Online Rumor Transmission Among Younger and Older Adults

Social Media Use and Psychological Responses to Rumor

Social media not only promotes efficient dissemination of information but also facilitates the spread of rumors (Starbird, Maddock, Orand, Achterman, & Mason, 2014; Takayasu et al., 2015). On one hand, it can provide a platform for users to verify information collaboratively (Aula, 2013; Hermida, 2013), reduce uncertainty, and make appropriate decisions during a crisis (Ki & Nekmat, 2014; Ma, 2008; Muralidharan, Rasmussen, Patterson, & Shin, 2011), all of which can help detect rumors. Thus, the higher the frequency of social media use, the less likely one will fall for rumors. On the other hand, social media also allows all users, including those without authority, to generate or share information, which makes it difficult for users to access credible information from multiple sources (Agichtein, Castillo, Donato, Gionis, & Mishne, 2008; K. S. Kim, Yoo-Lee, & Sin, 2012; Lindsay, 2011; Savolainen, 2011). One study indicated that frequent use of online information sources increased exposure to rumors and strengthened people’s beliefs in rumors during the U.S. presidential election (Garrett, 2011). Moreover, given the inherently participatory and information-sharing nature of social media, users find it easy to share information immediately even before verifying it. A study found that users tend to support rather than deny a rumor on Twitter before it is verified (Zubiaga et al., 2016). The results from a national survey showed that only 25.2% of Chinese netizens (Internet users) prefer to validate information before they share it with others (China Internet Network Information Center, 2016), indicating that many active WeChat users likely transmit rumors without even being aware of it. Therefore, it is also plausible that using WeChat more frequently may increase the probability of participating in generating and transmitting rumors.

If social media use is associated with the intention to transmit rumors, what psychological processes are responsible for this effect? Research on rumor transmission showed that psychological factors, such as uncertainty, belief, importance, lack of control, and anxiety, have been identified as individual-level motivations for transmitting rumors (DiFonzo & Bordia, 2007), among which anxiety plays a crucial role (Anthony, 1973; Jaeger, Anthony, & Rosnow, 1980; Walker & Beckerle, 1987). Higher anxiety was associated with a greater likelihood of transmitting rumors. A meta-analysis of rumor studies also confirmed that anxiety has a large effect (r = .48) on rumor transmission (DiFonzo & Bordia, 2002). Although only a few studies have directly focused on anxiety and online rumor transmission and anxiety has not been reported to be the most important predictor, K. H. Kwon and Rao (2017) found that anxiety is still a significant and important predictor of rumor transmission. Thus, we considered that the effect of WeChat use on the intention to transmit rumors could be explained by different levels of rumor anxiety in individuals. People with varying levels of WeChat use may have different levels of rumor anxiety when they encounter social media rumors, leading to different possibilities with respect to transmission of rumors. Therefore, we focused on the relationships among WeChat use, rumor anxiety, and the intention to transmit rumors, with a focus on the potential mediation role of rumor anxiety.

Age-Related Challenges

Previous studies identified age-related differences in the motivation and modes of social media use (Vroman, Arthanat, & Lysack, 2015; Zickuhr & Madden, 2012), suggesting that the associations among WeChat use, rumor anxiety, and intention to transmit rumors might also vary with age. Compared to other social media platforms where younger people tend to be overrepresented, WeChat users are more widely distributed in terms of age. Even older adults, known as digital immigrants (Prensky, 2001), account for 50 million of the WeChat users (WeChat, 2017). WeChat exposed Chinese older adults to a considerable amount of controversial information and viewpoints for the first time (Guo, 2017). It is important to recognize that the challenges of using social media might be more severe among older people. For example, as people age, they may experience more difficulty in managing information overload if they experience age-related cognitive decline (O’Sullivan et al., 2001; Salthouse, 2009; Salthouse, Atkinson, & Berish, 2003). Previous studies have reported that older people are more disturbed by news overload (Holton & Chyi, 2012). Moreover, they are reported to be less efficient at searching for information online (Czaja, Sharit, Ownby, Roth, & Nair, 2001; Sharit, Hernández, Czaja, & Pirolli, 2008), and thus, they tend to rely on other people (e.g., families and friends) when confronted with uncertain information. Research has indicated that compared to their younger counterparts, older people are more likely to spread rumors (Lai & Wong, 2002). Thus, the effect of WeChat use and rumor transmission might differ by age. Furthermore, older people are more likely to experience various problems related to physical health, independence, and safety (Rush, Murphy, & Kozak, 2012), and they might be easily triggered by rumors that evoke anxiety (Rosnow, 1988). Given the lack of direct research into the possible association between social media use and psychological responses to rumors, particularly considering the potential moderating role of age, we proposed the following research questions. Do the associations among WeChat use, rumor anxiety, and rumor transmission intention vary by age? Particularly, does age moderate the relationship among rumor anxiety, WeChat use, and rumor transmission intention, and if so, how?

The conceptual model shown in Figure 1 based on a moderated mediation model illustrates the research questions.

OPEN IN VIEWER

Figure 1.

Conceptual model.

Method

Results

Descriptive and Preliminary Analyses

With respect to the frequency of WeChat use, 41.6% (n = 3,893) of the participants reported spending more than 1 hour per day, 52.3% (n = 4,902) reported using WeChat 1 hour or less per day, and only 6.1% (n = 573) of the participants reported that they do not use WeChat.

The correlation between WeChat use and rumor reactions is presented in Table 1. Participants who were more active on WeChat had a greater intention to transmit rumors (r = .039, p < .001). Meanwhile, rumor anxiety was strongly and positively correlated with rumor transmission intention (r = .755, p < .01), indicating that anxiety is an important psychological factor of rumor transmission. However, in the simple correlation analysis, the association between WeChat use and rumor anxiety was not significant (r = .008, p = .459). Moreover, when participants encountered rumors, female participants had a higher intention to transmit rumors than did male participants (M = 3.22 vs. M = 3.18, t = 3.13, p < .01).

OPEN IN VIEWER

Table 1.

Correlations, Means, and Standard Deviations of Variables (N = 9,368).

Variables	M	SD	1	2	3	4	5	6	7	8	9
1. WeChat use	4.28	1.66	1
2. Rumor anxiety	3.04	0.65	.008	1
3. Rumor transmission intention	3.20	0.71	.039**	.755**	1
4. Age	36.48	13.26	−.215	.020*	.006	1
5. Rumor exposure	2.56	0.70	.060**	.410**	.413**	−.020	1
6. Gender	0.50	0.50	−.029**	−.037**	−.032**	.015	−.008	1
7. Income	2.59	1.16	.048**	.046**	.065**	.193**	.102**	.151**	1
8. Education	2.55	1.02	.119**	.019	.026*	−.334*	.103**	.007	.225**	1
9. Subjective social status	5.55	1.64	.019	−.022*	.026*	.014	.065**	.007	.148**	.059**	1

*

p < .05. **p < .01 (two-tails).

Analyses of Moderated Mediation Model

As the relationship between WeChat use and rumor anxiety was not significant, we are unable to report the simple mediation effect of rumor anxiety according to the traditional stepwise method. However, we propose that the mediating effect might differ by further considering the moderating effect of age. Thus, we directly tested the moderated mediation model using Model 8 of Hayes’ PROCESS macro (Table 2; Hayes, 2013). The frequency of WeChat use was entered as the independent variable (X). Age was entered as the moderating variable (W), rumor anxiety was entered as the mediator variable (M), and rumor transmission intention was entered as the criterion variable (Y) in the model. Gender, education, income, subjective social status, and rumor exposure were included as covariates.

OPEN IN VIEWER

Table 2.

Results of the Moderated Mediation Analysis (N = 9,368).

	Consequent (Y)
	Rumor anxiety		Rumor transmission intention
Antecedent	b	SE	b	SE
WeChat use	−0.051***	0.011	−0.015+	0.008
Rumor anxiety	—	—	0.768***	0.008
Age	−0.004***	0.001	−0.003***	0.001
WeChat use × age	0.001***	0.000	0.001**	0.000
Rumor exposure	0.379***	0.009	0.118***	0.007
Gender	−0.046***	0.012	−0.010	0.010
Income	0.008	0.006	0.012*	0.004
Education	−0.012	0.007	−0.009	0.005
Subjective social status	−0.020***	0.004	0.013***	0.003
Constant	2.387***	0.062	0.564***	0.052
Model summary	R2 = .17		R2 = .59
	F(8, 9,359) = 248.03, p < .001		F(9, 9,358) = 1,465.44, p < .001

†

p < .1. *p < .05. **p < .01. ***p < .001 (two-tails).

As shown in Table 2, rumor anxiety was a significant factor for predicting rumor transmission intention—people with higher rumor anxiety were more likely to transmit rumors, b = 0.768, SE = 0.008, (95% confidence interval [CI]: 0.7526-0.7839), p < .01. Moreover, age significantly moderated the associations between WeChat use and perceived rumor anxiety (b = 0.0012, SE = 0.0003 [95% CI: 0.0007-0.0017], p < .001) and between WeChat use and rumor transmission intention (b = 0.0007, SE = 0.0002 [95% CI: 0.0003-0.0011], p < .01), indicating that the relationships between WeChat use and rumor anxiety and between WeChat use and rumor transmission intention varied by age. To identify the age range for which the effect of WeChat use on rumor anxiety was significant and to determine the direction of the effect, we further probed this interaction using the Johnson-Neyman technique (Spiller, Fitzsimons, Lynch Jr, & McClelland, 2013). As shown in Figure 2, the conditional effect of WeChat use on rumor anxiety was negatively significant when the respondent’s age was below 32 years (less than 32.2311), but the effect was positively significant when the respondent’s age was above 41 years (more than 40.2332). For respondents aged between 33 and 40 years, the conditional effect of WeChat use on rumor anxiety was not statistically significant.

OPEN IN VIEWER

Figure 2.

The conditional effect of WeChat use on rumor anxiety as function of age.

To address the moderated mediation effect, a formal test of linear moderated mediation was performed with 10,000 bootstrap samples. The index of moderated mediation for rumor transmission intention was significant (index = 0.0009, SE = 0.0002, 95% bias-corrected bootstrap CI [0.0005-0.0013]), indicating that the indirect effects of WeChat use on the intention to transmit rumors through rumor anxiety were moderated by age. In addition, estimates of the conditional direct and indirect associations of WeChat use with rumor transmission intention at the ages of 23 (M_age − 1SD), 36 (M_age), and 50 (M_age + 1SD) years are shown in Table 3. For the direct effect, the conditional association of WeChat use with rumor transmission intention was not significant for participants under age 23. However, as the participant’s age increased, this direct association became significantly positive. Among participants above age 50, those who used WeChat more frequently had higher rumor transmission intention. For indirect effects, a pattern similar to the direct effect was observed. The conditional indirect effect of WeChat use on rumor anxiety was negatively significant for young people (effect = −0.017, bootstrap SE = 0.004, bootstrap CI = [−0.025 to −0.009]) but positively significant for older people (effect = 0.008, bootstrap SE = 0.004, bootstrap CI = [0.001-0.015]).

OPEN IN VIEWER

Table 3.

Conditional Direct and Indirect Effects of WeChat Use on Rumor Transmission Intention.

Direct effect				Indirect effect through perceived rumors anxiety
Age	Effect	SE	p	Effect	Bootstrap SE	Bootstrap CI
23.221	0.001	0.004	.840	−0.017	0.004	[−0.025 to −0.009]
36.478	0.010	0.0029	.001	−0.005	0.003	[−0.011 to 0.001]
49.735	0.018	0.004	.000	0.008	0.004	[0.001 to 0.015]

Note. CI = confidence intervals.

Discussion

With social media playing an increasingly important role in our daily lives, the frequency of using social media has also become a significant factor for predicting many important social outcomes, such as social engagement (Y. Kim et al., 2016), well-being (Wen et al., 2016), and so on. Based on data from a national survey in China, this study investigated the relationship between the use of WeChat and rumor transmission intention and identified the underlying moderating mechanism of age in the relationships among WeChat use, rumor anxiety, and rumor transmission intention.

Our results show that WeChat use was associated with the intention to transmit rumors, and age significantly moderated this association and the mediation effect of rumor anxiety. Specifically, more frequent use of WeChat reduced young people’s intention to transmit rumors; however, it increased this intention among older people. For the younger users, using WeChat more frequently was related to lower anxiety, which had a negative indirect effect on the intention to transmit rumors. Furthermore, WeChat use had no direct effect on young people’s intention to transmit rumors. However, as the age of users increased, WeChat use gradually had not only a direct positive effect on rumor transmission intention but also an indirect positive effect through rumor anxiety. We suggest that more frequent use of social media may help young people make better use of the Internet to collect information, analyze information, and discover the truth, thus effectively reducing their rumor anxiety and the possibility of transmitting rumors. In contrast, for older people, because of the threat of cognitive ability decline and a deficit in media literacy (Guess et al., 2019), more frequent use of social media cannot help them reduce rumor anxiety when facing information overload (O’Sullivan et al., 2001). Instead, it may enhance their anxiety and ultimately increase the possibility of transmitting rumors. Our finding is consistent with those of previous studies, which reported that older people are highly vulnerable to believing rumors on social media compared to younger people (Grinberg et al., 2019; Guess et al., 2019). According to Guess et al. (2019), older people aged above 65 years shared fake news articles on Facebook 5.69 times more than the youngest group, and age was shown to be the most robust demographic predictor, even after controlling for other variables (e.g., education and ideology).

Our study has theoretical implications for understanding the role of the digital literacy divide and providing new evidence for the theoretical framework of why people believe online rumors. First, our findings are helpful to understand the differences in the decision-making process for people of different age groups when exposed to harmful information, and it is suggested that simply bridging the divide in social media use rather than addressing the underlying digital literacy divide may further increase these differences. The higher risk of rumor transmission among older people could be related to the lack of digital media literacy and decline in cognitive ability (Guess et al., 2019). Previous studies of the social media divide mainly focused on the divide in social media use (Robinson et al., 2015). However, our findings suggest that the higher the frequency of social media use among older people, the higher the risk of rumor transmission. This means that we should focus on not only the level of media use but also the underlying literacy divide. Second, by examining the associations among social media use, rumor anxiety, and rumor transmission intention in people of different ages, we not only verified the phenomenon of older people being at a higher risk for the social media divide but also explored its mechanism. Most of the existing studies (Grinberg et al., 2019) have only found that older people are at a higher risk of believing rumors, but the underlying mechanism is still poorly understood. From the perspective of social media use, our findings clarify the role of rumor anxiety in this process and suggest that older people who use social media more frequently are more likely to become anxious and transmit rumors when exposed to online rumors. Third, our study adds to the research on the factors influencing rumor transmission by introducing social media use as a new dimension to depict the characteristics of Internet users, thereby enriching the existing theoretical framework of describing individual characteristics from cognitive and emotional perspectives. Future studies should consider synthesizing and comparing different individual characteristics, even situational factors and other factors, to provide more comprehensive evidence for why people trust and share online rumors.

This study also has practical implications for governing online rumors, providing guidance for targeting interventions to the most vulnerable people and allocating intervention resources. Online rumors have been described as a “digital wildfire” and identified as one of the numerous threats to governance worldwide (DiFonzo, 2013). Practical guidance generally indicated that rumor governance needs the efforts of not only the government and organizations but also the public and ordinary users (Webb et al., 2016). We propose that online rumor governance would greatly benefit from understanding who is more likely to share rumors on social media and why they do so. According to our results, older people could be considered a vulnerable group, and the risk of sharing rumors is higher among older people who use social media more frequently. Therefore, rumor governance and interventions, such as rumor debunking, rumor prevention, and rumor inoculation (Van der Linden, Maibach, Cook, Leiserowitz, & Lewandowsky, 2017), should be more targeted at older users, especially those who use social media more frequently. Our findings also suggest that the higher the use of social media among older people, the more likely they are to transmit rumors, which was also related to their elevated rumor anxiety to some extent. Therefore, practitioners should aim to decrease rumor anxiety in older people to attenuate the risk of sharing rumors. It has been found that social support can help older adults to reduce stress and improve their psychological well-being (Antonucci, 1990). Thus, maintaining contact with loved ones through social media might, in fact, help reduce loneliness and improve the physical and psychological health of older people (Chopik, 2016), which may also prevent them from believing online rumors.

This study has several limitations that open avenues for future research. First, because we relied on self-reports, our results might have been affected by the bias of favorable self-assessment stemming from participants’ self-enhancement motives. Therefore, future studies should analyze users’ rumor decisions based on behavioral data obtained from social media platforms. Second, our findings were based on a cross-sectional survey, which precludes inferring a causal relationship between WeChat use and rumor transmission. In the future, longitudinal studies or experiments should be conducted to gain a better understanding of this relationship. Finally, although WeChat is the most popular social media platform in China, it has the characteristics of a friendship network to some extent, which makes it different from other more open social media platforms, such as Twitter. To elucidate the relationship between social media use and rumor transmission, future studies should investigate and compare the effects of other social media platforms.

Conclusion

This study demonstrated that WeChat use could have different effects on individuals’ psychological responses to social media rumors based on age. For younger people, increased WeChat use was associated with lower rumor anxiety and a reduced probability of sharing rumors. However, for older individuals, using WeChat more frequently promoted the intention to transmit rumors as a result of increased anxiety. This study extends the findings of previous research on the challenges of social media use among older adults. When older people adopt a new technology, they may have more difficulties coping with social media rumors, giving rise to the question of how to better integrate technology into the lives of older adults.