0026rdquo;).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTo measure prior knowledge, we asked the same question as in the previous studies, also answered on a 7-point Likert scale. However, the labeling of the ends was switched to 1=\u0026ldquo;not at all\u0026rdquo; and 7=\u0026ldquo;fully applies\u0026rdquo;. Participants indicating 6 or 7 on the scale were, again, excluded from the analysis\u003c/p\u003e\n\u003ch2\u003e4.2. Results\u003c/h2\u003e\n\u003cp\u003eThe statistical analysis was conducted using R (Version 4.3.0)\u003csup\u003e47\u003c/sup\u003e. The successful manipulation of both the academic and personal expertise was confirmed using two Welch\u0026rsquo;s t-tests. For academic expertise, \u003cem\u003et\u003c/em\u003e(200.12)=21.56, \u003cem\u003ep\u003c/em\u003e\u0026lt;.001, \u003cem\u003ed\u003c/em\u003e=3.02, the conditions with high academic expertise were regarded significantly higher in academic experience (\u003cem\u003eM\u003c/em\u003e=5.85, \u003cem\u003eSD\u003c/em\u003e=1.27) than the conditions without academic expertise (\u003cem\u003eM\u003c/em\u003e=1.96, \u003cem\u003eSD\u003c/em\u003e=1.32). For personal expertise, \u003cem\u003et\u003c/em\u003e(160.56)=16.92, \u003cem\u003ep\u003c/em\u003e\u0026lt;.001, \u003cem\u003ed\u003c/em\u003e=2.29, the conditions with high personal expertise were regarded significantly higher in personal experience (\u003cem\u003eM\u003c/em\u003e=6.27, \u003cem\u003eSD\u003c/em\u003e=0.90) than the conditions without personal expertise (\u003cem\u003eM\u003c/em\u003e=2.95, \u003cem\u003eSD\u003c/em\u003e=1.86).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTo understand whether there was a main effect of academic expertise (H3), a main effect of personal expertise (H4), or an interaction between the two (ERQ4), a linear model was used. The overall model was significant, \u003cem\u003eF\u003c/em\u003e(3, 202)=3.65, \u003cem\u003ep\u003c/em\u003e=.013, \u003cem\u003eR\u003c/em\u003e\u0026sup2;=.05. Academic expertise was a significant positive predictor of perceived message credibility (\u0026beta;=0.52, \u003cem\u003ep\u003c/em\u003e=.002), whereas personal expertise (\u0026beta;=-0.12, \u003cem\u003ep\u003c/em\u003e=.459) and the interaction (\u0026beta;=0.21, \u003cem\u003ep\u003c/em\u003e=.524) were not significant. This supports H3, while H4 could not be supported and ERQ4 was negated. The credibility values of all four conditions are depicted in Figure 1.\u003c/p\u003e\n\u003cp\u003eTo investigate whether there was a combined effect of the types of expertise (H5) and a difference based on the types of expertise presented exclusively (ERQ5), we fitted a linear model with contrast coding (CH5, CE5, R1). Contrast coding was applied as follows: CH5=[0, -0.5, -0.5, 1], CE5=[0, -0.5, 0.5, 0], R1=[-0.75, 0.25, 0.25, 0.25], with [1, 2, 3, 4] referring to the experimental conditions 1) no expertise, 2) no academic / high personal expertise, 3) high academic / no personal expertise, and 4) high academic / high personal expertise. The overall model was significant, \u003cem\u003eF\u003c/em\u003e(3, 202)=3.65, \u003cem\u003ep\u003c/em\u003e=.013, \u003cem\u003eR\u003c/em\u003e\u0026sup2;=.05. The contrast CH5, representing the contrast between dual and the average of the singular types of expertise (H5), was not significant (\u0026beta;=0.20, \u003cem\u003ep\u003c/em\u003e=.144). Therefore, H5 could not be supported. However, the contrast CE5, which directly compared exclusively academic with exclusively personal expertise (ERQ5), was a significant positive predictor of credibility (\u0026beta;=0.64, \u003cem\u003ep\u003c/em\u003e=.007), answering ERQ5 affirmatively. The residual contrast R1, which compared all three expertise conditions with the control condition of no expertise, was not significant (\u0026beta;=0.20, \u003cem\u003ep\u003c/em\u003e=.266).\u003c/p\u003e\n\u003ch2\u003e4.3. Discussion\u003c/h2\u003e\n\u003cp\u003eStudy 3 showed a significant main effect of academic expertise, indicating higher perceived credibility for texts written by sources with high academic expertise than sources with no academic expertise. Academic-only sources were also rated more credible than personal-only sources when only one type of expertise was present. The significance of academic expertise is in accordance with literature indicating effects of expertise on credibility perception\u003csup\u003e23,27\u003c/sup\u003e. Furthermore, it is partially in accordance with the findings by Burgers et al.\u003csup\u003e45\u003c/sup\u003e and Ferreira and Wingrove\u003csup\u003e46\u003c/sup\u003e, who showed increased persuasiveness of people with professional compared to personal expertise and independent effects of professional expertise on credibility assessments, respectively.\u003c/p\u003e"},{"header":"5. General discussion","content":"\u003cp\u003eCorrectly judging the credibility of science communication texts without proper expertise in the field is an essential skill for laypeople. Besides the content, other properties of the communication environment, like information about the author, can be used to approximate message credibility. Therefore, it is important to critically regard the impact of author features on the subjective perception of messages in science communication.\u003c/p\u003e \u003cp\u003eIn three experimental studies, we investigated whether the alleged expertise and gender of an author impact the perceived message credibility of a gender-neutral science communication text. In terms of author gender, Study 1 and 2 showed no significant effects of author gender on perceived message credibility. The participants\u0026rsquo; own gender was also no moderator of the credibility judgments. We interpreted the results of Study 1 and 2 as supporting the current trend toward gender-independent credibility assessments of science communication\u003csup\u003e\u003cspan additionalcitationids=\"CR34\" citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eWhile Study 1 found an effect of author expertise on message credibility with alleged high author expertise leading to increased credibility ratings in a student sample, Study 2 failed to replicate this effect with a more diverse sample. While labelling the conditions in Studies 1 and 2 as high or low in expertise, the conditions comprised a mixture of academic and personal expertise. Strictly differentiating between the two concepts, we investigated their specific impact on message credibility in Study 3. We found that only academic, but not personal expertise had a significant effect on perceived message credibility. We also did not find additive or interactive effects of both types of expertise.\u003c/p\u003e \u003cp\u003eThese results are only partially in accordance with the literature, which indicates independent effects for both academic and personal expertise\u003csup\u003e\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e,\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e\u003c/sup\u003e. In accordance with the Elaboration Likelihood Model\u003csup\u003e\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e,\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e\u003c/sup\u003e and the Unifying Framework of Credibility Assessment\u003csup\u003e\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e\u003c/sup\u003e, we assume that the presented science communication text was judged based on heuristics activated by the author descriptions, which are more favorable to established cues like academic credentials and thereby reduce processing effort\u003csup\u003e\u003cspan additionalcitationids=\"CR52\" citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e\u003c/sup\u003e. This is supported by the consideration that the text itself was probably not personally relevant to the participants and was thus processed with low effort\u003csup\u003e\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e,\u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e54\u003c/span\u003e,\u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eThe symbolic power of academic titles probably carries specific weight in relation to the results of Study 1. Academic titles represent institutional validation by the scientific system\u003csup\u003e\u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e,\u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e\u003c/sup\u003e and the symbolic power of academic titles activates culturally embedded associations with authority\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e,\u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e58\u003c/span\u003e\u003c/sup\u003e. Therefore, academic expertise specifically acts as an established signal of credibility through institutional legitimation and formalized knowledge structures\u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e,\u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e,\u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e59\u003c/span\u003e\u003c/sup\u003e. In contrast, experience-based expertise represents an alternative form of knowledge legitimation through practical provenance\u003csup\u003e\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e\u003c/sup\u003e. The concept of a \u003cem\u003erecognition heuristic\u003c/em\u003e\u003csup\u003e\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e\u003c/sup\u003e suggests that the familiarity of academic titles can act as a signal of credibility. University students in particular are embedded in the academic realm and potentially used to accepting academic credentials as a sign of epistemic authority\u003csup\u003e\u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e60\u003c/span\u003e,\u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e61\u003c/span\u003e\u003c/sup\u003e. The incomplete separation of academic and personal expertise may have been processed more variably by the sample of Study 2, which might be more removed from academia. This could have potentially led to the non-significant findings in Study 2, while the student sample probably focused more on the academic credentials as a signal of epistemic authority, leading to the significant findings in Study 1.\u003c/p\u003e \u003cp\u003eHowever, the results have to be considered in the context of the study environment. The present studies used isolated online settings with controlled text presentations, whereas science communication often takes place in more complex, interactive media environments. Zimmermann et al.\u003csup\u003e27\u003c/sup\u003e, for example, demonstrated that on Instagram, political expertise significantly increased perceived credibility, while traditional source types had little influence. This highlights how much the context of the medium can influence expertise effects. Therefore, future studies should examine the robustness of the effects of different types of expertise in different digital media and communication environments.\u003c/p\u003e"},{"header":"6. Conclusion","content":"\u003cp\u003eOur findings support the notion that an author\u0026rsquo;s academic credentials provide a decisive advantage regarding the perceived credibility of a science communication text, while personal expertise does not add credibility in this context. The clear hierarchy between different forms of expertise confirms that not all expertise signals are equally effective. This clarifies previous assumptions about the effects of expertise on credibility perceptions and shows that the type of expertise, and not its presence in general, is crucial in science communication.\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eAcknowledgements / Funding statement\u003c/h2\u003e\n\u003cp\u003eThis research is part of the Metavorhaben \u0026ldquo;Digitalisierung im Bildungsbereich\u0026ldquo; (Digi-EBF ii) and was funded by the BMBFSFJ.\u003c/p\u003e\n\u003ch2\u003eAuthor contributions\u003c/h2\u003e\n\u003cp\u003eIn accordance with the CRediT taxonomy, the authors contributed to this paper as follows:\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eA.U.:\u003c/strong\u003e conceptualization, methodology, validation, formal analysis, investigation, data curation, writing (original draft), visualization, project administration. \u003cstrong\u003eL.P.:\u003c/strong\u003e conceptualization, formal analysis, investigation, data curation, writing (review and editing\u003cstrong\u003e)\u003c/strong\u003e. \u003cstrong\u003eC.S.:\u003c/strong\u003e conceptualization, methodology, formal analysis, investigation, data curation, writing (review and editing).\u003cstrong\u003e\u0026nbsp;K.P.:\u003c/strong\u003e conceptualization, methodology, formal analysis, investigation, data curation, writing (review and editing). \u003cstrong\u003eL.S.:\u003c/strong\u003e conceptualization, methodology, formal analysis, investigation, data curation, writing (review and editing). \u003cstrong\u003eM.E.:\u003c/strong\u003e conceptualization, methodology, formal analysis, investigation, data curation, writing (review and editing). \u003cstrong\u003eJ.K.:\u003c/strong\u003e conceptualization, methodology, resources, writing (review and editing), supervision, project administration, funding acquisition.\u003c/p\u003e\n\u003cp\u003eThe authors all read and approved this manuscript.\u0026nbsp;\u003c/p\u003e\n\u003ch2\u003eAdditional Information\u003c/h2\u003e\n\u003cp\u003eCompeting Interests Statement: The author(s) declare no competing interests.\u003c/p\u003e\n\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eThe datasets generated and analyzed during the current study are available in the Open Science Framework, https://osf.io/qjts4/overview?view_only=508b142e9ea54a08b9791056df455cf7\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eFischhoff, B. Evaluating science communication. \u003cem\u003eProc. Natl. Acad. Sci. USA\u003c/em\u003e. \u003cb\u003e116\u003c/b\u003e, 7670\u0026ndash;7675 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eH\u0026ouml;ttecke, D. \u0026amp; Allchin, D. Reconceptualizing nature-of-science education in the age of social media. \u003cem\u003eSci. Educ.\u003c/em\u003e \u003cb\u003e104\u003c/b\u003e, 641\u0026ndash;666 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePantic, M. \u0026amp; Ziek, P. Gatekeeping in a Digital Media Habitat: The Role of Secondary Gatekeepers. \u003cem\u003eElectron. News\u003c/em\u003e. \u003cb\u003e19\u003c/b\u003e, 94\u0026ndash;110 (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKrause, N. M., Freiling, I. \u0026amp; Scheufele, D. A. Our changing information ecosystem for science and why it matters for effective science communication. \u003cem\u003eProc. Natl. Acad. Sci.\u003c/em\u003e 122, e2400928121 (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOsborne, J. \u0026amp; Pimentel, D. Science education in an age of misinformation. \u003cem\u003eSci. Educ.\u003c/em\u003e \u003cb\u003e107\u003c/b\u003e, 553\u0026ndash;571 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eScharrer, L., Thomm, E., Stadtler, M. \u0026amp; Bromme, R. What Makes Sources Credible? How Source Features Shape Evaluation of Scientific Information. \u003cem\u003eJ. Exp. Educ.\u003c/em\u003e \u003cb\u003e0\u003c/b\u003e, 1\u0026ndash;24 (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOu, M. \u0026amp; Ho, S. S. Factors Associated With Information Credibility Perceptions: A Meta-Analysis. \u003cem\u003eJ. Mass. Commun. Q.\u003c/em\u003e \u003cb\u003e101\u003c/b\u003e, 346\u0026ndash;372 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTseng, S. \u0026amp; Fogg, B. J. Credibility and computing technology. \u003cem\u003eCommun. ACM\u003c/em\u003e. \u003cb\u003e42\u003c/b\u003e, 39\u0026ndash;44 (1999).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHanimann, A., Heimann, A., Hellmueller, L. \u0026amp; Trilling, D. Believing in Credibility Measures: Reviewing Credibility Measures in Media Research From 1951 to 2018. \u003cem\u003eInt. J. Commun.\u003c/em\u003e \u003cb\u003e17\u003c/b\u003e, 214\u0026ndash;235 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMetzger, M. J., Flanagin, A. J., Eyal, K., Lemus, D. R. \u0026amp; Mccann, R. M. Credibility for the 21st Century: Integrating Perspectives on Source, Message, and Media Credibility in the Contemporary Media Environment. \u003cem\u003eAnn. Int. Commun. Assoc.\u003c/em\u003e \u003cb\u003e27\u003c/b\u003e, 293\u0026ndash;335 (2003).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAppelman, A. \u0026amp; Sundar, S. S. Measuring Message Credibility: Construction and Validation of an Exclusive Scale. \u003cem\u003eJ. Mass. Commun. Q.\u003c/em\u003e \u003cb\u003e93\u003c/b\u003e, 59\u0026ndash;79 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFlanagin, A. J. \u0026amp; Metzger, M. J. The role of site features, user attributes, and information verification behaviors on the perceived credibility of web-based information. \u003cem\u003eNew. Media Soc.\u003c/em\u003e \u003cb\u003e9\u003c/b\u003e, 319\u0026ndash;342 (2007).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAnna) Kim, E., Shoenberger, H., Kwon, E. \u0026amp; Ratneshwar, S. (eds) (,(Penny) A narrative approach for overcoming the message credibility problem in green advertising. \u003cem\u003eJ. Bus. Res.\u003c/em\u003e 147, 449\u0026ndash;461 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eClark, J. K., Wegener, D. T., Sawicki, V., Petty, R. E. \u0026amp; Bri\u0026ntilde;ol, P. Evaluating the Message or the Messenger? Implications for Self-Validation in Persuasion. \u003cem\u003ePers. Soc. Psychol. Bull.\u003c/em\u003e \u003cb\u003e39\u003c/b\u003e, 1571\u0026ndash;1584 (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShamim, K. \u0026amp; Islam, T. Digital influencer marketing: How message credibility and media credibility affect trust and impulsive buying. \u003cem\u003eJ. Glob Sch. Mark. Sci.\u003c/em\u003e \u003cb\u003e32\u003c/b\u003e, 601\u0026ndash;626 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSlater, M. D. \u0026amp; Rouner, D. How Message Evaluation and Source Attributes May Influence Credibility Assessment and Belief Change. \u003cem\u003eJ. Mass. Commun. Q.\u003c/em\u003e \u003cb\u003e73\u003c/b\u003e, 974\u0026ndash;991 (1996).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWinter, S. \u0026amp; Kr\u0026auml;mer, N. C. A question of credibility \u0026ndash; Effects of source cues and recommendations on information selection on news sites and blogs. \u003cem\u003eCommunications\u003c/em\u003e \u003cb\u003e39\u003c/b\u003e, (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGottschling, S., Kammerer, Y., Thomm, E. \u0026amp; Gerjets, P. How laypersons consider differences in sources\u0026rsquo; trustworthiness and expertise in their regulation and resolution of scientific conflicts. \u003cem\u003eInt. J. Sci. Educ. Part. B\u003c/em\u003e. \u003cb\u003e10\u003c/b\u003e, 335\u0026ndash;354 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMehlis, K. Von der Sender- zur Nutzerqualit\u0026auml;t. in Journalismus und (sein) Publikum: Schnittstellen zwischen Journalismusforschung und Rezeptions- und Wirkungsforschung (eds (eds Loosen, W. \u0026amp; Dohle, M.) 253\u0026ndash;271 (Springer Fachmedien, Wiesbaden, doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/978-3-531-19821-7_14\u003c/span\u003e\u003cspan address=\"10.1007/978-3-531-19821-7_14\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKlebolte, K. Aspekte der Glaubw\u0026uuml;rdigkeitsbeurteilung \u0026ndash; Eine empirische Untersuchung zum Einfluss von Attraktivit\u0026auml;t, Expertise und Prominenz. \u003cem\u003eJr Manag Sci.\u003c/em\u003e \u003cb\u003e4\u003c/b\u003e, 524\u0026ndash;552 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIsmagilova, E., Slade, E., Rana, N. P. \u0026amp; Dwivedi, Y. K. The effect of characteristics of source credibility on consumer behaviour: A meta-analysis. \u003cem\u003eJ. Retail Consum. Serv.\u003c/em\u003e \u003cb\u003e53\u003c/b\u003e, 101736 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLis, B. \u003cem\u003eeWOM We Trust WIRTSCHAFTSINFORMATIK\u003c/em\u003e \u003cb\u003e55\u003c/b\u003e, 121\u0026ndash;134 (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEastin, M. S. Credibility Assessments of Online Health Information: the Effects of Source Expertise and Knowledge of Content. \u003cem\u003eJ. Comput. -Mediat Commun.\u003c/em\u003e \u003cb\u003e6\u003c/b\u003e, JCMC643 (2001).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJung, E. H., Walsh-Childers, K. \u0026amp; Kim, H. S. Factors influencing the perceived credibility of diet-nutrition information web sites. \u003cem\u003eComput. Hum. Behav.\u003c/em\u003e \u003cb\u003e58\u003c/b\u003e, 37\u0026ndash;47 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eThon, F. M. \u0026amp; Jucks, R. Believing in Expertise: How Authors\u0026rsquo; Credentials and Language Use Influence the Credibility of Online Health Information. \u003cem\u003eHealth Commun.\u003c/em\u003e \u003cb\u003e32\u003c/b\u003e, 828\u0026ndash;836 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMeinert, J. \u0026amp; Kr\u0026auml;mer, N. C. How the expertise heuristic accelerates decision-making and credibility judgments in social media by means of effort reduction. \u003cem\u003ePLOS ONE\u003c/em\u003e. \u003cb\u003e17\u003c/b\u003e, e0264428 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZimmermann, D., Klee, A. \u0026amp; Kaspar, K. Political news on Instagram: influencer versus traditional magazine and the role of their expertise in consumers\u0026rsquo; credibility perceptions and news engagement. \u003cem\u003eFront Psychol\u003c/em\u003e \u003cb\u003e14\u003c/b\u003e, (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAnders Ericsson, K., Towne, T. J. \u0026amp; Expertise \u003cem\u003eWIREs Cogn. Sci.\u003c/em\u003e \u003cb\u003e1\u003c/b\u003e, 404\u0026ndash;416 (2010).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003evan Antwerpen, N., Green, E. B., Sturman, D. \u0026amp; Searston, R. A. The impacts of expertise, conflict, and scientific literacy on trust and belief in scientific disagreements. \u003cem\u003eSci. Rep.\u003c/em\u003e \u003cb\u003e15\u003c/b\u003e, 11869 (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHocevar, K. P., Metzger, M. \u0026amp; Flanagin, A. J. Source Credibility, Expertise, and Trust in Health and Risk Messaging. in Oxford Research Encyclopedia of Communication (Oxford University Press, doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/acrefore/9780190228613.013.287\u003c/span\u003e\u003cspan address=\"10.1093/acrefore/9780190228613.013.287\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKnobloch-Westerwick, S., Glynn, C. J. \u0026amp; Huge, M. The Matilda Effect in Science Communication: An Experiment on Gender Bias in Publication Quality Perceptions and Collaboration Interest. \u003cem\u003eSci. Commun.\u003c/em\u003e \u003cb\u003e35\u003c/b\u003e, 603\u0026ndash;625 (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRossiter, M. W. The Matthew Matilda Effect in Science. \u003cem\u003eSoc. Stud. Sci.\u003c/em\u003e \u003cb\u003e23\u003c/b\u003e, 325\u0026ndash;341 (1993).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBundi, P. \u0026amp; Varone, F. Hanimann,Anina, Portmann, Leaand The future might be female: how does the public perceive experts? \u003cem\u003eJ. Eur. Public Policy\u003c/em\u003e 32, 843\u0026ndash;869 (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGorisek, C. \u003cem\u003eZusammenhang von Expertise, Reputation und Glaubw\u0026uuml;rdigkeit\u003c/em\u003e (Hochschule Rhein-Waal, 2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGreve-Poulsen, K., Larsen, F. K., Pedersen, R. T. \u0026amp; Alb\u0026aelig;k, E. No Gender Bias in Audience Perceptions of Male and Female Experts in the News: Equally Competent and Persuasive. \u003cem\u003eInt. J. Press.\u003c/em\u003e \u003cb\u003e28\u003c/b\u003e, 116\u0026ndash;137 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBigham, A. The Effect of Source Credibility and Gender on Message Persuasiveness. (2017). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://hdl.handle.net/2346/73515\u003c/span\u003e\u003cspan address=\"http://hdl.handle.net/2346/73515\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePratt, A. N., Tadlock, M. E., Watts, L. L., Wilson, T. C. \u0026amp; Denham, B. E. Perceptions of Credibility and Likeability in Broadcast Commentators of Women\u0026rsquo;s Sports. \u003cem\u003eJ. Sports Media\u003c/em\u003e. \u003cb\u003e13\u003c/b\u003e, 75\u0026ndash;97 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJones, C. W. \u0026amp; Mitchell, J. S. Women reporters as experts on security affairs in Jordan? Rethinking gender and issue competency stereotypes. \u003cem\u003eMediterr. Polit\u003c/em\u003e. \u003cb\u003e28\u003c/b\u003e, 434\u0026ndash;462 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLeiner, D. J. SoSci Survey. (2025). Available at \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.soscisurvey.de\u003c/span\u003e\u003cspan address=\"https://www.soscisurvey.de\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOpenAI. ChatGPT. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://chat.openai.com/\u003c/span\u003e\u003cspan address=\"https://chat.openai.com/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCollins, H. Studies of Expertise and Experience. \u003cem\u003eTopoi\u003c/em\u003e \u003cb\u003e37\u003c/b\u003e, 67\u0026ndash;77 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWeinstein, B. D. What is an expert? \u003cem\u003eTheor. Med.\u003c/em\u003e \u003cb\u003e14\u003c/b\u003e, 57\u0026ndash;73 (1993).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWagemans, J. H. M. The Assessment of Argumentation from Expert Opinion. \u003cem\u003eArgumentation\u003c/em\u003e \u003cb\u003e25\u003c/b\u003e, 329\u0026ndash;339 (2011).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHornikx, J. \u0026amp; Hoeken, H. Cultural Differences in the Persuasiveness of Evidence Types and Evidence Quality. \u003cem\u003eCommun. Monogr.\u003c/em\u003e \u003cb\u003e74\u003c/b\u003e, 443\u0026ndash;463 (2007).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBurgers, C., de Graaf, A. \u0026amp; Callaars, S. Differences in actual persuasiveness between experiential and professional expert evidence. \u003cem\u003eJ. Argum Context\u003c/em\u003e. \u003cb\u003e1\u003c/b\u003e, 194\u0026ndash;208 (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFerreira, P. A., Wingrove, T. \u0026amp; and Expert Witness Training History and Professional Experience Exert Separable Impacts on Expert Credibility Perceptions. \u003cem\u003eJ. Forensic Psychol. Res. Pract.\u003c/em\u003e \u003cb\u003e24\u003c/b\u003e, 524\u0026ndash;540 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eR Core Team. \u003cem\u003eR: A Language and Environment for Statistical Computing\u003c/em\u003e (R Foundation for Statistical Computing, 2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePetty, R. E. \u0026amp; Cacioppo, J. T. The Elaboration Likelihood Model of Persuasion. in \u003cem\u003eAdvances in Experimental Social Psychology\u003c/em\u003e vol. 19 123\u0026ndash;205Academic Press, (1986).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePornpitakpan, C. The Persuasiveness of Source Credibility: A Critical Review of Five Decades\u0026rsquo; Evidence. \u003cem\u003eJ. Appl. Soc. Psychol.\u003c/em\u003e \u003cb\u003e34\u003c/b\u003e, 243\u0026ndash;281 (2004).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHilligoss, B. \u0026amp; Rieh, S. Y. Developing a unifying framework of credibility assessment: Construct, heuristics, and interaction in context. \u003cem\u003eInf. Process. Manag\u003c/em\u003e. \u003cb\u003e44\u003c/b\u003e, 1467\u0026ndash;1484 (2008).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMetzger, M. J. \u0026amp; Flanagin, A. J. Credibility and trust of information in online environments: The use of cognitive heuristics. \u003cem\u003eJ. Pragmat.\u003c/em\u003e \u003cb\u003e59\u003c/b\u003e, 210\u0026ndash;220 (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSundar, S. S. The MAIN Model: A Heuristic Approach to Understanding Technology Effects on Credibility. (2008).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYalch, R. F. \u0026amp; Elmore-Yalch, R. The Effect of Numbers on the Route to Persuasion. \u003cem\u003eJ. Consum. Res.\u003c/em\u003e \u003cb\u003e11\u003c/b\u003e, 522 (1984).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eReinhard, M. A. \u0026amp; Sporer, S. L. Content versus source cue information as a basis for credibility judgments: The impact of task involvement. \u003cem\u003eSoc. Psychol.\u003c/em\u003e \u003cb\u003e41\u003c/b\u003e, 93\u0026ndash;104 (2010).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSundar, S. S., Knobloch-Westerwick, S. \u0026amp; Hastall, M. R. News cues: Information scent and cognitive heuristics. \u003cem\u003eJ. Am. Soc. Inf. Sci. Technol.\u003c/em\u003e \u003cb\u003e58\u003c/b\u003e, 366\u0026ndash;378 (2007).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLee, J. Y. \u0026amp; Sundar, S. S. To Tweet or to Retweet? That Is the Question for Health Professionals on Twitter. \u003cem\u003eHealth Commun.\u003c/em\u003e \u003cb\u003e28\u003c/b\u003e, 509\u0026ndash;524 (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWathen, C. N. \u0026amp; Burkell, J. Believe it or not: Factors influencing credibility on the Web. \u003cem\u003eJ. Am. Soc. Inf. Sci. Technol.\u003c/em\u003e \u003cb\u003e53\u003c/b\u003e, 134\u0026ndash;144 (2002).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLittle, D. \u0026amp; Green, D. A. Credibility in educational development: trustworthiness, expertise, and identification. \u003cem\u003eHigh. Educ. Res. Dev.\u003c/em\u003e \u003cb\u003e41\u003c/b\u003e, 804\u0026ndash;819 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBrown, D. K. The Social Sources of Educational Credentialism: Status Cultures, Labor Markets, and Organizations. \u003cem\u003eSociol. Educ.\u003c/em\u003e \u003cb\u003e74\u003c/b\u003e, 19\u0026ndash;34 (2001).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBokros, S. E. A deference model of epistemic authority. \u003cem\u003eSynthese\u003c/em\u003e \u003cb\u003e198\u003c/b\u003e, 12041\u0026ndash;12069 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePierson, R. The Epistemic Authority of Expertise. \u003cem\u003ePSA Proc. Bienn. Meet. Philos. Sci. Assoc.\u003c/em\u003e 398\u0026ndash;405 (1994). (1994).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBigham, A., Meyers, C., Li, N. \u0026amp; Irlbeck, E. The Effect of Emphasizing Credibility Elements and the Role of Source Gender on Perceptions of Source Credibility. \u003cem\u003eJ Appl. Commun\u003c/em\u003e \u003cb\u003e103\u003c/b\u003e, (2019).\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"},{"header":"Tables","content":"\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"589\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eTable 1\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"5\" valign=\"top\"\u003e\n \u003cp\u003e\u003cem\u003eSample size, gender, age and educational distribution for Studies 1-3\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eStudy 1\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eStudy 2\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eStudy 3\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eN\u003csub\u003etotal\u003c/sub\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e203\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e182\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e206\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eGender\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eFemale\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e139 (68.5%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e70 (38.5%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e50 (24.3%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eMale\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e61 (30.1%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e110 (60.5%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e154 (74.8%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eNon-binary\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e3 (1.5%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1 (0.6%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e2 (1.0%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eOther\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0 (0.0%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eNot indicated\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0 (0.0%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1 (0.6%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0 (0.0%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eAge (years)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eMean (SD)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e30.8 (13.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e43.5 (12.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e33.1 (10.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eRange\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e18-72\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e19-77\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e18-72\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eEducational level\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eUniversity degree\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e98 (48.3%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e86 (47.3%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e92 (44.7%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eTechnical college\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e2 (1.0%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e19 (10.5%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e18 (8.7%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eHigh school diploma\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e97 (47.8%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e37 (20.3%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e58 (28.2%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eCompleted apprenticeship\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e5 (2.5%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e26 (14.3%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e23 (11.2%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eSecondary school\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1 (0.5%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e11 (6.0%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e12 (5.8%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eSecondary modern school\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0 (0.0 %)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1 (0.6%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e3 (1.5%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"credibility, expertise, science communication, academic expertise, personal expertise","lastPublishedDoi":"10.21203/rs.3.rs-9010835/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9010835/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eEvaluating message credibility is critical for understanding scientific information and informed decision making. People\u0026rsquo;s ability to assess the credibility of science communication texts is influenced by several factors. In three experiments, we examined the impact of authors\u0026rsquo; gender as well as different types of expertise on readers\u0026rsquo; credibility perceptions. In Experiment 1 (n\u0026thinsp;=\u0026thinsp;203), we varied authors\u0026rsquo; alleged gender and expertise. We found that a text allegedly written by a high-expertise author was rated as more credible than the same text by a low-expertise author. There was no effect of authors\u0026rsquo; gender. Experiment 2 (n\u0026thinsp;=\u0026thinsp;182) was a replication of the first study with a different sample. We did not find any effects of gender or expertise on perceived message credibility. In Experiment 3 (n\u0026thinsp;=\u0026thinsp;206), we differentiated between academic and personal expertise and manipulated these types of expertise independently. The data indicated that academic expertise was a significant positive predictor of perceived message credibility, while personal expertise was not. We discuss these findings in terms of their significance for the trustworthiness of science communication and examine the relevance of differentiating the concept of expertise in this context.\u003c/p\u003e","manuscriptTitle":"Authors’ academic but not personal expertise affects message credibility in science communication: A series of experiments on text perception","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-03-25 09:26:42","doi":"10.21203/rs.3.rs-9010835/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2026-04-23T05:04:08+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-04-21T18:00:29+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"285783130513962030275928327187452858046","date":"2026-04-16T20:05:26+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-04-01T00:27:12+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"138418595891172943794493099177518352919","date":"2026-03-13T08:57:15+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"257556012810349174611212034254883945667","date":"2026-03-10T21:33:57+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-03-10T14:43:39+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-03-10T13:47:06+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2026-03-06T14:12:53+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-03-05T20:31:01+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2026-03-05T13:48:04+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"51491de4-560a-4239-aaa3-adc23801ee2d","owner":[],"postedDate":"March 25th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2026-05-06T23:08:14+00:00","versionOfRecord":[],"versionCreatedAt":"2026-03-25 09:26:42","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9010835","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9010835","identity":"rs-9010835","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}