Abstract
Across a wide range of issues and a variety of sources, scientific information often gets lost in translation, failing to properly inform, educate, and engage publics in a meaningful way. But science communication is like a wine—it is better when it is outside the box. One promising tactic for more effective scientific communication is the use of humor to both engage and educate less interested and knowledgeable citizens on important topics. Here, we investigate whether the identity of a speaker influences perceptions of their effectiveness at connecting with and engaging public audiences. We conducted a 2 (Gender: Female vs. Male) × 2 (Race: Black Speaker vs. White Speaker) × 2 (Credentials: Scientist vs. Comedian) between-subjects experiment and found that race and gender of the source, relative to their credentials, were not significant factors for predicting the perceptions of communicator effectiveness. We also found that experienced humor, or mirth, moderated the relationship between the speaker’s credentials and perceived effectiveness. We discuss the implications of our findings for science communication research and practice.
Climate disasters are like reality TV dating shows: They are increasing in volume, and we are the cause. Across the globe, drought, floods, and severe weather systems exacerbated by climate change have devastated communities. Then, in 2020, the coronavirus (COVID-19) pandemic emerged like a surprise contestant on The Bachelorette vying for our attention. The directive to “follow the science” became an important mantra as we attempted to navigate these challenges safely.
Partly as a means of stemming the growing tide of misinformation, disinformation, and fake news surrounding key public health debates and the global climate crisis, politicians, journalists, and celebrities alike encouraged a focus on sharing and relying upon scientific information when making important personal and community-level decisions. Of course, not all scientific information is created and disseminated equally. Scientific information can be complex and difficult for the average citizen to process and understand (McNutt, 2013). Scientists are not known as the most natural of communicators; we can struggle with communicating technical information in clear, concise, and consistent ways (Besley et al., 2018). Across a wide range of issues and a variety of sources, scientific information often gets lost in translation, failing to properly inform, educate, and engage public audiences in a meaningful way.
Science communication is like a wine—it is better when it is outside the box. One promising tactic for more effective scientific communication is the use of humor to both engage and educate less interested and knowledgeable citizens on important topics. For example, previous research has shown that exposure to humor focused on climate change reinforces existing beliefs, belief certainty, and the perceived risks associated with global warming, particularly among those who are initially less interested in the issue (Anderson & Becker, 2018; Brewer & McKnight, 2015, 2017). Political satire that addresses the issue of climate change is not only increasingly pro-science (Feldman, 2017) but it also reduces the partisan gap in risk perceptions related to climate change, increasing intentions to engage in activism related to the issue (Skurka et al.,2019, 2018, 2022). While most of the focus on humor and climate change communication has considered the effect of exposure to political comedy content from comedic sources like Stephen Colbert, John Oliver, Jimmy Kimmel, and even the pages of The Onion (e.g., Anderson & Becker, 2018), more recent work has started to consider what happens when scientists use jokes, humor, and comedy to try and convey scientific information on a range of topics in a funny and amusing way (e.g., Cacciatore et al., 2020).
For example, recent research has shown that watching a scientist tell jokes in an effort to convey scientific information in an engaging way is seen as a valid source of science information, particularly when audience members find the comedy humorous (Yeo, Anderson, et al., 2020). In fact, the funnier the scientist is, the greater their perceived scientific expertise, which is positively associated with perceptions that the comedy they are sharing is a valid source of scientific information (Yeo, Anderson, et al., 2020). But is this science communication tactic effective? Effective communication requires trust, a concept that has several dimensions, including competence, integrity, warmth, and openness (Besley et al., 2018). Scientists are often perceived as competent by lay audiences; what is often lacking in public perceptions of scientists is warmth or benevolence (Fiske & Dupree, 2014).
Effectiveness in communication has myriad meanings and is often dependent on the objectives and goals of a communicator. Indeed, this is how others have defined this elusive concept (K. M. Leonard et al., 2009). As there is no single definition of communicator effectiveness in the literature, we restrict its definition to the context of our study, that is, science, and use a relatively simple definition. In the present work, we conceptualize effectiveness as a combination of the following: (a) the extent to which a source can connect with public audiences around science, (b) make science interesting to publics, and (c) persuade them of its value.
Importantly, a fundamental question has yet to be considered in research on the effectiveness of humorous scientific communication: How does the identity of a speaker influence perceptions of their effectiveness at connecting with and engaging public audiences? Here, we use a between-subjects experiment to examine how a speaker’s credentials, race, and gender may matter when it comes to perceptions of speaker effectiveness in a video featuring stand-up science comedy. Accordingly, our review of the extant literature first covers how a communicator’s identity might affect audience perceptions of their effectiveness. Then, we review scholarship on humor as a tactic for science communication and how gender, race, and expertise interact with experienced humor, or mirth, 1 to affect audiences’ perceptions of effectiveness.
Considering the Relationships Between Communicator Identity and Effectiveness
It is no secret that certain populations, based on gender and race, are underrepresented in science. In 2007, women accounted for 26% of science employment in the United States and, a decade later, this was only marginally higher, 29% (National Science Board, 2020). The data on racial groups also evince a pervasive bias; in 2017, the share of black and Hispanic individuals in the science workforce was 5.6% and 7.5%, respectively (National Science Board, 2020). As a point of comparison, according to the U.S. Census Bureau’s 2020 American Community Survey, 13.6% of the American population identifies as black and 18.9% identifies as Hispanic or Latino (United States Census Bureau, 2022). This underrepresentation, combined with the persistent prototypical view of scientists as white males (Ferguson & Lezotte, 2020), a point to which we will return in our review of the literature, leads us to examine gender and race as critical characteristics when it comes to scientists’ communication efforts.
Before delving into the literature, however, it is worth noting that our study and results are confined to the geographic and cultural context of the United States as our sample consists only of American adults. That said, the relationships that we examine in the present work may differ in other contexts but the study we have conducted is U.S.-based, and we examine the literature that is most relevant to this situation.
Gender
The characteristics of effective communicators have been debated since the time of Aristotle (Cooper, 1932; Umeogu, 2012), with scholarly research on audience perceptions of male vs. female speakers going back to at least the mid-20th century (Hamilton et al., 1977; Whittaker, 1965). Hovland et al. (1953) defined source credibility as “expertise” and “trustworthiness,” with expertise akin to the knowledge or understanding the communicator has about the topic in question, and trustworthiness having to do with the intentions of the communicator toward their audience. Later, Kenton (1989) proposed a four-dimension model of source credibility that includes (a) perceived “prestige” or rank of the communicator, (b) verbal abilities or platform skills (self-presentation), which were added to the two existing dimensions of (c) goodwill and fairness and (d) expertise. Importantly, Kenton (1989) also argued that men and women have relative areas of strength and weakness—whether real or perceived—when it comes to these four dimensions of credibility, an argument that was supported in subsequent studies (Cameron, 1997; Cheshire & Trudgill, 1998; Coates, 1996). Some scholars have even suggested that “the difference approach” studies of the 1980s and 1990s—which focused largely on differences between the sexes—contributed to audience expectations and stereotypes about how men and women communicate (Holmes, 2006; Talbot, 2003).
There are data that show female sources are rated more highly than male sources (e.g., Bigham et al., 2019; Todd & Melancon, 2018), yet other research shows that people will rate source credibility higher for individuals of the opposite sex (Flanagin & Metzger, 2003). Traditionally, women have been seen as fair communicators who are more likely to show concern for others in their communications (Pyke & Johnson, 2003; Schippers, 2007), with men more often described as being competitive and aggressive in their communications (Connell, 2005; Connell & Messerschmidt, 2005; Kenton, 1989). Thus, women are expected to be rated more highly on the goodwill and fairness dimension of being an effective communicator. Conversely, he argued that men are typically rated more highly on knowledge or expertise when people are asked to rate communicator performance (Cuddy et al., 2008; Kenton, 1989). Interestingly, this appears to be driven largely by different expectations or standards for male and female communicators (Aries, 1987; Holmes, 2006; Lindvall-Östling et al., 2020; Talbot, 2003). For example, Bradley (1981) found that women who argued without proper supporting evidence or who used rhetorical questions like “don’t you agree?” were rated more harshly on their intelligence than men who employed the same communication or rhetorical tactics, while Ladegaard (2011) showed that even when men and women used the same preferred communication management styles, female leaders were more often challenged or questioned than male leaders.
Kenton (1989) argued that men also experienced advantages over women in terms of the power they hold when communicating. For instance, he noted that men are often rated as higher in prestige even when their levels of experience, training, and ability are equal to the women to whom they are being compared. Other work links higher prestige levels in men to their tendencies to utilize “power moves” when communicating—such as dominating interpersonal conversations (Connell, 2005; Leet-Pellegrini, 1980), controlling the direction of those conversations (Baird, Jr, & Bradley, 1979; Mohindra & Azhar, 2012), and being more willing to interrupt others (Connell & Messerschmidt, 2005; Zimmerman & West, 1975).
In terms of the self-presentation dimension, men are rated as being louder and more forceful, aggressive, and frank than women, among a host of other characteristics, when communicating (Cheshire & Trudgill, 1998; Kramer, 1977). Perhaps more importantly, men often self-describe themselves as especially confident in their communications, which includes attributing their successes to their own abilities, while women are more likely to attribute similar success to luck or contextual factors (Maccoby & Jacklin, 1977; Mohindra & Azhar, 2012). In fact, this gap in confidence is not always limited to judgments of their own work but also extends to work done by other women. For example, women have been found to not only rate traditionally masculine qualities as higher than traditionally feminine qualities but to also rate the work attributed to men to be better than that attributed to women (Kenton, 1989).
Of course, scholars have questioned models of gendered speech characteristics based on basic dichotomies of cooperative (female) versus competitive (male) elements, arguing that speech is more complex than simple dichotomies allow for (Cameron, 1997; Eckert, McConnell-Ginet, 2013). For example, context is believed to play a role in audience reception of communication, with characteristics like greater perceived involvement or interest in an issue possibly signaling greater expertise or knowledge of the topic. As Gainsburg and Sekaquaptewa (2020) note, female speakers might be perceived as having greater expertise on issues related to topics like gender equality. In the context of the present work, it is worth mentioning again that there is a large gap in the number of men vs. women who work in science, technology, engineering, and math (STEM) fields. And while reliable data about careers in comedy are difficult to come by, it is generally accepted that men far outnumber women in comedy, with some estimating that as much as 90% of all comedians are men (Levitt, 2018; though it is worth noting that there has recently been a proliferation of high-profile female comedians in mainstream media).
Overall, while the early research by Kenton (1989) argues that men hold perceived advantages in three of the four dimensions of source credibility, thereby leading audiences to often rate men higher than women when they are asked about their communications, more recent work suggests that contextual factors may play a powerful role in shaping such perceptions. Nevertheless, the gaps in both stand-up comedy and scientific workforce representation suggest that gender representation across both comedy and STEM fields is likely to advantage men, perhaps signaling an expectation that men will perform better when delivering stand-up comedy set focused on science. Therefore, this leads us to propose the following hypothesis:
Hypothesis 1 (H1): Relative to those who viewed a female speaker, subjects who watched the video with a male speaker will perceive the source as a more effective communicator.
Race
Like the perceived biases surrounding female and male communicators, a speaker’s race has been documented as a factor that influences audience perceptions. Research has found that audiences often rely on racialized stereotypes when forming judgments of different groups of communicators. For instance, work in the educational space has found that audiences often classify black speakers as “performers,” resulting in expectations of how energetic or lively a given lecture or speech should be (McGee & Kazembe, 2016). Ogawa (1971) found that white respondents stereotyped communication by black individuals as straightforward, argumentative, emotional, aggressive, critical, ostentatious, defiant, and hostile but also as sensitive, open, responsive, and intelligent.
The mixed pattern of stereotypes noted above was largely confirmed by a later study by Leonard and Locke (1993). However, that work also found black respondents often depicted white speakers as demanding, manipulative, organized, rude, critical, aggressive, arrogant, boastful, hostile, ignorant, deceptive, and noisy. Hughes and Baldwin (2002) replicated the work of Ogawa (1971) and Leonard and Locke (1993) and found evidence that many of the stereotypes had persisted, but that both white and black audiences were overall more positive in how they viewed communicators of the other race. For instance, “organized” emerged as the most highly rated perception of white communicators from black respondents, while “friendly” was the highest rated characteristic of black communicators among white respondents. They also noted that the hierarchy of stereotypes seemed likely to shift depending on context.
In a study done by Popp and colleagues (2003), white college students were asked to rate the speech style of either a fictional black male, black female, white male, or white female. The respondents tended to classify the words attributed to black speakers as being more direct, dominant, and emotional, and less socially appropriate and playful than equivalent speech that was attributed to white speakers. The authors further found that the effects of race were stronger than the effects of gender and speculated that communication style might be judged based on whether race or gender was more salient in a given situation. A replication of this work by Carlson and Crawford (2012) reproduced several of the findings, with black communicators rated as more emotional, lower in social appropriateness, and less articulate that their white counterparts. However, in a departure from the earlier work, the white communicators were rated as more dominant than the black communicators.
As the findings noted above suggest, and echoing results shared in the earlier section on gender, communication is more complex than simple patterns identified based on a demographic attribute. And once again, it is likely that context can influence the perceptions of communicator effectiveness. For instance, some have suggested that under-represented minorities are likely to enjoy higher levels of perceived credibility and knowledge on issues related to race, thereby making them especially effective communicators on matters in that space (Tatum & Sekaquaptewa, 2009). However, others note that being closely tied to an area of interest can hurt one’s credibility as a communicator as their arguments may be tied to their personal self-interest in the matter (Hekman et al., 2017).
Of particular interest to the present study is the context of science information being shared within a comedy act. To better understand how this specific context might influence audience perceptions, it is important to again look at representation in the fields of comedy and STEM. We have already acknowledged the bias in black representation within the U.S. scientific workforce, and unfortunately, reliable data of the racial diversity in comedy are lacking. However, one career website estimates white comics to make up approximately 60% of the vocation, with black comics represented at 13.5% (Anonymous, 2019), percentages that do not differ dramatically with the broader demographics of the United States (United States Census Bureau, 2022), but that do suggest that black comics are more the exception rather than the rule.
Nevertheless, given the literature and the context-dependent nature of perceived stereotypes, we opted to propose a research question rather than a hypothesis about the impact of communicator race on audience perceptions of their effectiveness:
Research Question 1 (RQ1): Will there be differences in perceived effectiveness of the speaker based on the speaker’s race?
Credentials and Expertise
As noted, Kenton (1989) argues that perceived prestige plays an important role in influencing the credibility of a communicator. This idea of prestige is particularly relevant to the third experimental manipulation in our work, which focuses on the official credential (scientist or comedian) we provided for the speakers in the video. On the one hand, even as the popular press decries the seemingly growing public skepticism toward science and scientists (Jewett, 2020), it is worth noting that scientists enjoy a privileged position as one of the more trusted groups in society (Pew Research Center, 2020). Furthermore, while cultivation studies in the 1980s lamented the largely negative depictions of scientists on television—noting they often had sinister or evil motivations, among other things (Gerbner, 1987; Gerbner et al., 1981, 1985)—more recent work has described scientist depictions on TV as, at best, positive, and at worst, mixed (Dudo et al., 2011). Thus, it is perhaps the case that scientists may have an inherent credibility advantage as communicators.
On the other hand, despite the shift in popular portrayals of science and scientists, it is still worth noting that scientists are not generally known to be effective communicators. To this day, one of the most common depictions of scientists is that of an anti-social or eccentric loner (Ferguson & Lezotte, 2020). Arguably the most viewed popular culture representation of scientists among mainstream audiences—the CBS sitcom The Big Bang Theory—while sometimes praised for presenting a group of scientists who are more diverse in gender and ethnicity than other popular culture depictions, has still been criticized for its portrayal of scientists as nerds, constructing the image of a nerd as “rude” or “not quite human” (Bednarek, 2012), with some likening those portrayals to a form of blackface (Weitekamp, 2017). Such stereotypes built around socially awkward loners, even if they have been shaped by a television comedy program, might prime audiences to expect an unfunny set of jokes when they are met with a scientist in a stand-up comedy routine.
Like scientists, comedians also enjoy a unique place in society as communicators. In recent years, the growth of satirical news programs, such as The Daily Show and The Colbert Report, along with the political slant of most late-night talk shows has helped engage audiences with the major political issues of the day. Some have even described comedic treatments of serious political topics as serving as a gateway toward encouraging audiences to pay more attention to traditional news media accounts of those same topics (Baum, 2005; Feldman et al., 2011; Xenos & Becker, 2009). There is also evidence that audiences are less likely to counterargue even the serious information to which they are exposed through comedy, and instead, rely on heuristic cues, such as the affection they have for a confident, likable, or funny communicator (Nabi et al., 2007). Indeed, recent work has found that comedians can serve the dual functions of being credible and likable communicators (Borum, Chattoo, & Green-Barber, 2021).
However, not all research points toward comedians being viewed as credible or even effective communicators. Comedy (and comedians) poses the risk of trivializing topics. For instance, one study tested a humor strategy to teach audiences about the issue of teen pregnancy and found that those who were exposed to the humorous intervention viewed teen pregnancy as a less serious issue than those who were exposed to communication with a more serious tone (Moyer-Gusé et al., 2011). Given the conflicting evidence concerning how audiences view comedians as communicators of serious topics, we propose the following research question:
Research Question 2 (RQ2): Will there be differences in perceived effectiveness of the communicator based on the speaker’s credentials?
Humor and Source Characteristics
Given the attention to humor in science communication literature, this study places the speaker characteristics described above in the context of how funny one is perceived to be. Humor is a desirable source characteristic (Martin, 2010). Humorous qualities in an individual involve a number of things, such as an appreciation for and comprehension of humor, but also include the way individuals express humor and mirth (Hehl & Ruch, 1985). In other words, how funny others perceive an individual to be is connected to likability and perceived source credibility (Nabi et al., 2007; Weinberger & Gulas, 1992). Scholarship in science communication has repeatedly shown experienced humor to have positive effects on science communication and engagement outcomes (Cacciatore et al., 2020; Yeo, Anderson, et al., 2020; Yeo, Su, et al., 2020), and we expect to find similar patterns here:
Hypothesis 2 (H2): Subjects who report higher levels of mirth will perceive the speaker to be a more effective communicator.
However, one’s mirth is not an objective experience. This subjective experience of mirth combined with cues about the source may shape viewer perceptions of communicator effectiveness. Specifically, we are interested in how a speaker’s gender, race, and expertise (in the form of credentials) might interact with experiences of humor. We, therefore, turn to the disposition theory of humor (Zillmann & Cantor, 1976), which is one among many humor theories, though it is the most appropriate for our examination of humor and race.
Following the perspective of the disposition theory of humor (Zillmann & Cantor, 1976), or that people respond to humor differently based on their in- or out-group, some studies have looked at how people respond to humor focused on a particular racial group. While this type of humor can heighten emotional responses (Banjo et al., 2017), people tend to be more accepting of racially disparaging jokes if the joke is targeted at a race to which the speaker belongs (Thai et al., 2019). Research on the role of a speaker’s race outside of the disparaging context is more limited. One study, composed only of white respondents, found that people view comedians who are black more favorably and think they are funnier than white comedians (Walther-Martin, 2015). The limited evidence, thus, shows that people who are not among the majority group in society make evaluations about sources based on their identification with the group. While there is limited evidence that the majority group (white people) perceives the non-majority group (black people) to be relatively funnier, white people tend not to form source perceptions differently based on the race of the source.
Research on the source characteristic of gender also demonstrates mixed results, as our above review of the literature clearly shows. When it comes to humor, there is a common perception that men are funnier than women despite the proliferation of mainstream female comedians such as Tina Fey, Tiffany Haddish, Amy Poehler, Amy Schumer, and Ali Wong, among others. Research supporting this perception, however, is quite mixed. Studies have found that both men and women are more likely to rate cartoons written by men as funnier than those written by women (Mickes et al., 2012). Another study found that while participants reported that men’s cartoons were funnier than women’s, they rated women’s captions slightly higher in the absence of a gender label (Hooper et al., 2016). Other research shows that jokes by women who hold high self-efficacy in their capacity to be funny are rated as funnier than men’s jokes (Caldwell & Wojtach, 2020). More recently, Becker (2022) found that gender of a comic host did not interact with viewers’ gender to affect message-related outcomes such as informativeness of a message and message discounting.
Our third source characteristic concerns a speaker’s credentials and, thus, expertise. Perceptions of scientists are often formed early in life, based on media depictions of scientists that are strange and detached from everyday life (Potts & Martinez, 1994; Weingart et al., 2003). Stereotypes of scientists are not grounded in warmth or friendliness (Fiske & Dupree, 2014), while comedians are seen as extroverted, agreeable, and open (Greengross & Miller, 2009; Jarreau et al., 2019). Comedians are, therefore, likely to be perceived as funnier than scientists. However, our measures of communicator effectiveness are specifically related to connecting and engaging with public audiences in the context of science. In this realm, scientists, as a relatively well-trusted source (Pew Research Center, 2020), may hold the advantage when it comes to effectiveness at communicating science issues. It remains to be seen whether experienced mirth in these different sources—scientists or comedians—shapes the perceived effectiveness of the communication differently.
Our review of the literature leads us to two conclusions. First, the source characteristics of gender, race, and credentials, combined with experiences of mirth, are likely to influence people’s perceptions of communicator effectiveness. Yet, we are unsure of the specific ways in which these factors affect perceived effectiveness. Therefore, we pose the following, three-part research question:
Research Question 3 (RQ3): Does experienced mirth moderate the effect of the speaker’s (a) gender, (b) race, and (c) credentials on perceived communicator effectiveness? If so, how?
Our second conclusion from the extant literature is that viewer characteristics are likely to also moderate the effect of source characteristics combined with mirth on perceived communicator effectiveness. However, given our focus on perceptions of communicator effectiveness and source identity characteristics, we opted to focus our experiment on elucidating those interactions instead of those that include viewer characteristics. Moreover, our sample size does not allow us to test these interactions given the number of experimental conditions.
Method
A 2 (Gender: Female vs. Male) × 2 (Race: Black Speaker vs. White Speaker) × 2 (Credentials: Scientist vs. Comedian) between-subjects experiment was administered via Amazon’s MTurk Human Intelligence Tasks (MTurk) platform in July 2021 (N = 487). Following a pre-test questionnaire that measured concepts such as attention to media and interest in science, subjects were randomly assigned to view one of eight experimental stimuli: a video clip of a stand-up comedy routine about science. The jokes were identical in all eight videos, but the race, gender, and credentials of the communicator varied (a chyron 2 identified the speaker as a scientist or a comedian).
To maximize ecological validity, the jokes were written by one of the co-authors, who is a professional science comedian, and selected based on their positive responses from audiences at in-person shows and an analysis based on the duration of laughter (in seconds) produced. The same set was performed by four professional comedians (a black female, a black male, a white female, and a white male), who were each coached to ensure joke delivery was uniform in pace, tone, emphasis, and cadence. Joke content consisted of neutral statements about the Earth, scientific studies, and wordplay with scientific jargon. Importantly, the jokes did not directly address current scientific controversies, such as those associated with COVID-19, anti-vaxxers, or climate denialism. This “neutral” content was designed to test comedy as a means of communicating scientific information while entertaining audiences.
All videos were approximately 2 minutes in length; subjects had to remain on the video page for the duration. A series of post-test questions followed, including experienced humor and perceptions about the speaker in the video.
Measures
Perceived effectiveness of the speaker (M = 5.40, SD = 1.08, Cronbach’s α = .80) was an index of four items measured on a 7-point Likert-type scale (1 = “Not at all,” 7 = “Extremely”) asking subjects how effective they thought the speaker was at: (a) connecting with the public, (b) persuading the public about the value of science, (c) sharing research findings with the public, and (d) making science interesting for the public.
The experimental conditions were included in the regression model as three dichotomous variables in which speakers who were white, female, and identified as a scientist were coded high. An index of mirth (M = 5.36, SD = 1.25, Cronbach’s α = .74) was created by averaging three items on 7-point semantic differential scales; subjects indicated whether the video they watched was “not funny-funny,” “not humorous-humorous,” and “not amusing-amusing” (Nabi et al., 2007; Yeo, Su, et al., 2020).
Demographic variables were also included in the analysis. Age was calculated from subjects’ year of birth (M = 37.2, SD = 10.5) and sex was a dichotomous variable with female coded high (35.5% female). As 78.6% of our sample consisted of white individuals, we dichotomized race into two categories, non-white and white, with respondents in the white category coded high. Education was measured by asking subjects to self-identify into one of seven education groups ranging from “some high school” to “graduate degree” (Median: Bachelor’s degree).
Data Analysis
Data were analyzed in R (version 4.1.1; “Funny-Looking Kid”). We first examined the mean levels of mirth among our eight experimental conditions (Figure 1). Then, to address our hypotheses and research questions, we conducted hierarchical ordinary least squares regression and entered the variables into the model in assumed causal order. The constituent variables of the interaction terms were centered prior to entering them into the model. Interaction terms were entered simultaneously into the model. Variables were entered into the model in the following blocks:
Experimental conditions (speaker gender, race, and credentials).
Viewers’ demographic variables (age, gender, race, education).
Experienced mirth (mirth).
Interaction terms (Speaker Gender × Mirth, Speaker Race × Mirth, Speaker Credentials × Mirth).
We then used the PROCESS add-on for R (Hayes, 2022) to probe significant interactions via floodlight analysis (Spiller et al., 2013) using Model 1, which is a simple moderation model. PROCESS is a computational tool for conducting moderation and mediation analysis based on path analysis (Hayes, 2022). The interactions were probed at the mean score of mirth and one standard deviation above and below the mean.
Mean and standard errors of mirth for each experimental condition.
Results
The variables in the regression model accounted for 41.0% of the variance in perceived effectiveness of the speaker (Table 1). We found no evidence to support the idea that males are seen as more effective science communicators than females (H1; B = .001, standard error [SE] = .076, p = .993). Similarly, the main effects of the race and credential manipulations on perceived effectiveness were not significant (RQ1 and RQ2; B = .001, SE = .078, p = .991 and B = −.098, SE = .077, p = .204, respectively). This was unsurprising given the overall high levels of reported mirth among respondents in all eight conditions (Figure 1). H2 proposed that experienced mirth and perceived effectiveness of the communicator are positively associated; the data and regression model bore out this expectation (B = .516, SE = .031, p < .001).
Predicting Perceived Effectiveness of the Communicator (N = 485).
We also questioned whether the effect of humor on participants’ perceptions of effectiveness would differ as a function of the speaker’s gender (RQ3a), race (RQ3b), and credentials (RQ3c). We found no evidence to support RQ3a or RQ3b, but there was a significant interaction between mirth and the experimental manipulation of the speaker’s credentials (B = .186, SE = .062, p = .003). Floodlight analysis revealed that the speaker’s credentials, whether they were labeled as a comedian or scientist, only mattered among respondents who reported levels of mirth lower than 5.06 on our 7-point scale (Figure 2). Among respondents who reported lower levels of mirth, a scientist, relative to a comedian, was perceived by participants to be a less effective communicator. Notably, 37.7% of our sample reported levels of mirth below this 5.06 threshold. Interestingly, the effect of a speaker’s credentials was not significant among participants who reported experiencing more humor.
Moderating and conditional effect of mirth on the relationship between the speaker credentials manipulation and perceived effectiveness of the speaker.
Discussion
Complex scientific topics can often be miscommunicated when experts exchange information with publics. Although generally perceived to be competent in their jobs, scientists are not the best communicators and score relatively poorly on perceptions of warmth and benevolence (Fiske & Dupree, 2014). There is some evidence that humor and laughter can increase the perceptions of communicator warmth without affecting perceived competence (Stewart et al., 2019) and generally have positive downstream effects on communicative outcomes (Cacciatore et al., 2020; Yeo, Anderson, et al., 2020; Yeo, Su, et al., 2020). Here, we conducted an experiment to examine how source characteristics, specifically race, gender, and credentials, affect perceived communicator effectiveness. Our study adds further empirical evidence of the positive effects of humor within the context of science comedy but is not without limitations.
Experiments allow us to isolate causal relationships but are limited by the artificiality of the experimental setting. In this case, the video that participants watched is not the same as attending a comedy show in the real world, especially prior to the start of the COVID-19 pandemic. That said, it does reproduce the format and settings to which comedians turned during the pandemic, that is, online shows. And while the delivery of the content by the individual comedians, who were compensated for their time and skill, are not completely identical, one of the authors wrote the entire script, and recruited and worked with the comedians on the delivery and appearance of the virtual stage to ensure that the conditions were as consistent as possible.
The second limitation is the nature of the sample used in this study. We obtained the sample from Amazon’s MTurk platform, which while not representative of the general adult population in the United States, is comparable to the representativeness offered by other online survey gathering platforms (Huff & Tingley, 2015; Levay et al., 2016). Specifically, this sample is likely more science-friendly than a general population sample and more comfortable with the online format for comedy (Hauser & Schwarz, 2016). Of course, this limitation does not mean that perceived differences in the humor production ability of people of different backgrounds do not exist in the larger context; we may simply have observed and measured an attenuated effect in this relatively more educated, and thus more science-inclined (National Science Board, 2020), sample.
We found that the race and gender of the source, relative to their credentials, were not significant factors for predicting the perceptions of communicator effectiveness (Table 1) among respondents in this study. This lack of significant differences based on race and gender of the communicators suggests that audiences are judging their experience with the stand-up routine on the basis of the jokes, rather than possible feelings of prejudice about who is delivering them (Becker, 2022). However, some viewer demographics, mirth, and the interaction between some source characteristics—specifically a speaker’s credentials and viewers’ experience of humor—did matter. Non-white respondents, those who identify as women, and those who are relatively more educated were more likely to perceive the speaker as effective at connecting with audiences around science. This is an area of study ripe for future research—understanding the dynamics and mechanisms through which audiences form opinions, perceptions, and attitudes at the intersection of the message, audience, and source characteristics is integral to designing effective communications that engage diverse publics to achieve specific goals and objectives.
In line with other work (Cacciatore et al., 2020; Yeo, Anderson, et al., 2020; Yeo, Su, et al., 2020), respondents who reported experiencing greater mirth as a result of watching the video were also more likely to perceive positive downstream effects. This effect was moderated by the speaker’s credentials (Figure 2). At lower levels of reported mirth, respondents perceived the speaker labeled as a comedian to be more effective than the scientist. However, among viewers who reported higher levels of mirth, perceived effectiveness of the scientist and comedian at connecting with audiences and communicating are at comparable levels. Here, humor appears to serve as an equalizer, and we posit that the mechanism through which this occurs may be through increasing perceptions of warmth, a concept on which we elaborate in the following paragraph.
Scientists are rated highly in terms of competence, a dimension of trust in communication sources (Fiske & Dupree, 2014). It is possible that humor improves how scientists are perceived on a dimension on which they score poorly relative to other professions, warmth (Fiske & Dupree, 2014). Comedians are likely to be perceived as having greater warmth relative to scientists, especially in the context of connecting with audiences. Indeed, a successful comedy act requires that a comedian is adept at connecting and reading their audience. Yet, if a scientist can make their audience laugh, this may increase audiences’ perceptions of warmth, which can, in turn, increase perceived communication effectiveness. As a result, there is a greater likelihood that their message will be more closely heeded by a public audience.
The mechanism of warmth may even be rooted in the relationship between psychological power and persuasion. Although power can be a difficult concept to define, most scholars have conceptualized it as some degree of asymmetrical control over resources in a social relationship (Dubois et al., 2016; Keltner et al., 2003). In general, high-power communicators tend to be more persuasive relative to low-power counterparts (Lammers et al., 2013). Yet, recent work has found that low-power states put more emphasis on information that conveys warmth relative to competence and that low-power communicators have been found to be more effective at persuading low-power audiences (Dubois et al., 2016).
Our audience, that is, our sample is likely to be one in a state of relatively low power. Indeed, there was little information about what the participants could expect from the video based on the instructions in the survey: “Please take a few moments to watch the following video. There is sound that accompanies the video clip, so please make sure your volume is turned up and/or use headphones if necessary.” The speaker labeled as a comedian could be considered to also be in a relatively low power state compared with the actor labeled as a scientist. This relative difference in perceived power likely arises from scientists being perceived as higher in competence and more knowledgeable (about science) than comedians.
Viewed through the lens of power, it is unsurprising that the comedian fared better with the audience relative to the scientist, that is, low-power communicator and audience. Nevertheless, among those who experienced greater mirth, the perceived effectiveness of the scientist (high-power communicator) matched that of the comedian, thus equalizing the power differential and lending weight to the possibility of humor having the potential to influence communicator effectiveness through warmth. Of course, this proposed pathway of influence is an untested hypothesis that warrants examination in future work.
Regardless of the pathway through which the effect of mirth influences communicator effectiveness, our findings can inform future studies and the practice of science communication. With regard to practice, our findings along with others (e.g., Cacciatore et al., 2020) highlight the utility of humor when it comes to communicating science. Whether it increases the perceptions of warmth remains an open question that, if answered, could greatly improve scientists’ communication efforts. Of course, humor is one of many tactics through which we can achieve (science) communication objectives and may not always be the best one in certain situations. Moreover, while our study offers some insight related to the using humor, we do not claim to have empirical evidence on how a communicator can be funny or make people laugh. A related consideration is our measure of mirth, which asked respondents to rate the extent to which they thought the video was “not funny-funny,” “not humorous-humorous,” and “not amusing-amusing.” These measures are similar to those used in other work (e.g., Nabi et al., 2007; Yeo, Su, et al., 2020) but are only one way of measuring experiences of humor elicited by a stimulus. Other means of measuring this affective experience (e.g., Duncan & Nelson, 1985) may yield insight into any nuances in mirth.
The findings of this study should be replicated with a sample that is more closely representative of the adult population in the United States. Moreover, these findings warrant replication in other cultural contexts as humor is a phenomenon that may be unique in various languages and cultures. Although our sample size was too small to address the interactions between the source characteristics, mirth, and audience characteristics, it would be informative for the practice of science communication to do so. While this project considered warmth as an explanation, future research could explore other psychological responses to funny media content, such as identification or transportation that focus more on engagement with the narrative and affective responses (humorous or otherwise) to characters (Kaplan & Dahlstrom, 2017). Future work should also consider how perceptions of others are subjective and can differ among individuals. For example, perceptions of a speaker’s gender can vary based on the degree of perceived femininity or masculinity. There are also expectations of what male or female comedians should look like that may play a key role in audience perceptions of communicator effectiveness. In addition, scientists may marshal humor as a tactic in a variety of settings, including in audio and video interviews with media or during public talks and engagement or outreach events (see, for example, Boykoff & Osnes, 2019). It may be that one type of event lends better to various types of humor as an effective tactic of science communication than others. These, too, are empirical questions that additional scholarship should investigate.
Inaccurate and misleading information remains a challenge in today’s media environment. When society is faced with challenges that are scientific in nature, such as the ongoing coronavirus pandemic or climate change, it is imperative that experts communicate pertinent information in a clear manner. While no single communication tactic and strategy can ensure that message receivers do not misunderstand complex information, empirical studies such as this one can help scientists and science communication professionals make informed choices about how to maximize communication effectiveness in a time of high uncertainty. Additional research in this area and, importantly, translation of such scholarship for practical use by communities involved in science communication are critical for ensuring that scientific research reaches beyond the ivory tower of academia and positively affects society.
Supplemental Material
sj-docx-1-scx-10.1177_10755470221132278 – Supplemental material for Humor Can Increase Perceived Communicator Effectiveness Regardless of Race, Gender, and Expertise—If You are Funny Enough
Supplemental material, sj-docx-1-scx-10.1177_10755470221132278 for Humor Can Increase Perceived Communicator Effectiveness Regardless of Race, Gender, and Expertise—If You are Funny Enough by Sara K. Yeo, Amy B. Becker, Michael A. Cacciatore, Ashley A. Anderson and Kasha Patel in Science Communication
Footnotes
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
Data Availability Statement
Supplemental Material
Supplemental material for this article is available online.
Notes
Author Biographies
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
