Does Physiological Arousal Increase Social Transmission of Information? Two Replications of Berger (2011)

Abstract

People share information for many reasons. For example, Berger (2011, N = 40) found that undergraduate participants manipulated to have higher physiological arousal were more likely to share a news article with others via email than people who had low arousal. Berger’s research is widely cited as evidence of the causal role of arousal in sharing information and has been used to explain why information that induces high-arousal emotions is shared more than information that induces low-arousal emotions. We conducted two replications (N = 111, N = 160) of Berger’s study, using the same arousal manipulation but updating the sharing measure to reflect the rise of information sharing through social media. Both studies failed to find an impact of incidental physiological arousal on undergraduate participants’ willingness to share news articles on social media. Our studies cast doubt on the idea that incidental physiological arousal—in the absence of other factors—impacts people’s decisions to share information on social networking sites.

Keywords

arousal social transmission sharing replication open data open materials preregistered

With the rise in popularity of social media, this medium has become a common way for people to consume news (Ling, 2020; Pentina & Tarafdar, 2014). On social media platforms, consumers are able to curate their news, selecting what to share with others in their network. Many studies have examined why people might share various types of information with others (for reviews, see Baptista & Gradim, 2020; Berger, 2014; Kümpel et al., 2015; Rimé, 2009). In a widely cited example, Berger (2011) found that physiological arousal plays a causal role in people’s decisions to share information. We conducted two replications of Berger’s influential research.

Motivation for Social Transmission

Many factors have been shown to influence whether people share information, including how unexpected the information is (Al-Rawi, 2017), whether it is believable (Heath, 1996), whether the information is perceived to be useful (Alexandrov et al., 2013), and the emotions the information elicits (Berger & Milkman, 2010). For example, Berger and Milkman (2010) examined how often online news articles were emailed to others and the emotional content in those articles. They found that articles that evoked stronger emotions—regardless of whether the emotions were positive or negative—were shared more often than articles that did not elicit emotional responses. Similarly, Stieglitz and Dang-Xuan (2013) found that emotionally charged tweets were retweeted (i.e., shared) more often and more quickly than ones that were not emotionally charged. Another study of people’s social media activity found that people posted about emotionally charged issues (e.g., politics and crime) more often than any other subject and that 67% of these public posts were done to promote an ideology or express anger (Lottridge & Bentley, 2018). The overall conclusion seems to be that information that evokes emotional arousal is shared more than nonemotional information, whether the information is shared via email or on social media (see also Berger & Milkman, 2012; Brady et al., 2017; Hasell & Weeks, 2016; Heath, 1996; Peters et al., 2009).

In an attempt to explain why emotions—particularly high-arousal emotions—increase sharing, Berger (2011) sought to investigate whether physiological arousal alone could act as a causal factor in determining the rate of social transmission of information. In the first of two studies, Berger induced high-arousal (anxiety or amusement) or low-arousal (sadness or contentment) emotions by having participants watch film clips. In an ostensibly unrelated task, participants were shown a neutral news article and video and asked if they would share them with friends, family members, or coworkers. Berger found that the high-arousal emotions (regardless of whether they were positive or negative) increased participants’ willingness to share the article and video with others relative to the low-arousal emotions.

Berger’s (2011) first study was important because it showed that incidental arousal (i.e., arousal not caused by the information to be shared) could influence people’s willingness to share information. One issue, however, was that the physiological arousal was inducted through emotion manipulations—that is, emotional and physiological arousal were confounded. Therefore, it was unclear whether the increase in sharing was caused by the emotions that were induced (anxiety and amusement) or by the increased physiological arousal. In other words, Berger’s first study was unable to determine whether physiological arousal—in the absence of emotional arousal—could increase social transmission of information. To address this issue, Berger conducted a second study aimed at isolating the potential impact of physiological arousal.

In Berger’s (2011) second study, 40 students were randomly assigned to either jog for 60 s or sit quietly for 60 s. Jogging in place has been shown to increase physiological arousal (Wegner & Giuliano, 1980) and (importantly for Berger’s purposes) did so without also influencing participants’ emotions. After the arousal manipulation, participants were presented with a distractor task before reading a neutral online news article. Finally, participants indicated whether they would email the article to anyone they wanted. Berger found that participants who jogged in place (i.e., those in a high-arousal state) were more than twice as likely to share the news article than participants who sat quietly (i.e., those in a low-arousal state). These results led Berger to conclude that “Situations that heighten arousal should boost social transmission, regardless of whether they are positive (e.g., inaugurations) or negative (e.g., panics) in nature” (p. 892). Although other studies have shown that emotions can influence what information people share, Berger’s second study is the only one that did not confound emotional and physiological arousal, thereby demonstrating that physiological arousal plays a causal role in the social transmission of information.

Statement of Relevance

The sharing of information permeates our daily lives, whether for protection, the spreading of knowledge, or simply for social communication. People’s decisions about what information to share with others is likely to be influenced by a wide variety of factors. Berger (2011) identified one of those factors as a person’s level of physiological arousal. Berger’s research has been widely cited and is one of the few studies that has examined the causal role of physiological arousal, in the absence of other factors. We conducted two replications of Berger’s research, updating the materials to reflect the rise of social media as a primary way for people to communicate with each other. Neither of our studies found evidence that physiological arousal increased sharing of information, casting doubt on the role of arousal in the social transmission of information.

Rationale for Replication

Berger’s (2011) research has been influential in shaping our understanding of the factors that influence people’s decisions to share information and in understanding why information might go viral (Berger & Milkman, 2012). As one measure of the impact of Berger’s (2011) research, the article has been cited over 820, times according to Google Scholar. As noted earlier, many studies have examined the role that emotional arousal plays in sharing information (e.g., Berger & Milkman, 2010, 2012; Brady et al., 2017; Hasell & Weeks, 2016; Heath, 1996; Lottridge & Bentley, 2018; Peters et al., 2009; Stieglitz & Dang-Xuan, 2013), but very few have examined the influence of physiological arousal in the absence of emotional arousal. Given that one mechanism proposed to explain why emotionally laden information is shared is that it induces physiological arousal (Berger & Milkman, 2012), Berger’s research stands as one of the lone tests of this claim. Furthermore, many reviews examining factors that influence the sharing of information cite Berger’s research as evidence that physiological arousal plays a causal role (e.g., Baptista & Gradim, 2020; Berger, 2014; Kümpel et al., 2015). Despite the impact of Berger’s (2011) research, his studies have not been directly replicated.

The lack of independent replications of Berger’s research is potentially problematic given the somewhat low rates of successful replications of psychological research (e.g., Camerer et al., 2018; Ebersole et al., 2016; Open Science Collaboration, 2015). The so-called replication crisis has led many to advocate for an increase in the number of replications conducted and published (Simons, 2014). Perhaps more important than the generally poor replicability of psychological research broadly is that aspects of Berger’s (2011) research are not considered best practices by today’s standards. For example, Berger’s studies were not preregistered and relied on relatively small samples. To be clear, these practices were commonplace and accepted when the studies were conducted. However, more recently, recommendations have been made to help address issues that have plagued psychological research (Simmons et al., 2011).

The Current Studies

Given the influence of Berger’s (2011) research and the lack of direct replications, we conducted two studies testing whether physiological arousal—when not confounded with emotional arousal—increases sharing of information. We closely followed Berger’s procedures while updating the measure of information sharing to better reflect how people are likely to share news. Specifically, rather than asking whether participants would email a news article to other people, we showed participants headlines and asked whether they would share the news articles on social-media sites (e.g., Facebook or X). Self-reported willingness to share information on social-media sites is a relatively common measure (e.g., Pennycook et al., 2018; Pennycook & Rand, 2018; Pereira et al., 2023) and has been shown to correlate with sharing behavior on actual social-media sites (Mosleh et al., 2020). In both studies, we tested Berger’s (2011) claim that incidentally induced physiological arousal can increase people’s willingness to share information with others.

Open Practices Statement

The materials and data can be found on the Open Science Framework (https://osf.io/azc87/). Study 1 (https://osf.io/gxsfa) and Study 2 (https://osf.io/28gr7) were both preregistered.

Studies 1 and 2

Studies 1 and 2 were identical (i.e., used the same procedures, stimuli, and measures) but were conducted approximately 4 months apart.¹ As described above, we followed the procedures used by Berger (2011)² with a few exceptions. The arousal manipulation (jog in place for 60 s vs. sit quietly for 60 s) and filler task (rate the brightness of images) were the same as in the original study, but the measure of social transmission of information was updated to better reflect how most people consume and share news. We also included manipulation-check measures to ensure the manipulation successfully influenced physiological arousal (see Table 1 for a description of how the replication studies deviated from the original). For both studies, we reported how we determined our sample size, all data exclusions, all manipulations, and all measures used in the studies.

Table 1.

Summary of Similarities and Discrepancies Between Original Study and Replication Study

Study element	Original study	Replication studies	Reasoning for the difference between original and replication
Sample size	40 participants	271 participants total (111 in Study 1, 160 in Study 2)	Larger sample to increase the precision of the effect size estimate and ensure a high likelihood of detecting a smaller effect size than the original study
Sample population	Undergraduate students	Undergraduate students	—
Manipulation	Jog in place for 60 s vs. rest for 60 s	Jog in place for 60 s vs. rest for 60 s	—
Distractor task	Rate brightness of five images	Rate brightness of five images	—
Stimuli	One news article	10 news articles	Increase generalizability and ensure that features of a specific news article did not exaggerate or weaken the effect
Measure of sharing	Share news article via email	Share news articles on social media	People are much more likely to share news on social media than by email; therefore, we updated the measure to reflect what people more commonly do
Manipulation check	No manipulation check reported	Conducted a manipulation check	Ensure the manipulation worked as expected
Exploratory mood measure	Participants rated how positively they felt	Measured positive and negative affect (PANAS)	Used a widely used and well-validated measure of both positive and negative affect

Note: PANAS = Positive and Negative Affect Schedule.

Method

Participants

Undergraduate students from Appalachian State University received partial course credit in exchange for their participation. For both studies, the target sample size was at least 100 participants, which is 2.5 times the sample size of the original study, as recommended by Simonsohn (2015). In Study 1, we recruited 114 participants. On the basis of our preregistered exclusion criteria, we excluded 1 participant who did not agree to jog in place for 60 s and 2 participants who did not have social-media accounts, leaving a final sample of 111 (66.7% female, 30.2% male, 3.1% nonbinary; M_age = 18.7 years, SD = 1.39). In Study 2, we recruited 162 participants and excluded 2 participants who did not have social-media accounts, leaving a final sample of 160 (65.6% female, 31.9% male, 2.5% nonbinary; M_age = 19.53 years, SD = 1.30). Both studies were approved by the Institutional Review Board at Appalachian State University.

Sensitivity power analyses revealed that we had an 80% chance of detecting effect sizes (ds) of 0.54 and 0.44 for Study 1 and Study 2, respectively. Both studies had a greater than 99% chance of detecting an effect equal to or larger than the effect observed in Berger’s (2011) study (d = 0.89).

Procedure

The study was advertised to potential participants as examining the influence of physical activity on people’s perceptions. On the study’s recruitment page and in the informed-consent document, the participants were informed that they would be asked to either sit quietly for 60 s or jog in place for 60 s. Each participant completed the study individually with one research assistant. In each study session, once consent was given, participants were instructed to begin a study on the computer. The first question reminded participants that they would be asked to either sit or jog for 60 s and asked whether they felt they were able to jog in place for 60 s, should they be asked to perform that task. As noted above, all but 1 participant in Study 1 and all participants in Study 2 indicated they were able to jog in place for 60 s.

Participants were then randomly assigned to a jogging condition (high physiological arousal) or sitting condition (low physiological arousal). Participants in the jogging condition were given instructions verbally by the research assistant that they would jog lightly in place for 60 s, whereas participants in the sitting condition were instructed they would sit quietly for 60 s (see the Supplemental Material available online for the exact instructions).

After jogging or sitting for 60 s, participants returned to the computer and completed the filler task used by Berger (2011). Specifically, participants rated the brightness of five landscape images on a scale ranging from 1 (very dark) to 7 (very bright).

Participants then read instructions about the article-sharing task (see the Supplemental Material). The participants were shown 10 news articles one at a time in a random order. Each news article included the article headline along with an image and the first sentence of the article (see the Supplemental Material). The presentation of the new articles was modeled after previous research examining factors related to sharing news on social-media sites (e.g., Mosleh et al., 2020; Pennycook et al., 2018; Pennycook & Rand, 2018) and was designed to resemble how people might see news headlines on social-media sites. Following Berger’s study, the news articles were selected to be relatively neutral so as not to evoke emotional responses. For each article, participants were asked, “If you came across this article on your own, would you share the article on social media?” and selected either yes or no.

Next, participants completed measures assessing physiological arousal that served as manipulation checks. Berger (2011) did not report including a manipulation check in the study we replicated (Study 2) but did report using a manipulation check in the other study reported in his article (Study 1). Therefore, we included Berger’s manipulation check that asked participants to indicate how they felt while going through the study; they responded on three 7-point response scales ranging from very passive—very active, very mellow—very fired up, and very low energy—very high energy. Responses to these items were highly correlated (α = .876 in Study 1 and α = .852 in Study 2), so they were combined into an overall arousal score. We added two new manipulation-check questions that asked people, “During this study, how fast did it feel like your heart was beating?” (1 = very slow, 4 = very fast) and “How relaxed did you feel while going through the study?” (1 = not at all relaxed, 4 = very relaxed).

After completing the manipulation-check questions, participants reported positive and negative affect using the Positive and Negative Affect Schedule (PANAS; Watson et al., 1988). The participants were then a checklist of eight popular social-media sites— Facebook, Instagram, TikTok, Snapchat, Pinterest, Twitter (now X), Reddit, and LinkedIn—and asked which social-media sites they regularly use; they were allowed write in any other sites they used that were not on the list. Finally, the participants were asked their gender and age, thanked for their participation, and dismissed.

Results

Manipulation checks

We first compared participants’ self-reported arousal for the high- and low-arousal conditions. In both studies, participants in the high-arousal condition reported higher levels of arousal than participants in the low-arousal condition—Study 1: t(109) = 9.23, p < .001, d = 1.75, 95% confidence interval (CI) = [1.25, 2.24]; Study 2: t(158) = 7.59, p < .001, d = 1.20, 95% CI = [0.84, 1.56]. Next, we evaluated the additional manipulation-check questions and found that in both studies, participants in the high-arousal condition reported having a faster heart rate than participants in the low-arousal condition—Study 1: t(109) = 7.87, p < .001, d = 1.49, 95% CI = [1.03, 1.95]; Study 2: t(158) = 10.61, p < .001, d = 1.68, 95% CI = [1.27, 2.08]. Similarly, in both studies, participants in the high-arousal condition reported being less relaxed than participants in the low-arousal condition—Study 1: t(109) = 2.21, p = .029, d = 0.42, 95% CI = [0.04, 0.80]; Study 2: t(158) = 3.39, p < .001, d = 0.54, 95% CI = [0.21, 0.86].

Taken together, these analyses suggest that the manipulation of physiological arousal was successful because in both studies participants in the high-arousal condition reported higher self-reported arousal and a faster heart rate; they also reported feeling less relaxed than participants in the low-arousal condition.

Confirmatory analyses

For each participant, we calculated the percentage of news articles they reported they would share on social media out of the 10 news articles they saw. Overall, participants in Study 1 reported they would share 37.4% of the articles, and participants in Study 2 reported they would share 32.7% of the articles. We next compared the percentage of articles shared for the high- and low-arousal conditions. In Study 1, participants in the high-arousal condition (M = 38.3%, SD = 24.7%) shared about the same number of articles as participants in the low-arousal condition (M = 36.5%, SD = 24.3%), t(109) = 0.40, p = .693, d = 0.08, 95% CI = [−0.30, 0.45]. Similarly, in Study 2, participants in the high-arousal condition (M = 33.3%, SD = 23.5%) shared about the same number of articles as participants in the low-arousal condition (M = 32.1%, SD = 23.6%), t(158) = 0.32, p = .750, d = 0.05, 95% CI = [−0.26, 0.36]. In summary, and in contrast to Berger (2011), we did not find evidence that increased physiological arousal corresponded to an increase in the number of articles shared by the participants (see Table 2 for a comparison of our studies and Berger’s study).

Table 2.

Comparison of Berger’s (2011) Study and the Replication Studies

		Percentage shared
Study	N	Low arousal	High arousal	p value	Effect size (d)	95% confidence interval
Berger (2011)	40	33%	75%	< .01	0.89	[0.22, 1.57]
Study 1	111	36.5%	38.3%	.693	0.08	[−0.30, 0.45]
Study 2	160	32.1%	33.3%	.750	0.05	[−0.26, 0.36]
Mini-meta analysis	311	33.8%	39.7%	.159	0.16	[−0.06, 0.39]

Exploratory analyses

One of the goals of Berger’s study was to manipulate physiological arousal independent of mood. Berger did not report whether mood was influenced by the arousal manipulation, but we did examine whether positive and negative affect varied across our arousal conditions. In both studies, negative affect did not vary across the arousal conditions—Study 1: t(109) = 0.66, p = .51, d = 0.13, 95% CI = [−0.25, 0.50]; Study 2: t(158) = 0.72, p = .47, d = 0.11, 95% CI = [−0.20, 0.42]. However, in both studies, participants in the high-arousal condition reported higher levels of positive affect than participants in the low-arousal condition—Study 1: t(109) = 4.32, p < .001, d = 0.82, 95% CI = [0.42, 1.22]; Study 2: t(158) = 3.82, p < .001, d = 0.61, 95% CI = [0.28, 0.93]. Therefore, it is possible that jogging in place increased people’s mood relative to sitting still. It is also possible that this effect was driven by the positive-affect measure that includes terms related to high physiological arousal (e.g., alert, excited, and active). As follow-up analyses, we examined which specific items differed across the two arousal conditions and found that in both studies participants in the high-arousal condition gave significantly higher ratings for excited, strong, enthusiastic, proud, alert, determined, and active; there were no differences between conditions in participants’ reported levels of interested, inspired, or attentive.

Next, we examined the relationships between dependent variables (i.e., self-reported arousal, self-reported heart rate, relaxation, positive and negative affect, and sharing of the articles; see the Appendix Table A1). Berger (2011) reported that sharing was unrelated to participants’ positivity ratings. In Study 1, sharing was positively correlated with positive affect, r(109) = .32, p < .001, 95% CI = [.14, .48], but in Study 2, these two variables were not significantly correlated, r(158) = .04, p = .64, 95% CI = [−.12, .19].

One other set of correlations was noteworthy. As would be expected from the lack of an effect of the arousal manipulation on sharing, in Study 1, the correlation between self-reported arousal and sharing was not significant, r(109) = .14, p = .131, 95% CI = [−.04, .32]. However, in Study 2, self-reported arousal was positively related to sharing, r(158) = .23, p = .004, 95% CI = [.08, .37]. Similarly, in Study 1 self-reported heart rate was not associated with sharing, r(109) = −.01, p = .91, 95% CI = [−.20, .18], but in Study 2, self-reported heart rate was positively related to sharing, r(158) = .18, p = .02, 95% CI = [.03, .33].

The significant correlation between sharing and self-reported arousal could be interpreted as providing support for Berger’s (2011) research showing that physiological arousal drives social transmission of information. However, as noted above, sharing articles was significantly related to self-reported measures of arousal in only one study. Perhaps more importantly, the decision to share news articles preceded the measure of self-reported arousal. Therefore, from a methodological standpoint, a more appropriate conclusion would be that choosing to share information on social media influences self-reported arousal. Given the inconsistency in findings and the issue with temporal precedence, we are reluctant to draw definitive conclusions from this correlation.

Finally, we examined two possible reasons for the null results of our preregistered analysis. First, because we asked participants whether they would share 10 different articles, it is possible they felt compelled to agree to share at least some of the articles.³ To test this explanation, we examined just the first article with which each participant was presented. In Study 1, there was not a significant difference between the low arousal condition (24.6%) and the high-arousal condition (33.3%) in terms of the percentage of people who shared the first article they encountered, χ²(1, N = 111) = 1.04, p = .308. ϕ = .097. Similarly, in Study 2, there was not a significant difference between the number of the first encountered articles shared by participants in the low-arousal condition (29.6%) and the high-arousal condition (22.8%), χ²(1, N = 160) = 0.97, p = .325. ϕ = .078.

The second possible explanation for the null results is that the physiological arousal manipulation only influenced sharing for some, but not all, of the articles we selected. To test this possibility, we conducted chi-square tests comparing how many people shared the article across the two arousal conditions. In each chi-square analysis, there was not a significant difference between the two conditions, suggesting that a subset of articles did not drive the null results (see the Supplemental Material for these analyses).

Evaluating the Evidence

There are multiple ways to evaluate the balance of evidence testing the claim that physiological arousal—independent of emotional arousal—can influence the social transmission of information. Because these methods have strengths and weaknesses, we used four different methods to evaluate how our replication studies compare to the original study.

Statistical significance

Perhaps the most straightforward evaluation is to examine whether the replication studies found statistically significant results in the same direction as the original study. Our two replication attempts, of course, did not find statistically significant results. Therefore, on balance, our two studies did not find the predicted effect; only Berger’s (2011) did. Given that our replications had much larger sample sizes than the original study, the evidence does not support Berger’s conclusion that arousal increases sharing of social information.

Comparing effect sizes

LeBel et al. (2019) described a framework for comparing effect sizes of replication studies with the effect size in the original study. Specifically, they recommended examining the replications for a signal (i.e., checking whether the 95% CI of the effect size of the replication included zero) and probing the consistency of the replications with the original effect size (i.e., checking whether the 95% CI of the effect sizes of the replications included the original effect size). In both studies, the 95% CIs included zero and did not include the original effect size. Therefore, using LeBel et al.’s terminology, our replications would be considered “no signal – inconsistent” and the “least favorable outcome” in terms of evidence providing support for the findings in the original study.

Mini meta-analysis

A third method for evaluating evidence is to examine the effect size when the original study and replications are combined. Following the procedures described by Goh et al. (2016), we conducted a mini meta-analysis that included the original study and our two replications. This analysis found that the average effect size weighted by sample size was small, d = 0.16, 95% CI = [−.063, .385], and not significantly different from 0, p = .159. In other words, combining the original study with the two replications did not produce a significant effect. It is possible that the effect of arousal on social transmission exists but is small. However, if the true effect size is d = 0.161, as was found in our mini meta-analysis, a study would need 1,214 participants to have an 80% chance of detecting the effect. Berger’s (2011) study with a sample size of 40 would have only an 8% chance of detecting an effect that small.

Detectability

Simonsohn (2015) described an alternative method for evaluating replications by examining whether the effects found in the replications were “. . . close enough to zero that the original study would have been unable to meaningfully study an effect that small” (p. 560). This is done by assuming that the original study was powered at only 33%. The original study had a sample size of 40. If it had 33% power, the effect would be d = 0.49 (referred to as d_33% = 0.49 by Simonsohn, 2015). The effect sizes found in Study 1 (d = 0.08) and Study 2 (d = 0.05) were both significantly smaller than d = 0.49 (p = .012 and p = .004 for Studies 1 and 2, respectively). This is inconsistent with the idea that the effect was large enough for Berger (2011) to detect given the sample size he used. Therefore, according to Simonsohn (2015), our replication studies would be considered informative failures; we failed to replicate the original finding.

Summary of evaluating the evidence

All four methods of evaluating the replications lead to the same conclusion: Neither of our studies found that increasing arousal significantly increased sharing of information on social media. The effect sizes of both of our studies were not different from zero and were significantly smaller than that of the original study; the average effect size of the two replications and the original study was not different from zero; and the results of the replications suggest that the sample size of the original study was not sufficient to detect an effect. In summary, considering the original study and the replication studies, the balance of evidence suggests that increasing physiological arousal—in the absence of emotional arousal—does not increase the sharing of social information on social media.

General Discussion

Berger (2011) examined the impact of incidental physiological arousal on people’s willingness to share information with others on social media. We conducted two replications of Berger’s second study, as that was the only study that isolated the influence of physiological arousal from other factors (e.g., emotional arousal). Despite the success of the arousal manipulation, participants in the high- and low-arousal conditions did not differ in their willingness to share news articles with other people. We used four different methods to evaluate the replications, and all led to the same conclusion: We did not find evidence that increased incidental physiological arousal influences sharing of information on social-networking sites.

There are, of course, a number of reasons as to why the results of our studies differed from the results of Berger’s (2011) study. Although we used the same arousal manipulation, we changed the measure of social transmission. Instead of asking people whether they would share one news article via email, we showed participants 10 articles and asked whether they would share them on social media. Therefore, it is possible that arousal increases social transmission via some methods (e.g., email) but not others (e.g., social media). Sharing new stories via email is a more private means of communication, where users can choose with whom they share the information. On the other hand, social media is a more public way of sharing information, making it hard for users to control who sees their posts. The difference between the nature of email and social media may be responsible for the difference between Berger’s results and our findings.

Another potential contributor to the differences between the findings is that Berger’s (2011) studies were conducted more than 10 years ago. It is conceivable that factors that previously influenced social transmission no longer have the same impact. A third possibility is that the influence of arousal on the sharing of information is much smaller than Berger’s study suggests. Although our studies had much larger sample sizes than Berger’s, our studies were powered only to detect medium effects.

A final possibility is that the original study was a false positive. As noted earlier, Berger (2011) had only 40 participants complete the study. Small sample sizes have been commonplace in psychological research, but neglecting power has been known for some time to contribute to erroneous conclusions (Sedlmeier & Gigerenzer, 1989). When combined, our studies had over six times as many participants as Berger’s (2011) study. Therefore, on balance, we favor conclusions drawn from our studies over the much smaller original study.

Although our studies do not support the notion that physiological arousal increases sharing of information, it is important to point out that the arousal was incidental to the articles the participants read. Therefore, what our studies suggest is that physiological arousal that is unrelated to the information to be shared does not appear to influence whether people share the information on social media. Berger’s (2011) study was designed to isolate the influence of arousal on sharing of information, which is why arousal was manipulated independently of the article the participants read. Our studies (and Berger’s) can speak only to whether physiological arousal, by itself, can increase social transmission. It is quite possible that arousal induced by the information might increase the likelihood that people would share that information. Similarly, our studies do not address whether emotional arousal can influence social transmission of information. There are, for example, numerous studies finding that when information elicits emotional responses, people are more likely to share that information (e.g., Berger & Milkman, 2012; Brady et al., 2017; Hasell & Weeks, 2016; Lottridge & Bentley, 2018; Peters et al., 2009; Stieglitz & Dang-Xuan, 2013).

One implication of our studies is that the influence of emotions on social transmission is not mediated by physiological arousal, as suggested by Berger (2011). People might be more willing to share emotion-inducing information as an attempt to build social connections (Rimé, 2009) or as an emotion-regulation strategy (Berger, 2014). Regardless of the specific mechanism, the conclusion from our studies is clear; in contrast to Berger’s (2011) study, we did not find evidence that incidental arousal increases social transmission of information.

Supplemental Material

sj-docx-1-pss-10.1177_09567976241257255 – Supplemental material for Does Physiological Arousal Increase Social Transmission of Information? Two Replications of Berger (2011)

Supplemental material, sj-docx-1-pss-10.1177_09567976241257255 for Does Physiological Arousal Increase Social Transmission of Information? Two Replications of Berger (2011) by Skyler Prowten, Emily Walker, Brian London, Elizabeth Pearce, Angela Napoli, Bailey Chenevert, Christian Clevenger and Andrew R. Smith in Psychological Science

Footnotes

Appendix

Table A1.

Correlations Between Dependent Variables in Studies 1 and 2

	Articles shared	Arousal	Heart rate	Relaxed	Positive affect	Negative affect
Articles shared	—	.23**	.18*	−.10	.04	.01
Arousal	.14	—	.47***	−.20*	.56***	.10
Heart rate	−.01	.64***	—	−.39***	.26***	−.14
Relaxed	.08	−.19*	−.36***	—	−.03	−.26**
Positive affect	.32***	.66***	.38***	−.05	—	.23**
Negative affect	.11	.03	.22*	−.20*	.16	—

Note: Values below the diagonal are from Study 1, and values above the diagonal are from Study 2.

p < .05. **p < .01. ***p < .001.

Transparency

Action Editor: Mark Brandt

Editor: Patricia J. Bauer

Author Contributions

Skyler Prowten: Conceptualization; Data curation; Formal analysis; Methodology; Project administration; Writing – original draft; Writing – review & editing.

Emily Walker: Conceptualization; Data curation; Investigation; Methodology; Project administration; Writing – original draft; Writing – review & editing.

Brian London: Conceptualization; Data curation; Formal analysis; Methodology; Project administration; Writing – original draft; Writing – review & editing.

Elizabeth Pearce: Conceptualization; Data curation; Formal analysis; Methodology; Project administration; Writing – original draft; Writing – review & editing.

Angela Napoli: Conceptualization; Data curation; Formal analysis; Methodology; Project administration; Writing – original draft; Writing – review & editing.

Bailey Chenevert: Conceptualization; Data curation; Formal analysis; Methodology; Project administration; Writing – original draft; Writing – review & editing.

Christian Clevenger: Conceptualization; Data curation; Formal analysis; Methodology; Project administration; Writing – original draft; Writing – review & editing.

Andrew R. Smith: Conceptualization; Data curation; Formal analysis; Methodology; Project administration; Writing – original draft; Writing – review & editing.

ORCID iD

Andrew R. Smith

Supplemental Material

Additional supporting information can be found at

Notes

References

Alexandrov

Lilly

Babakus

(2013). The effects of social and self-motives on the intentions to share positive and negative word of mouth. Journal of the Academy of Marketing Science, 41(5), 531–546. https://doi.org/10.1007/s11747-012-0323-4

Al-Rawi

(2017). Viral news on social media. Digital Journalism, 7(1), 63–79. https://doi.org/10.1080/21670811.2017.1387062

Baptista

J. P.

Gradim

(2020). Understanding fake news consumption: A review. Social Sciences, 9(10), Article 185. https://doi.org/10.3390/socsci9100185

Berger

(2011). Arousal increases social transmission of information. Psychological Science, 22(7), 891–893. https://doi.org/10.1177/0956797611413294

Berger

(2014). Word of mouth and interpersonal communication: A review and directions for future research. Journal of Consumer Psychology, 24(4), 586–607. https://doi.org/10.1016/j.jcps.2014.05.002

Berger

Milkman

K. L.

(2010). Social transmission, emotion, and the virality of online content. Wharton Research Paper, 106, 1–52. http://thearf-org-unified-admin.s3.amazonaws.com/MSI/2020/06/MSI_Report_10-114.pdf

Berger

Milkman

K. L.

(2012). What makes online content viral? Journal of Marketing Research, 49(2), 192–205. https://doi.org/10.1509/jmr.10.0353

Brady

W. J.

Wills

J. A.

Jost

J. T.

Tucker

J. A.

Van Bavel

J. J.

(2017). Emotion shapes the diffusion of moralized content in social networks. Proceedings of the National Academy of Sciences, USA, 114(28), 7313–7318. https://doi.org/10.1073/pnas.1618923114

Camerer

C. F.

Dreber

Holzmeister

T. H.

Huber

Johannesson

Kirchler

Nave

Nosek

B. A.

Pfeiffer

Altmejd

Buttrick

Chan

Chen

Forsell

Gampa

Heikensten

Hummer

Imai

. . . Wu

(2018). Evaluating the replicability of social science experiments in Nature and Science between 2010 and 2015. Nature Human Behaviour, 2(9), 637–644. https://doi.org/10.1038/s41562-018-0399-z

10.

Ebersole

C. R.

Atherton

O. E.

Belanger

A. L.

Skulborstad

H. M.

Allen

J. M.

Banks

J. B.

Baranski

Bernstein

M. J.

Bonfiglio

D. B. V.

Boucher

Brown

E. R.

Budiman

N. I.

Cairo

A. H.

Capaldi

C. A.

Chartier

C. R.

Chung

J. M.

Cicero

D. C.

Coleman

J. A.

Conway

J. G.

. . . Nosek

B. A.

(2016). Many Labs 3: Evaluating participant pool quality across the academic semester via replication. Journal of Experimental Social Psychology, 67, 68–82. https://doi.org/10.1016/j.jesp.2015.10.012

11.

Goh

J. X.

Hall

J. A.

Rosenthal

(2016). Mini meta-analysis of your own studies: Some arguments on why and a primer on how. Social and Personality Psychology Compass, 10(10), 535–549. https://doi.org/10.1111/spc3.12267

12.

Hasell

Weeks

B. E.

(2016). Partisan provocation: The role of partisan news use and emotional responses in political information sharing in social media. Human Communication Research, 42(4), 641–661. https://doi.org/10.1111/hcre.12092

13.

Heath

(1996). Do people prefer to pass along good news or bad news? Valence and relevance of news as a predictor of transmission propensity. Organizational Behavior and Human Decision Processes, 68, 79–94.

14.

Kümpel

A. S.

Karnowski

Keyling

(2015). News sharing in social media: A review of current research on news sharing users, content, and networks. Social Media + Society, 1(2). https://doi.org/10.1177/2056305115610141

15.

LeBel

E. P.

Vanpaemel

Cheung

Campbell

(2019). A brief guide to evaluate replications. Meta-Psychology, 3–12. https://doi.org/10.15626/MP.2018.843

16.

Ling

(2020). Confirmation bias in the era of mobile news consumption: The social and psychological dimensions. Digital Journalism, 8(5), 596–604. https://doi.org/10.1080/21670811.2020.1766987

17.

Lottridge

Bentley

F. R.

(2018). Let’s hate together: How people share news in messaging, social, and public networks. Proceedings of the 2018 CHI Conference on Human Factors in Computing Systems, 1–13. https://doi.org/10.1145/3173574.3173634

18.

Mosleh

Pennycook

Rand

D. G.

(2020). Self-reported willingness to share political news articles in online surveys correlates with actual sharing on Twitter. PLOS ONE, 15(2), Article e0228882. https://doi.org/10.1371/journal.pone.0228882

19.

Open Science Collaboration. (2015). Estimating the reproducibility of psychological science. Science, 349(6251), Article aac4716. https://doi.org/10.1126/science.aac4716

20.

Pennycook

Cannon

T. D.

Rand

D. G.

(2018). Prior exposure increases perceived accuracy of fake news. Journal of Experimental Psychology: General, 147(12), 1865–1880. https://doi.org/10/gfkddm

21.

Pennycook

Rand

D. G.

(2018). Lazy, not biased: Susceptibility to partisan fake news is better explained by lack of reasoning than by motivated reasoning. Cognition, 188, 39–50. https://doi.org/10.1016/j.cognition.2018.06.011

22.

Pentina

Tarafdar

(2014). From “information” to “knowing”: Exploring the role of social media in contemporary news consumption. Computers in Human Behavior, 35, 211–223. https://doi.org/10.1016/j.chb.2014.02.045

23.

Pereira

Harris

Van Bavel

J. J.

(2023). Identity concerns drive belief: The impact of partisan identity on the belief and dissemination of true and false news. Group Processes & Intergroup Relations, 26(1), 24–47. https://doi.org/10.1177/13684302211030004

24.

Peters

Kashima

Clark

(2009). Talking about others: Emotionality and the dissemination of social information. European Journal of Social Psychology, 39(2), 207–222. https://doi.org/10.1002/ejsp.523

25.

Rimé

(2009). Emotion elicits the social sharing of emotion: Theory and empirical review. Emotion Review, 1(1), 60–85. https://doi.org/10.1177/1754073908097189

26.

Sedlmeier

Gigerenzer

(1989). Do studies on statistical power have an effect on the power of studies? Psychological Bulletin, 105, 309–316.

27.

Simmons

J. P.

Nelson

L. D.

Simonsohn

. (2011). False-positive psychology: Undisclosed flexibility in data collection and analysis allows presenting anything as significant. Psychological Science, 22(11), 1359–1366. https://doi.org/10.1177/0956797611417632

28.

Simons

D. J.

(2014). The value of direct replication. Perspectives on Psychological Science, 9(1), 76–80. https://doi.org/10.1177/1745691613514755

29.

Simonsohn

(2015). Small telescopes: Detectability and the evaluation of replication results. Psychological Science, 26(5), 559–569. https://doi.org/10.1177/0956797614567341

30.

Stieglitz

Dang-Xuan

(2013). Emotions and information diffusion in social media—sentiment of microblogs and sharing behavior. Journal of Management Information Systems, 29(4), 217–248. https://doi.org/10.2753/MIS0742-1222290408

31.

Watson

Clark

L. A.

Tellegen

(1988). Development and validation of brief measures of positive and negative affect: The PANAS scales. Journal of Personality and Social Psychology, 54(6), 1063–1070. https://doi.org/10.1037/0022-3514.54.6.1063

32.

Wegner

D. M.

Giuliano

(1980). Arousal-induced attention to self. Journal of Personality and Social Psychology, 38(5), 719–726. https://doi.org/10.1037/0022-3514.38.5.719

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.

0.00 MB

1.06 MB