e-Health Campaigns for Promoting Influenza Vaccination: Examining Effectiveness of Fear Appeal Messages from Different Sources

Abstract

Background:

In recent years, influenza has become a severe disease and pandemic threat. There are more than 290,000 to 650,000 influenza-related deaths globally each year. Influenza vaccination is the best way to prevent influenza and potentially serious influenza-related complications. The current study aims to examine the effectiveness of fear-induced health campaigns on social media in promoting influenza vaccination with the focus on different sources.

Methods:

A 2 × 3 × 2 (visible source × receiver source × technological source) factorial online experiment was designed to investigate the effectiveness of fear appeal messages offered by different sources on social media. A total of 534 college students were recruited to participate in the experiment.

Results:

Individuals who receive messages from a verified visible source have greater intention to perform flu vaccination and seek flu-related information than those who acquire messages from an unverified one. Besides, visible source, receiver source, and technological source interact to affect flu-related information seeking.

Conclusions:

In addition to the message itself, different levels of message sources on social media should be considered for e-health campaign design, especially visible sources.

Introduction

Influenza viruses are highly contagious, which can cause an estimated 1 billion cases every year.¹ Being in constant exposure in proximate spaces, such as shared restrooms, enables influenza viruses to transmit quickly and widely making college students vulnerable to influenza.^2
–4 Once infected, they would experience illness of 8 days or more, which could lead to school absenteeism and impaired academic performance.²

Taking vaccination annually is the most effective way to prevent influenza.⁵ However, in the U.S. universities, only 8%–39% of the students have taken influenza vaccination.² In China, the flu vaccination rate is 2.4% and the vaccination rate among college students also hovers at low levels.^6
–8 Therefore, encouraging college students to get vaccinated against influenza is of great importance.

To solve this problem, an e-health campaign through social media may be an ideal option. First, there are an increasing number of college students who have used social media to access a large amount of health information.⁹ Moreover, social media can offer cost-effective access to reach a broad audience¹⁰ Besides, it should be noted that except the content, the source of the message on social media, which could be divided into visible sources, receiver sources, and technological sources,¹¹ is also important because it can influence people's judgment of information credibility and thus determine the effectiveness of health promotion.^12,13

Therefore, the current study intends to examine the effectiveness of influenza fear-induced health campaigns from different sources on social media in promoting flu vaccination and flu-related information seeking among college students.

Literature Review

Fear Appeal

Fear appeal message refers to “a persuasive communication attempting to arouse fear in order to promote precautionary motivation and self-protective action.”¹⁴ Fear-induced message models have been widely used in numerous studies.^15
–17 Given that prior theories focus on the cognitive elements of fear appeals, Witte emphasized fear's motivational component and introduced the extended parallel process model (EPPM), which makes it the most influential theoretical model to explain the way fear appeals are proceed.¹⁸ In EPPM, three main components exist: (1) fear, (2) threat, and (3) efficacy. The components of severity, susceptibility, response efficacy, and self-efficacy along with two appraisals—(1) danger control and (2) fear control, convincingly predict persuasive message acceptance and adaptive behaviors.^18,19 More importantly, the high threat–high efficacy combined messages had the highest persuasive effect compared with other messages.²⁰

The effectiveness of fear-induced messages on health campaigns has been documented emprically.^20

–24 Also, in the field of influenza research, fear appeal plays a decisive role of persuading recommended actions such as health information seeking and disease prevention,²⁵ and people such as college students have been given scholarly attention.²⁶

Fear Appeal in E-Health Campaign

There were six key overarching benefits of social media in health campaigns, including increased users' interactions with others and a wider range of accessibility for people to reach more available, shared, and tailored information.²⁷

Social media, as a channel to contact with extensive users quickly, deliver all information online with no aid of any additional media. A recent study has indicated that there is no significant difference between the persuasiveness of fear appeal messages between social media and traditional media. Thus, social media have been widely used for health campaigns.²¹

As social media information has become increasingly abundant and diverse, excepting the message itself, individuals are also able to see the source of message. It has become an important factor that has long been considered when valuing the credibility of information given by media.¹⁵ Some studies have suggested that source could determine the effects of media messages in changing users' behavior intentions.¹¹ Various sources can yield different media effects.¹³ A study has found that the high-credible sources on social media rather than low-credible sources could improve their attention to and tend to retweet the message.²⁸ Besides, compared with lay persons, an expert source, such as the Centers for Disease Control and Prevention, can play a significant role on debunking health misinformation in social media, which effectively reduces the misperception of users.²⁹ Therefore, understanding different message sources is fundamental for designing e-health campaigns.

Information Sources on Social Media

Social media provides a platform, which allows the generation and diffusion of user-generated content.³⁰ Compared with traditional media, information source on social media is more diverse.³¹ Some scholars indicated that information sources in social media can be divided into three different categories: visible sources, technological sources, and receiver sources.¹³

Visible source is the source seen by the receiver initially delivering the message, which means people or organizations that originally send and present content are considered a visible source. Receiver source refers to audiences themselves who could actively choose their own information for consumption. Besides, technological source refers to the media or channels of content presentation.¹³

Specifically, in social media, the account presenting original message is the visible source and the message receiver simultaneously becomes a receiver source if he or she retweets, comments, or likes the message. Meanwhile, the social media platforms' sociotechnical affordances, which allow users to set the messages' openness, make the medium itself the technological source. Some studies have suggested that people perceive different levels of credibility toward various account types on social media, so that various visible and receiver sources could exert different effectiveness.³² Extant studies have found that people usually rely on a heuristic model to process social media messages. Account verification status on social media could provide important heuristic cues for credibility evaluation. Verified or not which represents the users' authoritativeness can influence individuals' perceived credibility of message.^33
–35 Lin and Spence³⁶ examined that when viewing information about food safety, messages posted by a lay user were perceived as lower levels of credibility by Twitter users compared with messages posted by an expert, which represent high authority. In addition, the information openness, which is provided by technical support of social media, also poses an impact on persuasiveness. The different publicness, namely private and public, is deemed to be a heuristic cue for users to evaluate information credibility.³⁷ Users can either post a message to the public, which means the content is open to every user; or exclusively send it to specific user or group, which means the content is private to people. Liu and Wei³⁸ found that private help-seeking messages on social media can motivate individuals to provide support, while public messages might downplay their effort to help. In other words, messages sent by a private technological source are more persuasive than a public one.

Furthermore, in the information transmission, the sender creates or gathers a message, then a technological medium/channel transmits it to the audience, which makes a completed communication chain.¹³ However, in social media, the audience may not receive the message directly from the sender, but from another receiver, because each receiver is empowered to retweet the original message created by the sender. Therefore, the current study intends to explore the interactive effect among three different levels of sources.

In addition, scholars have indicated that health campaigns may not affect actual behavior directly, but stimulate the recipients to engage in information, such as paying attention to and discussing about relevant content, which in turn lead to actual behavior.³⁹ Thus, for information engagement, fear appeal messages on social media could effectively motivate message viewers to click on health-related information, in that Lazarus indicated clicking is a typical behavior of appraisal, recipients value the relevance and importance of messages carefully through this process.^40,41 Chen and his colleagues also found that threat and efficacy on social media stimulated information engagement, such as liking, commenting, and sharing. Thus, besides actual behaviors, information engagement could be another behavioral outcome when receiving fear appeal messages on social media.⁴²

Based on the above considerations, we posit the following three hypotheses and research question:

H1: People who receive messages from a verified visible source have greater intention to perform (1) flu vaccination and (2) flu-related information seeking than those who acquire messages from an unverified one.

H2: People who receive messages from a verified receiver source have greater intention to perform (1) flu vaccination and (2) flu-related information seeking than those who acquire messages from an unverified one.

H3: People who receive messages privately have greater intention to perform (1) flu vaccination and (2) flu-related information seeking than those who acquire messages publicly.

RQ1: To what extent do the visible, receiver and technological sources interact to affect the effectiveness of fear appeal message on (1) flu vaccination and (2) flu-related information seeking?

Methods

Experimental Design

This study was a 2 (visible source: verified vs. unverified) × 3 (receiver source: nonreceiver vs. verified vs. unverified) × 2 (technological source: public vs. private) between-subjects factorial experiment based on Weibo, which is an equivalent of twitter in China launched in 2009 and has become one of the most popular social media sites in China.⁴³ We conducted a convenience sampling, which is also used in many other online experiment and survey studies,^44,45 through the internet to recruit Chinese college students. The online experiment link was distributed via Chinese social media such as WeChat and Weibo, and e-mail. Ethics approval of the current study was obtained. All participants involved in this study gave their informed consent. A total of 534 subjects participated in this study, which is much larger than a minimum sample size. A priori power analysis for multivariate analysis of variance (MANOVA) with a medium effect size (f of 0.25), a desired statistical power level of 0.95, and an α level of 0.05, calculated through G-Power version 3, indicated a minimum sample size of 171.

Procedure

An online questionnaire containing 12 different experimental conditions was administered to participants. After completing an online informed consent, the online questionnaire system randomly showed 1 of the 12 pictures that contain influenza-related fear appeal information to the participants. Then, participants were asked to evaluate their intention to get flu vaccination and seek flu-related information after the stimulus.

Stimulus

The stimulus of this study is 12 pictures created within the template of Weibo. Each picture contains a fear appeal message about influenza on Weibo, which is released by a visible source and retweeted by a receiver source in one of two publicness of technological sources—post wall (public) or chat frame (private). To ensure whether the messages could successfully affect participants' levels of threat and efficacy, the materials with or without threat and efficacy were allocated to two groups of people and a pilot test was conducted. Sixty-four college students (24 males and 40 females) participated in this pilot test and the t-test analysis results showed a significant difference of threat [t(64) = 3.04, p < 0.01] and efficacy [t(64) = 3.00, p < 0.01] between these two groups, which means the manipulation material was effective.

Data Analysis

Basic descriptive analysis, t-test, and MANOVA through software SPSS 23.0 were conducted to test stimulus used in this study as well as examine the hypotheses and research question we proposed.

Results

The descriptive analysis showed that about 58% (n = 310) of participants were female, and the average age of respondents was 21.4 [standard deviation (SD) = 2.19] years. All of participants are from cities all over China (Eastern China 62%, Central China 20%, Western China 15%, and Northeast China 3%).

The MANOVA results revealed a significant main effect for the visible source of the stimuli on both behavioral intentions, which means that participants who had received message from a verified visible source had higher intentions to perform vaccination [mean (M) = 5.08, SD = 1.41] and flu-related information seeking (M = 5.20, SD = 1.41), while those who had received message from an unverified visible source had lower intentions to perform vaccination [M = 4.70, SD = 1.54, F(1, 522) = 8.61, p < 0.01, partial η ² = 0.02] and flu-related information seeking [M = 4.91, SD = 1.58, F(1, 522) = 4.80, p < 0.05, partial η ² = 0.01], which are consistent with H1. However, the main effect for the receiver and technological source of the stimuli on the flu vaccination and flu-related information seeking behavioral intention were neither significant. Thus, H2 and H3 were not supported (Tables 1 –3 ).

Table 1.

Descriptive Statistics for Intention to Perform Flu Vaccination and Flu-Related Information Seeking

VISIBLE	RECEIVER	TECHNOLOGICAL	N	MEAN		SD
VISIBLE	RECEIVER	TECHNOLOGICAL	N	VACCINATION	INFORMATIONSEEKING	VACCINATION	INFORMATIONSEEKING
Verified	Nonreceiver	Public	44	5.19	5.55	1.59	1.37
		Private	44	5.11	4.70	1.17	1.52
		Total	88	5.15	5.13	1.39	1.50
	Verified	Public	45	5.08	5.27	1.26	1.21
		Private	44	5.22	5.38	1.62	1.35
		Total	89	5.15	5.32	1.44	1.27
	Unverified	Public	42	4.76	4.98	1.48	1.65
		Private	43	5.12	5.29	1.31	1.23
		Total	85	4.95	5.14	1.40	1.45
	Total	Public	131	5.02	5.27	1.45	1.42
		Private	131	5.15	5.12	1.37	1.39
		Total	262	5.08	5.20	1.41	1.41
Unverified	Nonreceiver	Public	45	4.49	4.70	1.96	1.74
		Private	48	4.71	5.03	1.54	1.48
		Total	93	4.60	4.87	1.75	1.61
	Verified	Public	42	4.85	4.98	1.23	1.51
		Private	42	5.09	5.00	1.57	1.77
		Total	84	4.97	4.99	1.41	1.63
	Unverified	Public	48	4.56	4.77	1.48	1.49
		Private	47	4.53	4.98	1.38	1.52
		Total	95	4.55	4.87	1.42	1.50
	Total	Public	135	4.63	4.81	1.59	1.58
		Private	137	4.76	5.00	1.50	1.57
		Total	272	4.70	4.91	1.54	1.58
Total	Nonreceiver	Public	89	4.84	5.12	1.81	1.62
		Private	92	4.90	4.88	1.38	1.50
		Total	181	4.87	5.00	1.60	1.56
	Verified	Public	87	4.97	5.13	1.25	1.36
		Private	86	5.16	5.19	1.59	1.57
		Total	173	5.06	5.16	1.42	1.47
	Unverified	Public	90	4.66	4.87	1.48	1.56
		Private	90	4.81	5.13	1.37	1.39
		Total	180	4.74	5.00	1.42	1.48
	Total	Public	266	4.82	5.04	1.53	1.52
		Private	268	4.95	5.06	1.45	1.49
		Total	534	4.89	5.05	1.49	1.50

SD, standard deviation.

Table 2.

Visible × Receiver × Technological Factorial Analysis of Variance for Intention to Perform Flu Vaccination

SOURCE	df	F	η²	p
Visible	1.00	8.61	0.02	0.00^**
Receiver	2.00	1.99	0.01	0.14
Technological	1.00	1.22	0.00	0.27
Visible × receiver	2.00	0.69	0.00	0.50
Visible × technological	1.00	0.00	0.00	1.00
Receiver × technological	2.00	0.08	0.00	0.93
Visible × receiver × technological	2.00	0.64	0.00	0.53
Error	522

p < 0.05, ^** p < 0.01, and ^*** p < 0.001.

Table 3.

Visible × Receiver × Technological Factorial Analysis of Variance for Intention to Perform Flu-Related Information Seeking

SOURCE	df	F	η²	p
Visible	1.00	4.80	0.01	0.03^*
Receiver	2.00	0.63	0.00	0.53
Technological	1.00	0.03	0.00	0.86
Visible × receiver	2.00	0.03	0.00	0.97
Visible × technological	1.00	1.57	0.00	0.21
Receiver × technological	2.00	1.36	0.01	0.26
Visible × receiver × technological	2.00	2.70	0.01	0.07
Error	522

p < 0.05, ^** p < 0.01, and ^*** p < 0.001.

Besides, the results revealed that the visible × receiver interaction effect [F(2, 522) = 0.69, p = 0.50, partial η ² = 0.003; F(2, 522) = 0.03, p = 0.97, partial η ² = 0.00], visible × technological interaction effect [F(1, 522) = 0.00, p = 0.998, partial η ² = 0.00; F(1, 522) = 1.57, p = 0.21, partial η ² = 0.003], and receiver × technological interaction effect [F(2, 522) = 0.08, p = 0.93, partial η ² = 0.00; F(2, 522) = 1.36, p = 0.26, partial η ² = 0.01] were not significant on people's intentions to perform flu vaccination and flu-related information seeking.

Furthermore, the interaction effect among three sources on individuals' intention to get vaccination was not significant [F(2, 522) = 0.64, p = 0.53, partial η ² = 0.002]. However, when it came to the intention to perform flu-related information seeking, those three sources' interaction effect was marginally significant [F(2, 522) = 2.70, p = 0.068, partial η ² = 0.01] (Tables 2–3). Specifically, in the public condition, if a verified account (visible source) posted a message, individuals who obtained the message directly had a higher intention to seek flu-related information than those who obtained it through a receiver source. Besides, once being retweeted by a verified receiver source, message from an unverified visible source could exert the greatest persuasiveness. In the private condition, message directly sent by a verified account generated the lowest intention of people to seek flu-related information. However, if the message was retweeted by other accounts whether it is verified or not, individuals' behavior intention had dramatically increased. Meanwhile, if the visible source is an unverified account, persuasiveness was not much different among the three receiver sources. The interactions of three sources are depicted in Figure 1a and b.

Fig 1.

(a) Three-way interaction between visible source and receiver source on intention to seek flu-related information in the public technological source condition. (b) Three-way interaction between visible source and receiver source on intention to seek flu-related information in the private technological source condition.

Discussion

The current study examined the effectiveness of influenza-related fear appeal messages from different sources on social media in promoting vaccination and flu-related information seeking.

The results revealed that there was a significant main effect for visible source on the effectiveness of fear appeal messages in increasing people's intention to perform flu vaccination and information seeking, consistent with the extant study.⁴⁶ According to the heuristic-systematic model of information processing, unless people are motivated to determine accurate judgments and have sufficient cognitive competence, they usually tend to use least effort to acquire an attitude toward information.⁴⁷ It is reported that many college students do not believe that they are at risk of contracting influenza and do not realize the effectiveness of vaccination.² Thus, they might be less motivated and tend to process information in a heuristic way, using noncontent cues (such as communicator credibility) to judge the message.⁴³ Our research demonstrated that in the context of influenza vaccination, messages from verified accounts have better effects in persuasion, which reconfirms that verification of an account indicates a certain degree of authoritativeness and people are more likely to trust messages sent by this kind of accounts.^33,35

However, the effectiveness of fear appeal message in promoting people's vaccination behavior and information seeking behavior was not affected by receiver source and technological source. One explanation could be that in the current study, we used a single-time retweeted message with neither retweet number of the original content nor extra comments, which may lead to insufficient cues for people to process the message sent by receiver sources. As for technological source, since the accounts used to send message in the current study were either official accounts or fictitious grassroots accounts, the psychological distance between these accounts and the message receiver is relatively far, which may cause out-group bias and thus weaken the persuasiveness of privately sent message. Therefore, there is no much difference between private or public message about influenza in this study.

In addition, on social media, various sources are a complicated synthesis.¹³ Previous studies have discussed the influence of different sources, respectively,^30,35,48 but no study investigated their interaction effects yet. One of the major contributions offered by the current study is that we examined the interactive effects of fear appeal messages composed by different sources. The results indicated that the visible, receiver, and technological sources interacted to affect flu-related information seeking. According to the results, in a public condition, individuals who obtained the message directly from a verified account had a higher intention to seek flu-related information than those who obtained the message retweeted by a receiver source. However, if the message was published in an unverified account, retweeted by a verified account could increase the persuasiveness. In this regard, the message was endorsed by the high credible. This reaffirmed the crucial role of credibility of visible source. Thus, the message retweeted by a verified receiver source is more persuasive than an unverified receiver source. These findings are consistent with prior studies,^15,35 which indicated that authoritativeness and credibility of the source centered within the health information. In contrast, in a private condition, who obtained the message directly from a verified account had the least intention to seek flu-related information, meaning the private publicness will somehow decrease the persuasiveness of verified visible source. It is plausible that recipients may suspect it as propaganda and even refuse to pay attention to information because the verified account normally does not send private message to individuals directly. However, when the messages posted by an unverified account were retweeted by three kinds of receiver sources, the persuasive effects were relatively average, since the original source is low in credibility. This encourages health professionals to attach more importance to the original message sender and means of message delivering in the future health campaigns.

Implications and Limitations

The contributions of our study are twofold. About practical contributions, the current study sheds light on the promotive status of high-credible sources such as health professionals in online health campaigns. Specifically, our study suggests that health information should directly release by verified accounts in public way to increase individuals' intention to perform health behaviors. Besides, messages posted by an unverified account (i.e., grassroots) publicly may find the chance of being retweeted by a verified account to increase its persuasiveness. Factors of verified account status and open publicness should be included in e-health persuasion. As for theoretical contributions, our study extends the fear appeal theories by exploring the effectiveness of different sources on social media. Furthermore, we contribute to the field of media effects by providing a novel research perspective of information sources on social media, that is, as an important factor of credibility evaluation, the sources on social media should be examined, respectively, and interactively.

This study also has some limitations. First, our research mainly explored how information sources on social media affect the effectiveness of fear appeal message, while cues of information credibility on the internet are various. Prior research has indicated that users' commenting, liking, and other cues on social media could affect perceived trustworthiness and credibility.^12,44 In the future, more attention should be paid to the effects of other cues on social media. Second, we merely used the verification status of accounts to indicate the relative expertise and trustworthiness of visible source and receiver source. Future study should explore the role of other traits of sources, such as social distance,⁴⁹ and self-disclosed personal profile information.⁵⁰ Third, in our study, only one social media platform was investigated. However, the classification of technological source may be different among various platforms. Future study could replicate this study on other social media platforms. Finally, we barely examined the persuasiveness of fear appeals on social media; future studies could investigate the effectiveness of other persuasive messages in e-health campaigns.

Footnotes

Acknowledgments

This study is based on the authors' conference article “Social Media Health Campaigns for Promoting Influenza Vaccination: Examining Effectiveness of Fear Appeal Messages from Different Sources” presented at the 2020 Association for Education in Journalism and Mass Communication. We appreciate the dedication of the three anonymous reviewers to this article.

Disclosure Statement

No competing financial interests exist.

Funding Information

This work was supported by the National Social Science Fund [19CXW033].

References

World Health Organization (WHO). Global influenza strategy 2019–2030. Available at https://www.who.int/influenza/Global_Influenza_Strategy_2019_2030_Summary_English.pdf?ua=1 (last accessed June 15, 2020 ).

National Foundation for Infectious Diseases (NFID). Addressing the challenges of influenza vaccination on US college campuses: A report by the National Foundation for Infectious Diseases. Available at https://www.nfid.org/wp-content/uploads/2019/08/college-flu-summit-report-1.pdf (last accessed June 15, 2020 ).

Nichol

, D'Heilly

, Ehlinger

. Influenza vaccination among college and university students: Impact on influenzalike illness, health care use, and impaired school performance. Arch Pediatr Adolesc Med, 2008; 162:1113–1118.

Xiao

, Ma

, Wang

, Liu

, Li

. Prevalence and risk factors of influenza like illness among 2112 college students in Guangzhou, 2017–2018 epidemic season. Chin J Public Health, 2019; 35:829–832.

World Health Organization (WHO). Vaccines against influenza WHO position paper November. 2012. Available at https://www.who.int/immunization/position_papers/PP_influenza_november2012_summary.pdf?ua=1 (last accessed June 15, 2020 ).

Fan

, Cong

, Wang

, Bao

, Wang

, Feng

, et al. Influenza vaccination rate and its association with chronic diseases in China: Results of a national cross-sectional study. Vaccine, 2020; 38:2503–2511.

Tuohetamu

, Pang

, Nuer

. The knowledge, attitudes and practices on influenza among medical college students in Northwest China. Hum Vaccin Immunother, 2017; 13:1688–1692.

, Yang

, Li

, Ma

, Zhang

, Wang

. The coverage rate and obstructive factors of influenza vaccine inoculation among residents aged above 18 years in Beijing from 2007 to 2010. Chin J Prev Med, 2011; 45:1077–1081.

, Kim

. College students' use of social media for health in the USA and Korea. Inform Res, 2014; 19:n4.

10.

Korda

, Itani

. Harnessing social media for health promotion and behavior change. Health Promot Pract, 2013; 14:15–23.

11.

Sundar

, Nass

. Conceptualizing sources in online news. J Commun, 2001; 51:52–72.

12.

, Shyam Sundar

. Effects of online health sources on credibility and behavioral intentions. Commun Res, 2010; 37:105–132.

13.

Borah

, Xiao

. The importance of “Likes”: The interplay of message framing, source, and social endorsement on credibility perceptions of health information on facebook. J Health Commun, 2018; 23:399–411.

14.

Ruiter

, Abraham

, Kok

. Scary warnings and rational precautions: A review of the psychology of fear appeals. Psychol health, 2001; 16:613–630.

15.

Hovland

, Weiss

. The influence of source credibility on communication effectiveness. Public Opin Q, 1951; 15:635–650.

16.

Leventhal

. Findings and theory in the study of fear communications. In: Berkowitz

, ed. Advances in experimental social psychology. Amsterdam, The Netherlands: Elsevier, 1970:119–186.

17.

Rogers

. A protection motivation theory of fear appeals and attitude change. J Psychol, 1975; 91:93–114.

18.

Witte

. Putting the fear back into fear appeals: The extended parallel process model. Commun Monogr, 1992; 59:329–349.

19.

Witte

. Fear control and danger control: A test of the extended parallel process model (EPPM). Commun Monogr, 1994; 61:113–134.

20.

Witte

, Allen

. A meta-analysis of fear appeals: Implications for effective public health campaigns. Health Educ Behav, 2000; 27:591–615.

21.

Chen

, Yang

. Using EPPM to evaluate the effectiveness of fear appeal messages across different media outlets to increase the intention of breast self-examination among Chinese women. Health Commun, 2019; 34:1369–1376.

22.

Dillard

, Plotnick

, Godbold

, et al. The multiple affective outcomes of AIDS PSAs: Fear appeals do more than scare people. Commun Res, 1996; 23:44–72.

23.

Kotowski

, Smith

, Johnstone

, Pritt

. Using the EPPM to create and evaluate the effectiveness of brochures to reduce the risk for noise-induced hearing loss in college students. Noise Health, 2011; 13:261–271.

24.

Chen

, Guo

, Shi

. Social support seeking on social media among Chinese gay men living with HIV/AIDS: The role of perceived threat. Telemed J EHealth, 2019; 25:655–659.

25.

Dedeoglu

, Ventura

. Consumer responses to swine flu (H1N1) threat and fear arousing communications: The case of turkey. In: Campbell

, ed. The customer is not always right? Marketing orientations in a dynamic business world. Cham, Switzerland: Springer, 2017:249–258.

26.

Roberto

, Mongeau

, Liu

, Hashi

. “Fear the flu, not the flu shot”: A test of the Extended Parallel Process Model. J Health Commun, 2019; 24:1–8.

27.

Moorhead

, Hazlett

, Harrison

, et al. A new dimension of health care: Systematic review of the uses, benefits, and limitations of social media for health communication. J Med Internet Res, 2013; 15:e85.

28.

Liu

, Liu

, Li

. Determinants of information retweeting in microblogging. Internet Res, 2012; 22:443–466.

29.

Vraga

, Bode

. Using expert sources to correct health misinformation in social media. Sci Commun, 2017; 39:621–645.

30.

Kaplan

, Haenlein

. Users of the world, unite! The challenges and opportunities of social media. Bus Horiz, 2010; 53:59–68.

31.

Van Der Heide

, Lim

Y-S

. On the conditional cueing of credibility heuristics. Commun Res, 2015; 43:672–693.

32.

Nekmat

, Gower

, Zhou

, Metzger

. Connective-collective action on social media: Moderated mediation of cognitive elaboration and perceived source credibility on personalness of source. Commun Res, 2015; 46:62–87.

33.

Zhang

, Peng

, Zhang

, et al. Content or context: Which matters more in information processing on microblogging sites. Comput Hum Behav, 2014; 31:242–249.

34.

Metzger

, Flanagin

. Credibility and trust of information in online environments: The use of cognitive heuristics. J Pragmat, 2013; 59:210–220.

35.

Edgerly

, Vraga

. The blue check of credibility: Does account verification matter when evaluating news on Twitter?. Cyberpsychol Behav Soc Netw, 2019; 22:283–287.

36.

Lin

, Spence

. Identity on social networks as a cue: Identity, retweets, and credibility. Commun Stud, 2018; 69:461–482.

37.

Bazarova

. Public intimacy: Disclosure interpretation and social judgments on Facebook. J Commun, 2012; 62:815–832.

38.

Liu

, Wei

. Modeling social support on social media: Effect of publicness and the underlying mechanisms. Comput Hum Behav, 2018; 87:263–275.

39.

Wakefield

, Loken

, Hornik

. Use of mass media campaigns to change health behaviour. Lancet, 2010; 376:1261–1271.

40.

Zhang

, Zhou

. Clicking health risk messages on social media: Moderated mediation paths through perceived threat, perceived efficacy, and fear arousal. Health Commun, 2019; 34:1359–1368.

41.

Lazarus

. Psychological stress and the coping process. New York: McGraw-Hill, 1966:31–40.

42.

Chen

, Yang

, Fu

, Liu

, Yuan

. Using the Extended Parallel Process Model to examine the nature and impact of breast cancer prevention information on Mobile-Based social media: Content analysis. JMIR mHealth uHealth, 2019; 7:e13987.

43.

Miles

, Zhang

. China turns to tweeting: Exploring the problematic use of tweeting in China. Int J Bus Soc Sci, 2012; 3:91–94.

44.

Ershad-Sarabi

, Rahmati

, Mangolian-Shahrbabaki

. The effect of education on healthcare workers' knowledge of oro-dental health in a group of Iranian rural children. J Oral Health Oral Epidemiol, 2020; 9:32–37.

45.

Yang

, Chen

, Feng

. Risk perception of food safety issue on social media. Chin J Commun, 2016; 9:124–138.

46.

McGlone

, Bell

, Zaitchik

, McGlynn

, 3rd. Don't let the flu catch you: Agency assignment in printed educational materials about the H1N1 influenza virus. J Health Commun, 2013; 18:740–756.

47.

Chaiken

The heuristic—systematic model in its broader context. In: Shelly C, Yaacov T, eds. Dual-process theories in social psychology. New York, NY: Guilford Press, 1999:73.

48.

Gierth

, Bromme

. Attacking science on social media: How user comments affect perceived trustworthiness and credibility. Public Underst Sci, 2020; 29:230–247.

49.

Sohn

, Choi

. Social embeddedness of persuasion: Effects of cognitive social structures on information credibility assessment and sharing in social media. Int J Advert, 2019; 38:824–844.

50.

Park

, Xiang

, Josiam

, Kim

. Personal profile information as cues of credibility in online travel reviews. Anatolia, 2014; 25:13–23.