Sage Journals: Discover world-class research

Abstract

Emojis are paralinguistic elements that are used in computer-mediated communication to express emotions, convey tone, or simply depict objects or statements. This study examines how the (excessive) use of emojis affects knowledge, message credibility, and source trustworthiness. To this end, we conducted three online experiments using social media posts, in which we systematically varied the number of emojis and the source of the message. The results show that displaying many emojis decreases message credibility and source trustworthiness. As a mechanism explaining this effect, we also demonstrate that displaying many emojis is perceived as a persuasive attempt, triggering affective and cognitive reactance, and ultimately reducing message credibility. Moreover, the study shows that using (a large amount of) emojis in a message negatively affects factual knowledge.

Keywords

emojis message credibility source trustworthiness persuasive intent reactance media effects

The Internet and its subsequent technological advancements have provided individuals with numerous means of online communication, including instant messaging, social media, live streaming, video platforms, and discussion forums, all of which use computer-mediated communication (CMC; Cavalheiro et al., 2022). People use CMC for various reasons, including staying in touch with friends, carrying out work tasks, and interacting with companies or brands (Cavalheiro et al., 2022; Prada et al., 2022). CMC also allows users to communicate in multiple formats (e.g., text, images, animated GIFs, and video; Cavalheiro et al., 2022; Dunlap et al., 2016). Furthermore, CMC allows users to extend their textual communication with nonverbal elements through the use of emojis (Kralj Novak et al., 2015). Emojis are paralinguistic elements (Lee et al., 2021; Prada et al., 2022) that allow the sender to express feelings, moods, and emotions (Cavalheiro et al., 2022). Emojis are used in multiple modes of CMC. For example, they appear in interpersonal communication through email and instant messaging, as well as in public communication through social media and digital advertising (Jibril & Abdullah, 2013; Lee et al., 2021; Prada et al., 2022; Yakın & Eru, 2017).

Emojis represent a wide variety of concepts and ideas in a simplified form (Bai et al., 2019; Kralj Novak et al., 2015). These symbols are diverse, ranging from facial expressions, emotions, and feelings to abstract concepts, activities, and objects (Bai et al., 2019). Emojis evolved from emoticons [e.g., “;-)”], which are mostly shorthand for facial expressions. Emoticons were used before the widespread adoption of smartphones and social media, and thus represent a graphical advancement (Ljubešić & Fišer, 2016). However, subsequent technological advancements have led to emojis being used more frequently than emoticons and to wider range of symbols that go beyond facial expressions (Prada et al., 2018) and also represent values and norms (Robertson et al., 2018). There is reason to believe that, although “emoticons are not the same as emojis due to their more basic design, it can be extrapolated that the effects of emoticons could be extended and intensified through the use of emojis” (Willoughby & Liu, 2018, p. 77). This assumption is supported by the results of an online experiment by Ganster et al. (2012), which suggests that emojis and emoticons hardly differ in their effects.

Studies investigating the effects of emojis mostly emphasize the positive effects of using them, such as increased attention (Bai et al., 2019; Willoughby & Liu, 2018), support for reading comprehension (Lo, 2008; Riordan, 2017), or messages containing emojis being perceived as more informative, attractive, and innovative (Prada et al., 2022; Yakın & Eru, 2017). Emojis can also increase the likelihood of interaction (McShane et al., 2021) and have positive effects on purchase intention and product recommendation behavior (Distel et al., 2022; Indwar & Mishra, 2022; Prada et al., 2022). However, most studies have focused on comparing the effects of no emoji use to using only a few emojis. What happens when a substantial number of emojis are used? Willoughby and Liu (2018) show that using many emojis can have negative effects on information processing and credibility. Building upon their findings, our study seeks to explore the influence of emoji quantity on knowledge, message credibility, and source trustworthiness.

Functions of Emojis

The use of emojis in CMC contexts serves multiple functions, and we distinguish three primary purposes for their utilization. The first function refers to emojis’ ability to convey emotions (Danesi, 2022; Derks et al., 2008; Stark & Crawford, 2015). In light of this function, emojis can be described as “socio-emotional suppliers” (Jibril & Abdullah, 2013, p. 201) that set an emotional tone in messages (Distel et al., 2022; Kaye et al., 2016). For instance, examples such as “Please call me back ” or “Please call me back ” illustrate this aspect. In professional environments, emojis are used to strengthen emotional bonds with colleagues or clients (Stark & Crawford, 2015).

The second function pertains to the role of emojis in clarifying and understanding the context of a message (e.g., whether the message should be interpreted as ironic or sarcastic; Dresner & Herring, 2010). In addition to the verbal content, each utterance carries an intended action known as the illocutionary act. A single statement can have diverse meanings, such as a warning, a request, or a joke (Dresner & Herring, 2010). While in personal, face-to-face communication, these meanings are conveyed through voice and facial expressions, such nuances are challenging to express in CMC environments and text messaging. Here, emojis serve as valuable illocutionary elements that provide the necessary context for better understanding (Dresner & Herring, 2010). For instance, using the emoji “” can indicate that a statement is intended as a joke, thus reducing discourse ambiguity (Kaye et al., 2016).

The third function of emojis is linked to their capacity to enhance message comprehension (Lo, 2008; Riordan, 2017; Walther & D’Addario, 2001) and to accentuate or emphasize textual messages (Derks et al., 2008; Sampietro, 2019). In addition, emojis are also used as an interaction tool; for example, people use “” to communicate their agreement and to “enhance the attractiveness of the message to receivers” (Bai et al., 2019, p. 4). This example shows that emojis can also be used to illustrate a statement, to replace a word, or even an entire sentence. In advertising campaigns, incorporating emojis has been shown to enhance creativity and innovation ratings compared to campaigns that do not utilize emojis (Yakın & Eru, 2017).

In summary, the functions of emojis are to communicate emotions, convey the intensity of a message, and make statements clearer and easier to understand. Overall, emojis express the sender’s intentions and the emotions behind purely textual information (Bai et al., 2019). Emojis are used not only in private communication between users on social media but also in digital public communication and in digital advertising (Jibril & Abdullah, 2013; Lee et al., 2021; Yakın & Eru, 2017).

Effects of Emojis

Studies investigating the use and effects of emojis in CMC often focus on brand-consumer communication contexts and their effects on purchase intention or product recommendation (Distel et al., 2022; Indwar & Mishra, 2022; Prada et al., 2022). Furthermore, the majority of studies have concentrated on comparing the effects of completely abstaining from emoji use to utilizing only a few emojis. As a result, they have not explored the potential consequences of employing a significant quantity of emojis. While some studies have suggested that emojis enhance message comprehensibility compared to messages without emojis (Lo, 2008; Riordan, 2017; Walther & D’Addario, 2001), other research has indicated that an excessive use of emojis may have a negative effect on information processing (Willoughby & Liu, 2018). This could be attributed to the potential distraction caused by emojis, leading recipients to process the textual information less fluently. Processing fluency refers to how easily content is processed (Reber et al., 2004). Several factors can affect the processing fluency of a message. For instance, the use of color contrast or large font sizes can make a statement easier to read, thus enhancing processing fluency. Conversely, messages written in smaller font sizes or with low contrast can be more challenging to read, resulting in lower processing fluency and, consequently, reduced familiarity (Reber & Schwarz, 1999; Unkelbach, 2007).

Emojis may interfere with the reader’s ability to process the message fluently. When a message contains emojis, the reader must process both the textual content of the message and the visual cues provided by the emojis (Willoughby & Liu, 2018). This additional cognitive load may disrupt the fluency of the message and interfere with the reader’s ability to process the message efficiently. For example, studies show that reading fluency is related to reading comprehension (Klauda & Guthrie, 2008; Walczyk & Griffith-Ross, 2007) and that perceptual fluency is related to knowledge acquisition (Rau, 2018). Consequently, when emojis disrupt the fluency of a message, its content may not be as effectively retained in memory. In contrast, messages without emojis might be processed more smoothly since they lack additional visual elements that could distract the reader. Drawing from the previous findings of Willoughby and Liu (2018), we posit the following hypothesis:

H1: Using a greater number of emojis in a post leads to a decreased factual knowledge regarding the message content.

Apart from its potential impact on factual knowledge, the use of (an excessive amount of) emojis could also affect perceptions of credibility. Statements that are easier to process, for example, smoother and more effortless, are perceived as more familiar than statements that are more difficult to read (Parks & Toth, 2006). This feeling of familiarity is often associated with a sense of comfort, as we tend to prefer and trust things that we are familiar with. When we encounter a statement that feels familiar, it creates a sense of trust. This can lead us to perceive the statement as more credible. Thus, fluency has been shown to influence credibility judgments (Reber & Schwarz, 1999; Reber et al., 1998; Unkelbach, 2007). Messages that are easy to process are perceived as more credible (Koch & Zerback, 2013), while messages that are difficult to process are perceived as less credible (Unkelbach, 2007).

A few studies have examined the effects of emojis on message credibility and/or source trustworthiness (Beattie et al., 2020; Seghers et al., 2021). However, these studies did not consider the impact of varying numbers of emojis in their analyses. Willoughby and Liu (2018) showed that participants who viewed messages without emojis rated the messages as more credible than participants who viewed messages with emojis. Based on these findings and our theoretical assumptions about processing fluency, we hypothesize that:

H2: Messages containing a greater number of emojis are perceived as less credible compared to messages with fewer emojis.

H3: If messages contain a greater number of emojis, their source is perceived as less trustworthy.

Finally, we expect the postulated effects to differ depending on the type of source. Cavalheiro et al. (2022) show that acceptance of emoji use in brand communication was higher when the presence of emojis was perceived as more familiar or typical. We argue that a similar pattern might also apply to sources that incorporate emojis in their messages. It is plausible that people would expect less professional sources to use emojis more often than more professional sources. This mechanism can be explained by expectancy violations theory (Burgoon, 1993; Burgoon & Le Poire, 1993), which suggests that people form expectations based on the prior actions of others. This idea can also be applied to perceptions of organizations: Their communication is evaluated against the background of people’s expectations. When expectations and observed actions do not match, previously built expectations are violated. This can lead to a positive or negative reassessment of the source (Burgoon, 1993; Burgoon & Le Poire, 1993). For example, people may have different expectations of international organizations (e.g., UN, WHO)/research institutions and community groups/volunteer organizations as sources of information; consequently, people’s trust varies depending on the type of source. Thus, the effects on message credibility and source trustworthiness may be moderated by people’s expectations of the source. This idea leads to the final hypothesis:

H4: The effects postulated in H2 and H3 are moderated by the type of source.

We tested the four hypotheses in the first two studies.

Study 1a

Method

Sample

We recruited 324 participants (58% female; M_age = 48.67 years, SD_age = 15.50 years) via an online access panel for social science research (SoSci Panel; Leiner, 2016). To determine the sample size needed, an a priori power analysis using G*Power (Faul et al., 2007) was conducted for an analysis of variance (ANOVA; df = 2, 6 groups) based on the effect sizes from a previous study by Willoughby and Liu (2018). With a desired power of 85% and significance level (α) set at .05, the analysis indicated a required sample size of N = 210. Participants did not receive compensation for their voluntary participation.

Design and Procedure

We conducted a 3 × 2 between-subjects online experiment. As independent variables, we manipulated the number of emojis in a Facebook post (no emojis vs a low number of emojis vs a high number of emojis) and the source of the message (a more professional source [a health department] vs a less professional source [an NGO]). The participants were randomly assigned to one of the six experimental conditions and instructed to carefully read the social media post. We chose Facebook because it is one of the largest social media platforms overall and has a large presence of brands and organizations (Stelzner, 2023).

Stimulus

The post informed readers about chlamydia, a sexually transmitted infection, explaining what chlamydia is, its symptoms, and how to protect against it, using information based on the most recent scientific knowledge (Centers for Disease Control and Prevention [CDC], 2022a). The first factor was the number of emojis used in the post. Most experiments compared the use of no emojis with the use of a few emojis, typically ranging from three to eight emojis (Tang & Hew, 2019). However, to explore the effects of larger amounts of emojis, we chose seven emojis as our “medium” category, which is consistent with the range commonly studied. In addition, we included a “many emoji” category with 20 emojis to explore the effects of excessive emoji use. Thus, we created three different versions: no emojis, 7 emojis, and 20 emojis. We used different types of emojis, including facial emojis, symbols, and objects (all stimuli can be found on OSF: https://osf.io/7bvzc/?view_only=8bc226c2b25e4fddafbd372d69a2d635). The emojis did not give additional meaning to the sentences, but rather showed redundant information (illustrative function). Thus, we ensured that the meaning of all the sentences remained the same with or without the emojis. As the second factor, we also manipulated the source of the message. The Facebook profile was either a fictitious health department or a fictitious NGO. For the government organization, we chose the name “State Health Department,” which conveys authority and professionalism. In addition, we used the official German “Bundesadler” (federal eagle) as the profile picture to further enhance the perception of professionalism. For the NGO, on the other hand, we presented it as “Sex Education,” a name that conveys a more casual or informal impression. To reinforce this perception, the profile picture depicted a finger on an orange, symbolizing sexual stimulation, chosen to convey a less formal or professional image. We used fictitious profiles to control for the influence of pre-existing attitudes. We chose the topic and the wording of the text to appear authentic for both sources—the health department and the NGO—and to make sure that the use of emojis was not inappropriate or too obvious. The posts differed only in terms of the profile name and profile picture and the number of emojis. All other characteristics, including text, images, and the numbers of likes and comments, were identical.

Measures

We measured factual knowledge by presenting the participants with seven statements about the Facebook post they had seen. The statements were as follows: The article. . . “. . .was about a rather rare sexually transmitted infection,” “. . .pointed out that chlamydia can lead to infertility and abdominal inflammation,” “. . .informed the reader that the infection is usually accompanied by severe symptoms,” “. . .mentioned antibiotics as a suitable medication to fight the infection,” “. . .also provided information about another sexually transmitted infection,” “. . .clarified that condoms can significantly reduce the risk of infection,” and “. . .explained that it is important to be tested regularly for infection with chlamydia.” We asked the participants to indicate whether each statement was true or false. For each correct answer, participants received one point. We added the scores on individual items to form a single index ranging from 0 to 7 (M = 5.56, SD = 1.42).

We assessed message credibility using the three adjective pairs developed and validated by Appelman and Sundar (2016), assessed using five-point scales. The adjective pairs were “accurate–not accurate,” “authentic–not authentic,” and “believable–not believable.” These three variables were combined into an index (M = 3.77, SD = 1.00, α = .84).

We measured source trustworthiness following the approach of Ohanian (1990), using a five-point semantic differential scale of adjective pairs. We presented participants with three adjective pairs (“trustworthy–not trustworthy,” “authentic–not authentic,” and “credible–not credible”) in response to the item “I believe that the health department/the Sex Education NGO is. . .” The three variables were combined into an index (M = 3.75, SD = 1.03, α = .89).

Treatment Check

To check whether the amount of emojis was perceived as wanted, we asked the participants how many emojis they saw in the post. A univariate ANOVA showed that most participants who were exposed to the post with no emojis reported seeing no emojis (M = 0.83, SD = 3.01), whereas the participants who were exposed to the version with seven emojis reported seeing 8.28 emojis on average (SD = 5.12), and participants in the experimental group with 20 emojis reported seeing 20.77 emojis on average (SD = 11.90), F(2, 267) = 411.27, p < .001, η² = .61. Post hoc testing indicated that all three group mean values differed significantly from each other (p < .001). Furthermore, we asked the participants about the perceived professionality of the organization. An independent samples t-test showed that participants perceived the health department (M = 3.62, SD = 1.31) as more professional than the Sex Education NGO (M = 3.12, SD = 1.13), t(321) = 3.628, p < .001, on a five-point scale. Hence, both manipulations were successful.

Results

To test H1 (effects of emojis on factual knowledge), we performed a two-factor ANOVA using knowledge as the dependent variable. There was no significant main effect of the number of emojis on knowledge, F(2, 318) = 1.31, p = .271, η² = .01. Therefore, H1 was rejected. Moreover, there was no significant main effect of source, F(1, 318) = 0.01, p = .94, η² = .00, and no significant interaction effect, F(2, 318) = 0.57, p = .57, η² = .00.

To test H2 (effect of emojis on message credibility) and H4 (source as a moderator), we also calculated a two-factor ANOVA with message credibility as the dependent variable. The analysis showed a significant main effect of the number of emojis, F(2, 318) = 10.51, p < .001, η² = . 06. A post hoc test revealed that participants perceived a message as more credible when no emojis were used (M = 4.12, SD = 0.96) than when few (M = 3.68, SD = 0.94) or many (M = 3.53, SD = 1.00) emojis were used, supporting H2. The groups viewing 7 or 20 emojis, however, did not differ significantly from each other. There was no significant interaction effect, F(2, 318) = 0.49, p = .61, η² = .00. Moreover, there was also no significant main effect of source, F(1, 318) = 0.05, p = .83, η² = .00.

Finally, to test H3 (effect of emojis on source trustworthiness) and H4 (source as a moderator), a two-factor ANOVA with source trustworthiness as the dependent variable was performed. The ANOVA revealed a significant main effect of the number of emojis, F(2, 317) = 5.51, p = .004, η² = .03; a post hoc test showed that the post without emojis led to a more trustworthy perception of the source (M = 4.02, SD = 1.00) than the posts with few (M = 3.62, SD = 1.03) or many (M = 3.63, SD = 1.02) emojis, supporting H3. However, there was no significant main effect of source, F(1, 317) = 2.42, p = .12, η² = .01, and no significant interaction effect, F(2, 317) = 0.75, p = .47, η² = .01. Therefore, H4 cannot be supported for either source trustworthiness or message credibility.

Discussion

Study 1a shows that presenting seven or more emojis in a social media post can reduce message credibility and source trustworthiness. However, it makes no difference whether 7 or 20 emojis are used: In both cases, the message is judged to be less credible. We find the same results for source trustworthiness: When emojis are used in the message, regardless of whether there are a few or many emojis—the source is judged to be less trustworthy. Contrary to our expectations, these effects are not moderated by the type of source of the post (a health department or an NGO). In addition, we find no significant main effect of the number of emojis on participants’ factual knowledge. The topic of chlamydia may have influenced the results regarding participants’ factual knowledge. Because chlamydia is transmitted through sexual intercourse, the topic may have received a lot of attention. We decided to validate our findings by conducting a second experiment with a different scenario and topic.

Study 1b

Method

Sample

We recruited 216 participants via social media; after filtering out eight individuals who completed the questionnaire in less than 4 min and excluding six participants who admitted to not reading the stimulus material, a final sample of 204 individuals remained (66% female, M_age = 32.58, SD_age = 11.93). They did not receive compensation for their voluntary participation.

Design and Procedure

As in Study 1a, we conducted an online experiment with a 3 × 2 factorial between-subjects design, varying the number of emojis (no emojis vs a low number of emojis vs a high number of emojis) and the source of the message (an aviation authority vs a travel company). Again, the participants were randomly assigned to the six conditions and instructed to carefully read the social media post.

Stimulus

We again used a Facebook post as our stimulus (all stimuli can be found on OSF: https://osf.io/7bvzc/?view_only=8bc226c2b25e4fddafbd372d69a2d635). For the number of emojis included in the text, there were either many (26), few (6), or no emojis at all (Factor 1). The posts were posted by the fictitious “Federal Office of Aviation and Air Traffic Control” (more professional) or by a fictitious travel company, “Travelista” (less professional; Factor 2). As in Study 1a, the posts differed only in terms of the profile name and the number of emojis. All other characteristics, including text, image, and number of likes and comments, were identical.

Measures

To measure the participants’ factual knowledge, we asked them to indicate whether seven statements about the Facebook post they had seen were true or false. The statements were as follows: “The article was about the start of the holidays in Bavaria, Hesse, and Rhineland-Palatinate,” “The article pointed out that you can check in online 24 hours before departure,” “The article reported about 15 million–20 million travelers,” “The article mentioned the airports in Frankfurt and Munich,” “The article advised arriving at the airport four hours before departure,” “The article was about the beginning of winter vacations,” and “The article mentioned Mallorca and Greece as the top travel destinations for this year.” For each correct answer, participants received one point. We added the scores on individual items to form a single index ranging from 0 to 7 (M = 3.59, SD = 2.29).

We again assessed message credibility using a five-point scale, with the three adjective pairs of Appelman and Sundar (2016): “accurate–not accurate,” “authentic–not authentic,” and “believable–not believable.” The three variables were combined into an index (M = 3.21, SD = 0.97, α = .85).

The trustworthiness of the source was again measured following Ohanian’s (1990) approach of using a five-point semantic differential scale for adjective pairs (“trustworthy–not trustworthy,” “authentic–not authentic,” and “credible–not credible”). The three variables were also combined into an index (M = 3.30, SD = 0.91, α = .89).

Treatment Check

Again, to check whether the amount of emojis was perceived as wanted, we asked the participants how many emojis they saw in the post. We calculated a univariate ANOVA. The results showed that the participants who were exposed to the post displaying no emojis reported seeing 2.00 emojis on average (SD = 3.07), whereas the participants who were exposed to the version with six emojis reported seeing 6.31 emojis on average (SD = 3.93), and those exposed to the version with 26 emojis reported seeing 26.04 emojis on average (SD = 14.30), F(2, 137) = 89.13, p < .001, η² = .57. A post hoc test indicated that all three group mean values differed significantly from each other (p < .001). In addition, we asked the participants about the perceived professionality of the organization. An independent samples t-test indicated that the participants perceived the Federal Office of Aviation and Air Traffic Control (M = 3.36, SD = 1.23) as more professional than the travel company Travelista (M = 2.96, SD = 0.93), t(202) = 2.58, p < .01, on a five-point scale. Hence, both manipulations were successful.

Results

To test H1 (effect of number of emojis on factual knowledge), we conducted a two-factor ANOVA using knowledge as the dependent variable. The ANOVA showed a significant main effect of the number of emojis, F(2, 198) = 3.93, p = .02, η² = .04. A post hoc test showed that only the control group (M = 4.08, SD = 2.18) differed significantly from the group where 26 emojis were displayed (M = 3.00, SD = 2.35), and there was no difference from the group that saw 6 emojis (M = 3.73, SD = 2.25). Therefore, H1 can be partially supported. However, there was no main effect of source, F(1, 198) = 0.23, p = .63, η² = .00, and no significant interaction effect, F(2, 198) = 0.71, p = .49, η² = .01. Hence, using many emojis in a post seems to decrease recipients’ knowledge of the information presented.

To test H2 (effect of emojis on message credibility) and H4 (source as a moderator), we conducted a two-factor ANOVA with message credibility as the dependent variable. The analysis revealed a significant main effect of the number of emojis, F(2, 198) = 17.52, p < .001, η² = .15, showing that the post with many emojis (M = 2.69, SD = 0.96) was perceived as less credible than the post with few (M = 3.44, SD = 0.86) or no (M = 3.51, SD = 0.86) emojis. Consequently, H2 is supported. There was also a significant main effect of source, F(1, 198) = 5.87, p = .02, η² = .03, showing that the credibility of the message was higher when it was posted by the aviation authority (M = 3.36, SD = 0.96) compared with the travel company (M = 3.05, SD = 0.96). The interaction effect was not significant, F(2, 198) = 0.24, p = .78, η² = .00.

To test H3 (effect of emojis on source trustworthiness) and H4 (source as a moderator), we also performed a two-factor ANOVA with source trustworthiness as the dependent variable. We found a significant main effect of the number of emojis, F(2, 198) = 12.84, p < .001, η² = .12: Participants perceived the source as less trustworthy when many emojis were used (M = 2.89, SD = 0.93), compared with messages that included only a few (M = 3.49, SD = 0.86) or no (M = 3.55, SD = 0.80) emojis, partially supporting H3. The results also revealed a significant main effect of source, F(1, 198) = 18.18, p < .001, η² = .08: The aviation authority (M = 3.55, SD = 0.97) was perceived as more trustworthy than the travel company (M = 3.05, SD = 0.77). Again, the interaction effect was not significant, F(2, 198) = 0.40, p = .67, η² = .00. Therefore, H4 cannot be supported for either source trustworthiness or message credibility. The study thus suggests that a high number of emojis in a post decreases message credibility and source trustworthiness.

Discussion

Like Study 1a, Study 1b also shows that the presentation of emojis in a public social media post reduces message credibility and source trustworthiness. Participants judge a message with many emojis as less credible than a message with no or a few emojis. As in Study 1a, these effects are not moderated by the type of source of the message (the Federal Office of Aviation and Air Traffic Control or a travel company). In contrast to Study 1a, in Study 1b, we find a significant effect of the number of emojis on participant’s knowledge: Using many emojis in a post decreases the recipient’s knowledge of the information presented. Both experiments show that using many emojis in strategic communication decreases message credibility and source trustworthiness—but why does this effect occur? We decided to conduct a third experiment to investigate the mechanisms of these effects.

Study 2

Why does the use of many emojis reduce message credibility and trust in the source? We postulated that this effect is caused by a disrupted processing fluency. However, another possible explanation could be that recipients perceive the use of many emojis in strategic communication as a persuasive attempt. Recipients feel that emojis are being used strategically to persuade them. The Persuasion Knowledge Model proposed by Friestad and Wright (1994) is a theoretical framework that explains how people cope with persuasive attempts. Throughout their lives, people acquire knowledge about persuasive tactics: They learn how to identify persuasive attempts and how to “skillfully cope with these” (Friestad & Wright, 1994, p. 1). The ability to identify persuasive attempts and the understanding of strategies and tactics are referred to as conceptual persuasion knowledge (Beckert & Koch, 2022; Beckert et al., 2021; Boerman et al., 2018; Campbell & Kirmani, 2008). Although the Persuasion Knowledge Model originally referred to advertising and face-to-face conversations, it has been widely applied to more contemporary (and covert) forms of persuasive attempts in various online contexts (Beckert & Koch, 2022). Persuasion knowledge determines how recipients perceive strategic communication in social media posts (Beckert et al., 2021; De Veirman & Hudders, 2020; Krouwer et al., 2017).

When recipients feel that someone is trying to persuade them, they respond to this persuasion attempt with certain coping behaviors (Ham et al., 2015). Because they perceive persuasion attempts as a restriction on their personal freedom of choice, they are motivated to restore this freedom; this motivation is referred to as reactance (Brehm & Brehm, 1981). Reactance can be thought of as a combination of anger and negative cognitions (Dillard & Shen, 2005), which can negatively affect the perception of the communicated content (Meirick & Nisbett, 2011). Koch and Zerback (2013) show that the activation of reactance leads to a decrease in message credibility. Based on this finding, we hypothesize that two mechanisms may occur when considering the impact of emojis: the presence of emojis disrupts reading fluency and excessive use of emojis may create perceptions of persuasive intent, leading to reactance and subsequently decreasing credibility and trustworthiness. Therefore, we argue that it is plausible to assume that these two mechanisms, interrupted fluency and persuasion knowledge/reactance, operate in parallel to influence the effects of emojis on message credibility and source trustworthiness. This leads to the following two hypotheses (see Figure 1):

H5a: Messages displaying many emojis are perceived as persuasive attempts, which triggers cognitive reactance and thus decreases message credibility (indirect effect I).

H5b: Messages displaying many emojis are perceived as persuasive attempts, which triggers affective reactance and thus decreases message credibility (indirect effect II).

H6: Messages displaying many emojis directly decrease message credibility.

Figure 1.

Conceptual model.

Method

Sample

We recruited 338 participants (60% female, M_age = 49.46 years, SD_age = 15.28) via an online access panel (SoSci Panel; Leiner, 2016). They did not receive compensation for their voluntary participation.

Design and Procedure

As in Studies 1a and 1b, we conducted an online experiment with a 3 × 2 factorial between-subjects design. The number of emojis (no emojis vs a low number of emojis vs a high number of emojis) and the source of the message (an association of dermatologists vs a medical student union) were varied. The participants were randomly assigned to the conditions and instructed to carefully read the social media post.

Stimulus

We used a Facebook post that informed participants about skin cancer and how to prevent it as the stimulus (all stimuli available on OSF: https://osf.io/7bvzc/?view_only=8bc226c2b25e4fddafbd372d69a2d635). The number of emojis included in the text was 0, 9, or 22 (Factor 1). The post was posted either by a fictitious association of dermatologists of skin cancer (more professional) or by a fictitious medical student union (less professional; Factor 2). The posts differed only in terms of the profile name and the number of emojis; all other characteristics were identical. Again, the information presented was based on recent scientific publications (CDC, 2022b).

Measures

As in Studies 1a and 1b, we measured factual knowledge by asking the participants to indicate whether seven statements about the Facebook post they saw were true or false. The statements were as follows: “The article also reported about other skin diseases besides skin cancer,” “The article mentioned melanoma as the fifth most common cancer,” “The article explained that children are at risk for skin cancer,” “The article stated that the risk of skin cancer has been increasing for decades,” “The article named UV radiation as the main risk factor for skin cancer,” “The article recommended regular skin cancer screenings from the age of 20,” and “The article reported that people are entitled to a screening examination every year from the age of 35.” For each correct answer, the participants received one point. We added the scores on individual items to form a single index ranging from 0 to 7 (M = 4.76, SD = 1.45).

As in Studies 1a and 1b, we measured message credibility using a five-point scale with the following three adjective pairs (Appelman & Sundar, 2016): “accurate–not accurate,” “authentic–not authentic,” and “believable–not believable.” The three variables were combined into an index (M = 3.21, SD = 0.97, α = .85).

The trustworthiness of the source was again assessed using a five-point semantic differential scale (Ohanian, 1990) of adjective pairs (“trustworthy–not trustworthy,” “authentic–not authentic,” and “credible–not credible”). These three items were combined into an index (M = 3.57, SD = 0.98, α = .91).

In addition, we measured the participants’ perceived persuasive intent using a five-point Likert-type scale (strongly disagree–strongly agree), similar to the scales used by Tutaj and van Reijmersdal (2012) and Beckert et al. (2021). The three statements used in this measurement were as follows: “I had the feeling that the post was trying to manipulate me,” “It tried to steer my views on the subject in a certain direction,” and “The post was trying to convince me.” An index of the three items was created (M = 2.87, SD = 0.98, α = .64).

Here, we also assessed reactance as a two-dimensional construct comprising anger and negative cognitions (Brehm & Brehm, 1981; Dillard & Shen, 2005). We measured both components using a scale similar to the ones used by van Reijmersdal et al. (2016) and Beckert et al. (2021). Both constructs were assessed with three items using five-point Likert-type scales (strongly disagree–strongly agree). To assess anger, the participants were asked to rate how much they agreed or disagreed with the following statements: “The presentation of the post made me angry,” “I was annoyed by the way the post was presented,” and “I was upset by the way the post was written.” The items were combined into an index (M = 2.70, SD = 1.49, α = .91). Negative cognitions were measured using the following three items: “I immediately wanted to contradict the core statements of the post,” “I questioned the statements of the post after reading it,” and “I had doubts about the message of the post”; again, an index was created (M = 1.89, SD = 1.00, α = .84).

Treatment Check

Again, we asked the participants how many emojis they saw in the post. A univariate ANOVA showed that most subjects exposed to the post displaying no emojis indeed reported seeing no emojis (M = 0.21, SD = 0.58), whereas those who were exposed to the version with 9 emojis reported seeing 9.20 emojis on average (SD = 5.42) and those exposed to the version displaying 22 emojis reported seeing 21.66 emojis on average (SD = 9.35), F(2, 293) = 288.60, p < .001, η² = .66. A post hoc test indicated that all three group mean values differed significantly from each other (p < .001). Furthermore, we asked about the perceived professionality of the organization. An independent samples t-test revealed that the participants perceived the association of dermatologists (M = 3.47, SD = 1.23) as more professional than the medical student union (M = 2.94, SD = 1.28), t(334) = 3.86, p < .001, on a five-point scale. Hence, both manipulations were effective.

Results

We calculated a two-factor ANOVA using factual knowledge as the dependent variable to test H1 (effect of number of emojis on knowledge). The analysis revealed a significant main effect of the number of emojis, F(2, 332) = 6.96, p = .001, η² = .04. A post hoc test showed that the mean value of the control group (M = 5.13, SD = 1.17) differed significantly from that of the experimental groups where 9 (M = 4.69, SD = 1.45) and 22 (M = 4.45, SD = 1.62) emojis were displayed, partially supporting H1. There was no main effect of source, F(1, 332) = 1.88, p = .71, η² = .01, and no significant interaction effect, F(2, 332) = 1.37, p = .25, η² = .01. As in Study 1b, displaying (many) emojis in a post seemed to decrease the user’s knowledge.

To test whether the number of emojis and different sources affect message credibility (H2), we calculated a two-factor ANOVA. The analysis showed no significant main effect of source, F(1, 332) = 2.65, p = .10, η² = .01, and no significant interaction effect, F(2, 332) = 1.07, p = .34, η² = .01. However, there was a significant main effect of the number of emojis, F(2, 332) = 41.29, p < .001, η² = .20. Post hoc tests showed that the participants perceived a message as less credible when many emojis were used (M = 3.10, SD = 1.05), compared to a message with few emojis (M = 3.40, SD = 1.04); messages with a few emojis, in turn, were less credible than those with no emojis (M = 4.20, SD = 0.76). This supports H2.

Moreover, we performed a two-factor ANOVA to test whether emojis affect source trustworthiness (H3). There was a significant main effect of the number of emojis, F(2, 330) = 14.89, p < .001, η² = .08, showing that a post with many emojis led to the source being perceived as less trustworthy (M = 3.26, SD = 0.95) than a post with few (M = 3.51, SD = 1.04) or no (M = 3.94, SD = 0.85) emojis, partially supporting H3. The results also showed a significant main effect of source, F(1, 330) = 8.61, p = .004, η² = .03, indicating that the association of dermatologists was perceived as more trustworthy (M = 3.72, SD = 0.95) than the medical student union (M = 3.42, SD = 1.00). There was no significant interaction effect, F(2, 330) = 0.34, p = .72, η² = .00. Therefore, H4 cannot be supported for source trustworthiness or message credibility. As in Studies 1a and 1b, the results revealed that presenting many emojis in a post can reduce message credibility as well as source trustworthiness.

We conducted a mediation analysis with 5,000 bootstrap samples to test H5a (mediation via persuasive intent and anger) and H5b (mediation via persuasive intent and negative cognition). The number of emojis (0 = 0 emojis, 1 = 9 emojis, 2 = 22 emojis) was included as the independent variable, message credibility was the dependent variable, and perceived persuasive intent and affective and cognitive reactance were entered as mediating variables. The results are displayed in Figure 2.

Figure 2.

Results of the mediation analysis.

The analysis showed that a large number of emojis significantly increased perceived persuasive intent (b = 0.32, p < .001), as well as affective reactance/anger (b = 1.06, p < .001) and cognitive reactance/negative cognitions (b = 0.16, p < .01). Perceived persuasive intent, in turn, increased affective reactance (b = 0.25, p < .001), as well as cognitive reactance (b = 0.44, p < .001) and decreased message credibility (b = −0.17, p < .001). Affective reactance (b = −0.18, p < .001), as well as cognitive reactance (b = −0.47, p < .001), also decreased message credibility.

Taken together, the mediation analysis showed five significant indirect effects of the number of emojis on message credibility. The first indirect effect was mediated by perceived persuasive intent (b = −0.05, 95% CI = [−0.10, −0.02]). The second indirect effect was mediated by affective reactance (b = −0.19, 95% CI = [−0.28, −0.10]). The third indirect effect was mediated by cognitive reactance (b = −0.08, 95% CI = [−0.14, −0.02]). The fourth indirect effect was mediated by perceived persuasive intent and affective reactance (b = −0.01, 95% CI = [−0.03, −0.01]). The fifth indirect effect was mediated by perceived persuasive intent and cognitive reactance (b = −0.07, 95% CI = [−0.10, −0.04]). These finding support H5a and H5b: A large number of emojis is perceived as a persuasive attempt, which triggers cognitive as well as affective reactance and thus decreases message credibility. Moreover, the study showed that a large number of emojis significantly decreased message credibility directly (b = –0.15, p < .05), supporting H6.

Discussion

The third experiment (Study 2) also shows that displaying emojis in social media posts decreases message credibility and source trustworthiness. As in the previous studies, these effects are not moderated by source type. Thus, all three experiments show that the use of emojis in strategic social media communication has negative effects on message credibility and source trustworthiness. To explain why this occurs, we also measured whether people recognize many emojis, in particular, as a persuasive attempt. As hypothesized, the experiment shows that emojis are perceived as a persuasive attempt, which triggers both cognitive and affective reactance and thus decreases message credibility. In addition, consistent with Studies 1a and 1b, we found that a large number of emojis negatively affected participants’ knowledge.

General Discussion

Research on the use of emojis beyond their effects on purchase intention and product recommendation is scarce. Furthermore, such research often focuses on the positive effects of using emojis compared to not using them at all. Therefore, in three experimental studies, we focused on the effects of the number of emojis on knowledge, message credibility, and source trustworthiness in CMC. All three experiments consistently showed that displaying many emojis decreased message credibility and source trustworthiness. In the experiments, six, seven, and nine emojis were sufficient for participants to rate a post as less credible and the source of the post as less trustworthy compared to a post with no emojis. Displaying a greater number of emojis had the same effect. This finding suggests that strategic communicators need to be thoughtful about their use of nonverbal cues in CMC. They need to consider the potential negative impact that using too many emojis can have on the credibility of their message, as well as their own trustworthiness, and adjust their communication strategies accordingly. Emojis can add richness and nuance to messages, but their overuse or inappropriate use can have unintended consequences.

Moreover, we assumed that the type of the source would moderate these effects and that a message from a more professional source may require a different approach to the use of emojis than a message from a less professional source. In the former case, the use of too many emojis might be perceived as unprofessional and reduce the credibility of the message, while in the latter case, the use of emojis might be more acceptable and even expected. However, all three experiments did not show these moderating effects.

It is important to note that the emojis used did not add any additional meaning to the sentences: We made sure that the meaning of all sentences remained the same with or without emojis. In addition, we wanted to answer the question of why the observed effect occurs. Study 2 analyzed a possible mechanism of this effect and showed that the use of emojis is perceived as a persuasive attempt. Recipients feel that the source is trying to persuade them by using emojis. This, in turn, triggers cognitive and affective reactance. Moreover, the data also show that the use of emojis directly triggers both components of reactance and that the affective component (i.e., anger) in particular increases as a result. This activation of reactance then leads to a decrease in message credibility. In addition to the mediating effect of perceived persuasiveness, we also found a direct effect of the number of emojis on message credibility. This direct effect may be due to our assumption that the disruption of fluency caused by the presence of numerous emojis reduces the overall credibility of the message. However, further research is warranted to empirically test the validity of this explanation in future studies.

Two of the studies suggest that the use of emojis may have unintended consequences when it comes to conveying factual information (the first experiment did not find a significant effect, but provided suggestive evidence in the same direction): Using a large number of emojis may negatively affect participants’ factual knowledge. One possible explanation for this finding is that the use of many emojis may reduce processing fluency, leading to difficulties in reading and understanding a text. The use of many emojis may draw participants’ attention away from the actual information being conveyed and make it more difficult for them to focus on the textual content. In addition, the use of emojis may encourage participants to rely on visual cues rather than reading and comprehending the text, resulting in a reduced understanding of the content.

Our experimental studies are limited in several ways. First, because we used fictitious stimuli with fictitious sources and fictitious messages in all three studies, we cannot determine how pre-existing attitudes might affect the processing of emojis. Pre-existing attitudes toward a source are likely to influence how the use of emojis and the number of emojis are perceived. Accordingly, future studies should consider how pre-existing attitudes and perceptions about real sources influence the effects of emojis. Second, a variety of different sources use emojis on social media, and these can be applied to various topics. The present study investigated three specific topics and six specific sources. Therefore, no conclusions can be drawn for other topics or sources. Future studies could focus on other topics and sources to broaden the scope. Third, we conducted all three experiments in Germany; therefore, our sample is geographically limited. Follow-up studies should be designed to compare different countries.

Fourth, it is important to acknowledge that emojis serve multiple functions, and their perception may vary depending on the specific function they serve and the type of emojis used. Emojis can convey emotions, help clarify and understand the context of the message, and make messages more understandable (by visually illustrating a point, replacing a word, or even conveying an entire sentence with a single symbol). The perception of emojis as persuasive or simply as annoying, funny, or sad depends on the specific functions they perform and how they are used within the message. In our study, we chose emojis that did not add additional meaning to the messages to increase the internal validity of the experiments. These emojis served primarily to illustrate the content without changing or enhancing the information conveyed. Therefore, future studies should further investigate the effects of different types of emojis.

Fifth, it is important to note that the generalizability of these findings to different stimuli and contexts may be limited. Especially the choice of Facebook as the social media platform limits the generalizability of the findings, as social media platforms serve different purposes, have different user demographics, and may elicit different user behaviors (Pelletier et al., 2020; Waterloo et al., 2018). Furthermore, functions of emojis may also depend on the context of use, for example, the same emojis may serve different functions on different platforms, in different communication constellations (e.g., face-to-face communication, conversations in larger groups, public vs private communication), and the source of a message (friend, stranger, organization, etc.). Thus, not only our choice of platform limits the generalizability of our study but also our choice of a public communication context with an organization as the source. However, the design ensured robust internal validity by isolating the specific effects of emoji quantity on message perception, but this emphasis on internal validity may come at the expense of external validity. Future studies should aim to replicate and extend our findings by examining emoji use in different contexts, including other social media platforms, different topics, public vs private contexts, and a wider range of sources. Our experiments are also limited regarding the rather high number of emojis in the experimental conditions (we used at least six emojis). Future research could replicate our studies using fewer emojis (e.g., just one, two, or three). Moreover, factors such as cultural differences, age groups, and individual preferences play a role in how emojis are perceived and interpreted; therefore, it is essential that future research includes participants from different cultural backgrounds, age groups, and demographic profiles. Finally, we placed the emojis not only at the beginning or at the end of a sentence, but also in the middle. This may have influenced the perceived artificiality of our stimuli and may also have affected the credibility of the message as well. Therefore, future studies should also consider the positioning of emojis in a text. Furthermore, we did not control for the perceived valence of the emojis. Especially when using face emojis, the meaning may be perceived differently. Future studies should, therefore, take the valence of the emojis into account.

The primary focus of this study was not to examine persuasion effects directly, as attitudes were not measured or manipulated. Instead, we were primarily interested in investigating the effects of emojis on credibility and trustworthiness. It is worth noting, however, that our findings indirectly touch on persuasion effects. Previous research has consistently shown that trustworthy sources are more persuasive than non-trustworthy sources and vice versa (Hovland et al., 1953). Consistent with this finding, our study suggests that excessive use of emojis could potentially hinder persuasion efforts, as both the source and the message may be perceived as less credible and trustworthy. Future studies should consider extending these findings by designing scenarios that explicitly explore persuasive effects and measuring the influence of emoji use on participants’ attitudes. Overall, this study supports the idea that the use of emojis can have negative effects on perceptions of both the source and the message itself. Therefore, organizations should think carefully about whether and how to use emojis.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Thomas Koch

Author Biographies

Thomas Koch (Ph.D., University of Erlangen-Nuremberg, Germany) is Professor of Strategic Communication at the Johannes Gutenberg-University of Mainz, Germany. His research interests include persuasive communication, strategic communication, the blurring lines between journalism and public relations, CSR, and media effects.

Nora Denner (M.A., Ludwig-Maximilians-University of Munich, Germany) is a researcher in the field of strategic communication at the Johannes Gutenberg-University of Mainz, Germany. Her research interests include personalization in corporate reporting, crisis communication, and media effects.

Melina Crispin (B.A., Johannes Gutenberg-University Mainz, Germany) is an M.A. student of gender, intersectionality and politics at the Free University of Berlin, Germany.

Trees Hohagen (M.A., Johannes Gutenberg-University Mainz, Germany) is a consultant for corporate communications.

References

Appelman

Sundar

S. S

. (2016). Measuring message credibility: Construction and validation of an exclusive scale. Journalism & Mass Communication Quarterly, 93(1), 59–79. https://doi.org/10.1177/1077699015606057

Bai

Dan

Yang

. (2019). A systematic review of emoji: Current research and future perspectives. Frontiers in Psychology, 10, 2221. https://doi.org/10.3389/fpsyg.2019.02221

Beattie

Edwards

A. P.

Edwards

. (2020). A bot and a smile: Interpersonal impressions of chatbots and humans using emoji in computer-mediated communication. Communication Studies, 71(3), 409–427. https://doi.org/10.1080/10510974.2020.1725082

Beckert

Koch

. (2022). Why context matters in a changing social media environment: The role of placement prominence as a moderator of disclosure effects in sponsored influencer content. Studies in Communication Sciences, 22(1), 3061. https://doi.org/10.24434/j.scoms.2022.01.3061

Beckert

Koch

Viererbl

Schulz-Knappe

. (2021). The disclosure paradox: How persuasion knowledge mediates disclosure effects in sponsored media content. International Journal of Advertising, 40(7), 1160–1186. https://doi.org/10.1080/02650487.2020.1859171

Boerman

S. C.

van Reijmersdal

E. A.

Rozendaal

Dima

A. L

. (2018). Development of the persuasion knowledge scales of sponsored content (PKS-SC). International Journal of Advertising, 37(5), 671–697. https://doi.org/10.1080/02650487.2018.1470485

Brehm

S. S.

Brehm

J. W

. (1981). Psychological reactance: A theory of freedom and control. Academic.

Burgoon

J. K

. (1993). Interpersonal expectations, expectancy violations, and emotional communication. Journal of Language and Social Psychology, 12(1–2), 30–48. https://doi.org/10.1177/0261927X93121003

Burgoon

J. K.

Le Poire

B. A

. (1993). Effects of communication expectancies, actual communication, and expectancy disconfirmation on evaluations of communicators and their communication behavior. Human Communication Research, 20(1), 67–96. https://doi.org/10.1111/j.1468-2958.1993.tb00316.x

10.

Campbell

M. C.

Kirmani

. (2008). I know what you’re doing and why you’re doing it: The use of the Persuasion Knowledge Model in consumer research. In Haugvedt

Herr

P. M.

Kardes

F. R.

(Eds.), Handbook of consumer psychology (pp. 549–571). Lawrence Erlbaum Associates.

11.

Cavalheiro

B. P.

Prada

Rodrigues

D. L.

Garrido

M. V.

Lopes

. (2022). With or without emoji? Perceptions about emoji use in different brand-consumer communication contexts. Human Behavior and Emerging Technologies, 2022, 1–8. https://doi.org/10.1155/2022/3036664

12.

Centers for Disease Control and Prevention. (2022a). Chlamydia—CDC basic fact sheet. https://www.cdc.gov/std/chlamydia/stdfact-chlamydia.html

13.

Centers for Disease Control and Prevention. (2022b). What can I do to reduce my risk of skin cancer? https://www.cdc.gov/cancer/skin/basic_info/prevention.htm

14.

Danesi

. (2022). Emotional wellbeing and the semiotic translation of emojis. In Petrilli

(Eds.), Exploring the translatability of emotions (pp. 323–344). Springer. https://doi.org/10.1007/978-3-030-91748-7_12

15.

Derks

Fischer

A. H.

Bos

A. E. R

. (2008). The role of emotion in computer-mediated communication: A review. Computers in Human Behavior, 24(3), 766–785. https://doi.org/10.1016/j.chb.2007.04.004

16.

De Veirman

Hudders

. (2020). Disclosing sponsored Instagram posts: The role of material connection with the brand and message-sidedness when disclosing covert advertising. International Journal of Advertising, 39(1), 94–130. https://doi.org/10.1080/02650487.2019.1575108

17.

Dillard

J. P.

Shen

. (2005). On the nature of reactance and its role in persuasive health communication. Communication Monographs, 72(2), 144–168. https://doi.org/10.1080/03637750500111815

18.

Distel

Egger

Petrovic

Phan

V. L.

Wiesinger

. (2022). The usage of emoji in tourism-related instagram posts: Suggestions from a marketing perspective. In Stienmetz

J. L.

Ferrer-Rosell

Massimo

(Eds.), Information and communication technologies in tourism 2022 (pp. 134–145). Springer. https://doi.org/10.1007/978-3-030-94751-4_13

19.

Dresner

Herring

S. C

. (2010). Functions of the nonverbal in CMC: Emoticons and illocutionary force. Communication Theory, 20(3), 249–268. https://doi.org/10.1111/j.1468-2885.2010.01362.x

20.

Dunlap

J. C.

Bose

Lowenthal

P. R.

York

C. S.

Atkinson

Murtagh

. (2016). What sunshine is to flowers. In Tettegah

S. Y.

Gartmeier

(Eds.), Emotions, technology, design, and learning (pp. 163–182). Elsevier. https://doi.org/10.1016/B978-0-12-801856-9.00008-6

21.

Faul

Erdfelder

Lang

A.-G.

Buchner

. (2007). G*Power 3: A flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behavior Research Methods, 39(2), 175–191. https://doi.org/10.3758/BF03193146

22.

Friestad

Wright

. (1994). The persuasion knowledge model: How people cope with persuasion attempts. Journal of Consumer Research, 21(1), 1. https://doi.org/10.1086/209380

23.

Ganster

Eimler

S. C.

Krämer

N. C

. (2012). Same same but different!? The differential influence of smilies and emoticons on person perception. Cyberpsychology, Behavior, and Social Networking, 15(4), 226–230. https://doi.org/10.1089/cyber.2011.0179

24.

Ham

C.-D.

Nelson

M. R.

Das

. (2015). How to measure persuasion knowledge. International Journal of Advertising, 34(1), 17–53. https://doi.org/10.1080/02650487.2014.994730

25.

Hovland

C. I.

Janis

I. L.

Kelley

H. H

. (1953). Communication and persuasion. Psychological studies of opinion change. Yale University Press.

26.

Indwar

R. G.

Mishra

A. S

. (2022). Emojis: Can it reduce post-purchase dissonance? Journal of Strategic Marketing, 31, 918–939. https://doi.org/10.1080/0965254X.2021.2013935

27.

Jibril

T. A.

Abdullah

M. H

. (2013). Relevance of emoticons in computer-mediated communication contexts: An overview. Asian Social Science, 9(4), p201. https://doi.org/10.5539/ass.v9n4p201

28.

Kaye

L. K.

Wall

H. J.

Malone

S. A

. (2016). “Turn that frown upside-down”: A contextual account of emoticon usage on different virtual platforms. Computers in Human Behavior, 60, 463–467. https://doi.org/10.1016/j.chb.2016.02.088

29.

Klauda

S. L.

Guthrie

J. T

. (2008). Relationships of three components of reading fluency to reading comprehension. Journal of Educational Psychology, 100(2), 310–321. https://doi.org/10.1037/0022-0663.100.2.310

30.

Koch

Zerback

. (2013). Helpful or harmful? How frequent repetition affects perceived statement credibility: Helpful or harmful. Journal of Communication, 63(6), 993–1010. https://doi.org/10.1111/jcom.12063

31.

Kralj Novak

Smailović

Sluban

Mozetič

. (2015). Sentiment of emojis. PLOS ONE, 10(12), e0144296. https://doi.org/10.1371/journal.pone.0144296

32.

Krouwer

Poels

Paulussen

. (2017). To disguise or to disclose? The influence of disclosure recognition and brand presence on readers’ responses toward native advertisements in online news media. Journal of Interactive Advertising, 17(2), 124–137. https://doi.org/10.1080/15252019.2017.1381579

33.

Lee

Kim

Lee

K. C

. (2021). Investigating the negative effects of emojis in facebook sponsored ads for establishing sustainable marketing in social media. Sustainability, 13(9), 4864. https://doi.org/10.3390/su13094864

34.

Leiner

D. J

. (2016). Our research’s breadth lives on convenience samples: A case study of the online respondent pool “SoSci Panel.” Studies in Communication—Media, 5(4), 367–396. https://doi.org/10.5771/2192-4007-2016-4-367

35.

Ljubešić

Fišer

. (2016). A global analysis of emoji usage. In Proceedings of the 10th web as corpus workshop (pp. 82–89). http://nlp.ffzg.hr/data/publications/nljubesi/ljubesic16-global.pdf

36.

S.-K

. (2008). The nonverbal communication functions of emoticons in computer-mediated communication. Cyberpsychology & Behavior, 11(5), 595–597. https://doi.org/10.1089/cpb.2007.0132

37.

McShane

Pancer

Poole

Deng

. (2021). Emoji, playfulness, and brand engagement on Twitter. Journal of Interactive Marketing, 53, 96–110. https://doi.org/10.1016/j.intmar.2020.06.002

38.

Meirick

P. C.

Nisbett

G. S

. (2011). I approve this message: Effects of sponsorship, ad tone, and reactance in 2008 presidential advertising. Mass Communication and Society, 14(5), 666–689. https://doi.org/10.1080/15205436.2010.530381

39.

Ohanian

. (1990). Construction and validation of a scale to measure celebrity endorsers’ perceived expertise, trustworthiness, and attractiveness. Journal of Advertising, 19(3), 39–52. https://doi.org/10.1080/00913367.1990.10673191

40.

Parks

C. M.

Toth

J. P

. (2006). Fluency, familiarity, aging, and the illusion of truth. Aging, Neuropsychology, and Cognition, 13(2), 225–253. https://doi.org/10.1080/138255890968691

41.

Pelletier

M. J.

Krallman

Adams

F. G.

Hancock

. (2020). One size doesn’t fit all: A uses and gratifications analysis of social media platforms. Journal of Research in Interactive Marketing, 14(2), 269–284. https://doi.org/10.1108/JRIM-10-2019-0159

42.

Prada

Rodrigues

D. L.

Garrido

M. V.

Lopes

Cavalheiro

Gaspar

. (2018). Motives, frequency and attitudes toward emoji and emoticon use. Telematics and Informatics, 35(7), 1925–1934. https://doi.org/10.1016/j.tele.2018.06.005

43.

Prada

Saraiva

Cruz

Xavier

Rodrigues

D. L

. (2022). Using emoji in response to customer reservation requests and service reviews. Human Behavior and Emerging Technologies, 2022, 1–10. https://doi.org/10.1155/2022/1433055

44.

Rau

M. A

. (2018). Making connections among multiple visual representations: How do sense-making skills and perceptual fluency relate to learning of chemistry knowledge? Instructional Science, 46(2), 209–243. https://doi.org/10.1007/s11251-017-9431-3

45.

Reber

Schwarz

. (1999). Effects of perceptual fluency on judgments of truth. Consciousness and Cognition, 8(3), 338–342. https://doi.org/10.1006/ccog.1999.0386

46.

Reber

Schwarz

Winkielman

. (2004). Processing fluency and aesthetic pleasure: Is beauty in the perceiver’s processing experience? Personality and Social Psychology Review, 8(4), 364–382. https://doi.org/10.1207/s15327957pspr0804_3

47.

Reber

Winkielman

Schwarz

. (1998). Effects of perceptual fluency on affective judgments. Psychological Science, 9(1), 45–48. https://doi.org/10.1111/1467-9280.00008

48.

Riordan

M. A

. (2017). The communicative role of non-face emojis: Affect and disambiguation. Computers in Human Behavior, 76, 75–86. https://doi.org/10.1016/j.chb.2017.07.009

49.

Robertson

Magdy

Goldwater

. (2018). Self-representation on Twitter using emoji skin color modifiers. https://homepages.inf.ed.ac.uk/sgwater/papers/icwsm18-skintone.pdf

50.

Sampietro

. (2019). Emoji and rapport management in Spanish WhatsApp chats. Journal of Pragmatics, 143, 109–120. https://doi.org/10.1016/j.pragma.2019.02.009

51.

Seghers

De Clerck

Lybaert

. (2021). When form deviates from the norm: Attitudes towards old and new vernacular features and their impact on the perceived credibility and usefulness of Facebook consumer reviews. Language Sciences, 87, 101413. https://doi.org/10.1016/j.langsci.2021.101413

52.

Stark

Crawford

. (2015). The conservatism of emoji: Work, affect, and communication. Social Media + Society, 1(2), 205630511560485. https://doi.org/10.1177/2056305115604853

53.

Stelzner

. (2023). Social media marketing industry report. https://www.socialmediaexaminer.com/social-media-marketing-industry-report-2023/

54.

Tang

Hew

K. F

. (2019). Emoticon, emoji, and sticker use in computer-mediated communication: A review of theories and research findings. International Journal of Communication, 13, 2457–2483.

55.

Tutaj

van Reijmersdal

E. A

. (2012). Effects of online advertising format and persuasion knowledge on audience reactions. Journal of Marketing Communications, 18(1), 5–18. https://doi.org/10.1080/13527266.2011.620765

56.

Unkelbach

. (2007). Reversing the truth effect: Learning the interpretation of processing fluency in judgments of truth. Journal of Experimental Psychology: Learning, Memory, and Cognition, 33(1), 219–230. https://doi.org/10.1037/0278-7393.33.1.219

57.

van Reijmersdal

E. A.

Fransen

M. L.

van Noort

Opree

S. J.

Vandeberg

Reusch

. . . Boerman

S. C

. (2016). Effects of disclosing sponsored content in blogs: How the use of resistance strategies mediates effects on persuasion. American Behavioral Scientist, 60(12), 1458–1474. https://doi.org/10.1177/0002764216660141

58.

Walczyk

J. J.

Griffith-Ross

D. A

. (2007). How important is reading skill fluency for comprehension? The Reading Teacher, 60(6), 560–569. https://doi.org/10.1598/RT.60.6.6

59.

Walther

J. B.

D’Addario

K. P

. (2001). The impacts of emoticons on message interpretation in computer-mediated communication. Social Science Computer Review, 19(3), 324–347. https://doi.org/10.1177/089443930101900307

60.

Waterloo

S. F.

Baumgartner

S. E.

Peter

Valkenburg

P. M

. (2018). Norms of online expressions of emotion: Comparing Facebook, Twitter, Instagram, and WhatsApp. New Media & Society, 20(5), 1813–1831. https://doi.org/10.1177/1461444817707349

61.

Willoughby

J. F.

Liu

. (2018). Do pictures help tell the story? An experimental test of narrative and emojis in a health text message intervention. Computers in Human Behavior, 79, 75–82. https://doi.org/10.1016/j.chb.2017.10.031

62.

Yakın

Eru

. (2017). An application to determine the efficacy of emoji use on social marketing ads. International Journal of Social Sciences and Education Research, 3(1), 230–230. https://doi.org/10.24289/ijsser.270652

Funny but not Credible? Why Using (Many) Emojis Decreases Message Credibility and Source Trustworthiness

Abstract

Keywords

Functions of Emojis

Effects of Emojis

Study 1a

Method

Sample

Design and Procedure

Stimulus

Measures

Treatment Check

Results

Discussion

Study 1b

Method

Sample

Design and Procedure

Stimulus

Measures

Treatment Check

Results

Discussion

Study 2

Method

Sample

Design and Procedure

Stimulus

Measures

Treatment Check

Results

Discussion

General Discussion

Footnotes

Declaration of Conflicting Interests

Funding

ORCID iD

Author Biographies

References