当期荐读 2022年第3期 | 理解公众常态化防疫期的新冠疫苗接种意愿——从媒体接触与交互、健康与社会规范信念的视角

图 | Internet

郭路生　周瑶瑶　周金凤

南昌大学公共政策与管理学院，南昌，330031

摘要 / Abstract

随着新冠肺炎疫情在我国的成功控制，公众的感知风险和接种意愿下降，给免疫屏障的快速建立带来了挑战。本研究调查公众心理、社会和媒体使用等因素对疫苗接种意愿的影响。于2021年4月对全国范围内的1030名成年人进行网络调查，调查内容包括感知风险、感知利益、感知障碍、感知社会规范、媒体使用和交互频率、人口统计学特征等。采用SmartPLS软件进行PLS-SEM分析。结果表明：疫苗接种意愿总体较高(85.6%)，但有所下降，37.9%的人不愿立即接种；感知社会规范是最大的直接预测因子，然后是感知利益、感知障碍和感知风险；媒体使用间接影响接种意愿，且正式媒体的作用大于社交媒体；媒体依赖是调节变量，如果存在媒介依赖，正式媒体使用的影响显著而社交媒体不显著，如果不存在媒体依赖，情况则恰恰相反。结果揭示了媒体接触与交互、健康和社会规范信念如何影响接种意愿，为传染性健康危机传播和疫苗接种干预提供了见解。

关键词

新冠肺炎　疫苗接种意愿　社会规范　健康信念　媒介接触　信息交互　媒体依赖

Abstract

With the successful control of COVID-19 in China, the public’s perceived risk and vaccination intention have declined, posing a challenge to quickly establish an immune barrier. This study aimed to investigate the psychological, social, and media use factors that influences the COVID-19 vaccination intentions during the well-contained phase. A nationwide survey was conducted online among Chinese adults (N=1030)from April 1, 2021 to April 20, 2021 on the public’s vaccination intentions and possible influencing factors including perceived risks, benefits, barriers, and social norms, frequency of media usage and interaction, and demographic characteristics. The data were analyzed by PLS-SEM using SmartPLS software. The results show that vaccination intentions are generally high (85.6%), but there is a drop, and 37.9% of individuals will not be vaccinated immediately. Perceived social norms are the largest direct predictor, followed by perceived benefits, perceived barriers, and perceived risks. Media usage indirectly influences vaccination intention, and the effect of formal media is greater than that of social media. Media dependence is a moderator. If there is media dependence, the effect of formal media usage is significant while the effect of social media is not. If there is no media dependence, the opposite is true. These findings reveal how media usage and interaction, health and social norm beliefs predict vaccination intentions, providing insights into infectious health crisis communication and vaccination interventions.

Keywords

COVID-19; Vaccination intention; Social norm; Health beliefs; Media usage; Information interaction; Media dependence; China

向上滑动阅览

01

 Introduction

The COVID-19 pandemic has taken a huge toll on global health, economy, and society^[1-2]. As of April 11, 2021, the cumulative number of people infected with COVID-19 globally is approximately 136 million and the death toll is 2.93 million, with 480,000 new confirmed cases per day^[1]. Although COVID-19 has been well controlled in China, the risk of a re-epidemic remains. For example, there were some secondary epidemics around the 2021 Spring Festival in Hebei, Jilin, and other regions^[3].

Building herd immunity through COVID-19 vaccination is key to ending this pandemic^[4-5]. Herd immunity requires the vaccination rate to reach more than 70%^[6]. If minors are excluded, the vaccination rate of adults should be more than 90%^[7]. At present, a variety of COVID-19 vaccines have been successfully developed and distributed for use, but there is some hesitation due to concerns about the quality of these rapidly developed vaccines^[7-10], and China is no exception^[11]. According to an Ipsos survey, 27% of the respondents were unwilling or unsure if they would get vaccinated against COVID-19; 36% of these respondents were in France, the USA, and Spain^[12]. With the elimination of COVID-19 in Wuhan on April 28, 2020, China has reached the so-called phase of regular epidemic prevention and control, “the well-contained phase during which the pandemic was under control, work, study and life of the general population were resumed and economic and social order were restored”^[11]. The perceived risk of the Chinese public has decreased and the intention to be vaccinated has declined^[11-12]. Specifically, the intention to get vaccinated immediately has decreased from 58.3% to 23%^[11], which has severely challenged the rapid establishment of an immune barrier in China. For more effective health communication and interventions to improve vaccination intention, there is an urgent need to investigate influencing factors among the Chinese public during the well-contained phase.

Previous studies have focused on the impact of health beliefs and socio-demographic characteristics on vaccination intentions, mostly based on the health belief model(HBM).Wang et al.^[13] showed that being male, married, high perceived risk and perceived efficacy would increase vaccination intentions against COVID-19 in China. Lin et al.^[14] and Zampetakis and Melas.^[15] confirmed that health beliefs could predict vaccination intentions. Those studies provided the basis for this study. However, COVID-19 is not a common disease but a highly infectious disease that poses a great threat to society. It has facilitated an emergent shared social identity and the formation of social norms to constrain common epidemic prevention behaviors. In other words, getting vaccinated against COVID-19 depends not only on one's attitudes but also on social norms. However, previous studies have ignored the impact of social norms.

Some studies have investigated the impact of media usage on vaccination intentions. It was generally accepted that more frequent use of traditional media could promote vaccination intentions^[16-17], but the role of social media usage varies by regions. Allington et al.^[17] have shown that more frequent social media usage in the US and UK leads to more conspiracy beliefs and more negative vaccination intentions; But Mahmood’s^[18] work in Pakistan and Du’s^[19] in China suggest it has a positive effect. Hence more research evidences are needed. In addition, social media has media attributes and social attributes. Users can obtain epidemic information through its media attributes (conceptualized as social media usage) and interact with others through its social attributes (conceptualized as social media interaction). Therefore, social media usage and social media interaction may have different effects on vaccination intentions, but previous studies did not distinguish.

In summary, previous studies have investigated the effects of socio-demographic characteristics^[11,13,20], health beliefs^[14-15,21], and media usage^[17-19] on vaccination intentions but few studies have considered the effects of perceived social norms and different attributes of social media. There is a lack of research on the mechanism behind the combined influence of these factors on vaccination intentions. To fill these gaps, this study used protection motivation theory to investigate the influence of health beliefs, perceived social norms, and media usage on the vaccination intention of the Chinese public during the well-contained phase, providing new insights into appropriate health communications and interventions. The specific research questions are as follows.

RQ1: Are perceived social norms and health beliefs (perceived risks, perceived benefits, and perceived barriers) predictors of COVID-19 vaccination intention in China during the well-contained phase? Which factor is the most important?

RQ2: Do different types of media usage (formal media vs. social media) affect the public's health beliefs and perceived social norms? Are the effects different?

RQ3: Do media usage and interaction indirectly influence vaccination intentions through health beliefs and perceived social norms?

RQ4: Does media dependence moderates the relationship between media usage, media interaction and vaccination intentions?

02

Literature Review and Hypotheses Development

2.1  Health belief model and protection motivation theory

Health belief model(HBM)^[22] is a prominent theory in health behavior research, which was proposed by Hochbaum in 1952 and perfected by Backer later. HBM holds that health behaviors are determined by their health beliefs, including perceived risks, perceived benefits, perceived barriers, and self-efficacy^[22-23]. Perceived risk refers to the subjective evaluation of the health threat (e.g., COVID-19), including perceived severity and perceived susceptibility. Perceived severity refers to “one’s feelings about the seriousness of a threatening event”, whereas perceived susceptibility refers to “beliefs about the probability of personally experiencing the threat”^[18]. Perceived benefit refers to the belief that the health prevention behavior (e.g., COVID- 19 vaccination) will reduce the risk or seriousness of the disease threat^[15]. Perceived barrier refers to the beliefs that health prevention behavior is restricted due to difficulties related to psychosocial, physical, or financial factors^[15]. Concerns about adverse reactions and side effects after vaccination are the main barriers to vaccination^[24]. Self-efficacy is defined as beliefs about one’s ability and confidence to carry out recommended health behaviors^[18]. An individual will engage in health behaviors only when the perceived risks are serious, the perceived benefits of health behaviors outweigh the perceived barriers, and he or she has enough ability and confidence to do so. HBM has been widely used to explain and predict health behaviors, such as breast cancer screening^[25], influenza vaccination^[26] and HPV vaccination^[27].

The protection motivation theory is the development of HBM, which was proposed by Rogers in 1975^[28].The biggest improvement is the addition of external factor, information sources, to the health belief model. It holds that information sources influence protective behaviors through the mediating of health beliefs^[28-29], which provides a theoretical framework for the prediction and intervention of health protective behaviors as well as this study.

2.2  Health beliefs and vaccination intentions

Vaccination is an individual's health protective action against COVID-19. According to HBM^[23] and protection motivation theory^[28], the vaccination intentions against COVID-19 are determined by health beliefs, including perceived risks, perceived benefits, perceived barriers and self-efficacy. Self-efficacy was not considered in this study due to China's free COVID-19 vaccine policy. Studies have shown that the higher the perceived risk of COVID-19, the higher the perceived benefits and the lower the negative effects of COVID-19 vaccine, the higher the individual's vaccination intentions^{[14-15,23,30]}. The hypotheses are as follows:

H1:Perceived risks are positively related to vaccination intentions.

H2: Perceived benefits are positively related to vaccination intentions.

H3:Perceived barriers are negatively related to vaccination intentions.

2.3 Perceived social norms and vaccination intentions

　　COVID-19 is not a common disease but highly contagious and remains a great threat to society. It has led to an emergent shared social identity^[31]. This sense of common destiny can lead people to greater awareness of the impact of the epidemic on others and society and to collective action^[32], especially in countries with collectivist cultures. “The implicit rules that constrain social behaviors and underlie the felt social pressure to enact a given behavior within a group” are called perceived social norms, including descriptive social norms and injunctive social norms^[33]. Descriptive social norms refer to “the extent to which the individual perceives that other people are engaging in the preventive behavior”^[33], whereas injunctive social norms refer to “the extent to which the individual perceives that close others would approve or disapprove the behavior”^[33].

　　The theory of planned behavior holds that perceived social norms are an important predictor of individual behaviors (such as vaccination)^[34-35]. Gouin et al.^[33] and Mat el al.^[36] confirmed that perceived social norms had a significant impact on preventive behavior during the COVID-19 pandemic. Therefore, we suggest that perceived social norms may be a predictor of COVID-19 vaccination.

　　H4：Perceived social norms are positively related to vaccination intention

2.4 Media usage, interaction and health, social norm beliefs

2.4.1　Media usage and interaction

Media usage refers to how often the public uses the media to get news and updates about COVID-19^[16], which can be divided into formal media usage and social media usage according to the type of media. Formal media are those whose contents are journalist-generated and checked by editors or institutions, such as traditional media, official media and large commercial media, whereas social media conveys user-generated content without editors’ check, such as Wechat, Weibo and Online BBS. Social media also enables users to engage in interactions with others about the COVID-19 epidemic and vaccines, such as online communication, comments and likes. The frequency with which the public engages in these interactions about COVID-19 through social media is conceptualized as social media interaction.

2.4.2　Media usage, interaction and health beliefs

As people learn about risk events and risk response mainly from various media rather than direct experience, media usage is the focus of health communication research^[37]. According to the protection motivation theory, the content and frequency of information exposed to individuals will affect their perceived risks and perceived efficacy of response measures^[38]. With more frequent exposure to correct epidemic risk information, vaccine knowledge will lead to higher perceived risks of COVID-19 and perceived efficacy of protective behavior and vaccines, thus promoting vaccination intentions^[16,18,37], while more frequent exposure to misinformation will lead to more conspiracy beliefs and more negative vaccination intentions^[39-40].

Information in formal media is of high quality and in line with government vaccination recommendations. Its positive effects on individuals' perceived risks, perceived efficacy of vaccination, and vaccination intentions are generally recognized^{[16,18,37,41]}, while the role of social media is controversial as it is associated with misinformation^[39,42]. Some studies in western countries have shown that more frequent social media use leads to more conspiracy beliefs and more negative vaccination intentions^[16-17,40], while studies in eastern countries have shown that social media use is positive for perceived risks^[18,43], perceived efficacy^[18] and protective behavior^[18,43-44]. A recent study by Du et al.^[19] in China also showed that more informal media usage, though weaker than formal media, could still increase perceived risks and perceived benefits, and reduce perceived barriers, thus promoting vaccination intentions. Therefore, we believe that both formal media and social media have positive effects on health beliefs in China. The hypotheses are as follows:

H5: Formal media usage is positively related to perceived risks.

H6: (a) Social media usage and (b) interaction are positively related to perceived risk.

H7: Formal media usage is positively related to perceived benefits.

H8: (a) Social media usage and (b) interaction are positively related to perceived benefits.

H9: Formal media usage is negatively related to perceived barriers.

H10: (a) Social media usage and (b) interaction are negatively related to perceived barriers.

2.4.3　Media usage, interaction and perceived social norms

Like health beliefs, media usage and interaction also influence the formation of social norm beliefs. Information on COVID-19 in the media can be divided into three categories: risk information, knowledge and advices on epidemic prevention, and information on social norms and social interactions. The more information the public receives about COVID-19 risks and social norms through the media, the more they are aware of the threats COVID-19 poses to their society and the needs and responsibilities to engage in collective protection and vaccination, i.e., the higher the perceived social norms. The hypotheses are as follows:

H11: Formal media usage is positively related to perceived social norms.

H12: (a) Social media usage and (b) interaction are positively related to perceived social norms.

2.4.4　Media dependence

Media dependence is an individual's media bias in accessing information about COVID-19. If COVID-19 information comes from formal media much more than social media, it is referred to as formal-media-dependence; conversely, it is referred to as social-media-dependence; if the proportion is similar, there is non-media-dependence. Studies have shown that young people prefer social media, while older people prefer traditional media^[16]. Media dependence reflects individual group characteristics as a moderator of media usage and beliefs about health and social norms, and vaccination intentions in this study.

The quality of user-generated content in social media is often uneven, while the quality of content in formal media, which is strictly controlled by editors, is higher and more in line with national epidemic prevention policies. Previous studies have shown that social media usage has a weaker positive effect on vaccination intentions than formal media usage, and is even negative for people in West who depend on social media^[16]. Therefore, a certain level of formal media exposure is necessary to develop appropriate risk perceptions and positive vaccination attitudes.

For those who depend on social media, their vaccination intentions are more likely to be affected by inconsistent vaccination attitudes of other social media users, and their social media usage has a weaker positive impact on vaccination intentions than those who do not depend on social media^[16]. Coombs^[45] argues that the frame of news media plays an arbitrating role when individuals are affected by multiple frames. Therefore, formal media usage in the social-media-dependent group is more critical in promoting vaccination intentions than in the non-media-dependent group. For those who depend on formal media, they are mainly affected by the strong effect of formal media, while the weak effect of social media is even more insignificant. In other words, compared with the non-media-dependent group, the effect of formal media is stronger, while the effect of social media is weaker in the formal-media-dependent group. Based on the above analysis, hypotheses are as follows:

H13: Compared with non-media-dependent group, formal media usage has a stronger impact on vaccination intentions (a) in the social-media-dependent group or (b) in the formal-media-dependent group;

H14: Compared with non-media-dependent group, social media usage has a weaker impact on vaccination intentions (a) in the social-media-dependent group or (b) in the formal-media-dependent group.

H15: Compared with non-media-dependent group, social media interaction has a weaker impact on vaccination intentions (a) in the social-media-dependent group or (b) in the formal-media-dependent group.

2.4.5　The conceptual model of this study

This study investigates how media usage and interactions among different media-dependent populations influence vaccination intentions through health and social norm beliefs. Based on the above analysis and the protection motivation theory, a three-layer model of "information-cognition-behavior" is constructed, as shown in Figure 1. The information layer includes three constructs: formal media usage, social media usage, and social media interaction. The cognitive layer includes four constructs: health beliefs (perceived risks, perceived benefits, and perceived barriers) and perceived social norms. The behavioral layer is the vaccination intention. Media dependence is a moderator variable and social demographic characteristics, such as gender, age, occupation, and education level, used as control variables.

Figure 1  Conceptual model of this study

03

Materials and methods

3.1　Survey and Samples

1,030 samples were collected from a nationwide survey conducted among Chinese adults from April 1 to April 20, 2021. Respondents were recruited from a group of over one million Chinese residents on the Tencent questionnaire platform (wj.qq.com). A nationwide non-probabilistic quota sampling design was adopted to create a nationally representative sample in terms of region, gender, and age. Questionnaires were sent to the respondents via Wechat or QQ. A total of 2,635 volunteers received the questionnaire and 1,210 questionnaires were returned, with a response rate of 45%. We excluded questionnaires that had a short answer time (less than 100 seconds) and consistent answers, with a total of 1,030 valid samples.

The characteristics of the samples are shown in Table 1. The socio-demographic characteristics of the samples, such as age, gender, residence, risk level and education level, are consistent with the situation in China.

Table 1  Sample characteristics(N=1030)

3.2　Measurements

　　This study investigated the Chinese public’s intention to get vaccinated against COVID-19 in a low-risk period and its influencing factors: health beliefs (including perceived risks, perceived benefits, and perceived barriers), perceived social norms, media usage and interactions, and socio-demographic characteristics.

3.2.1　Vaccination intention

　　Vaccination intentions are measured on a 7-level scale of three items: what is your willingness to be vaccinated voluntarily and what is your willingness to do so after being asked by the government, from 1 (very unwilling) to 7 (very willing) and how soon will you get vaccinated (7, within one month; 6, within three months; 5, within six months; 4, within one year; 3, within two years; 2, after two years; 1, never). For the individuals who chose "unwilling" (a score of 1-3 on a scale of 1-7) for voluntary vaccination, we asked why. The reasons are: 1)no local cases so no need; 2) doubt about the vaccine’s effectiveness; 3) worried about vaccine safety; 4) inconvenient vaccination; and 5) advocate for natural immunity.

3.2.2　Health beliefs and perceived social norms

　　According to HBM, health beliefs can be described as perceived risks, perceived benefits, and perceived barriers^[23,30] measured using a 7-level scale. Perceived risks are measured according to the scale proposed in Xi et al.^[46], and by asking participants about: 1) the possibility of them or their family being infected with COVID-19 (perceived personal susceptibility); 2) the possibility of another COVID-19 outbreak in their community (perceived social susceptibility); 3) the degree they are worried about COVID-19 infection (worry emotion); and 4) whether they often imagine the situation after infection (perceived severity). Perceived benefits are measured based on the scale of Gouin et al^[33]. They are detemined by the extent to which participants thought that COVID-19 vaccination can effectively protect 1) themselves or 2) others from the virus, and 3) blocking the spread of the epidemic in the community. Perceived barriers are measured using the VAX scale\[47\] by asking participants to what extent they 1) doubt the effectiveness of the vaccine, 2) are concerned about the safety of the vaccine, 3) believe vaccination is inconvenient, 4) believe it is a commercial practice, and 5) advocate for natural immunity.

　　Perceived social norms can be divided into descriptive social norms, injunctive social norms, and injunctive personal norms. These are measured by Gouin’s scale^[33] from 1 (strongly disagree) to 7 (strongly agree). To measure descriptive social norms, we asked respondents the extent to which they thought others were willing to be vaccinated. To measure injunctive social norms, we asked respondents to what extent they thought others would disapprove of them if they didn't get vaccinated. To assess injunctive personal norms, we asked respondents to what extent they felt that vaccination was an obligation.

3.2.3　Media usage and interaction

　　Media usage refers to how often the public uses the medium to obtain information about COVID-19. It is divided into formal media usage and social media usage, assessed using a 7-level scale ranging from 1 (almost never) to 7 (daily). The contents in formal media are journalist-generated and checked by editors or institutions, whereas the contents in social media are user-generated and unchecked by editors. Formal media usage is measured by asking users how often they use: 1) traditional media, such as newspapers and television; 2) the government’s official website, Weibo, and official accounts; 3) official network media,such as People's Daily Online and Xinhua net; and 4) Sina, Sohu and other business portals to obtain information about COVID-19. Social media usage is measured by asking respondents how often they use: 1) WeChat and Weibo, 2) WeChat Moments and Qzone, and 3) BBS, etc. to obtain COVID-19 information.

　　Social media enables users to participate in discussions about the COVID-19 epidemic and vaccines. Social media interaction is used to describe how often users participate in these discussions. We measured this by asking respondents how often they 1) shared, forwarded, and commented on WeChat moments, 2) discussed COVID in forums, or 3) participated in discussions in WeChat or QQ groups about COVID-19, ranging from 1 (almost never) to 7 (daily).

3.2.4　Social demographic characteristics

　　Socio-demographic factors include gender, age, education, occupation, current residence, and highest COVID-19 risk level of residence in the past four months. The city’s epidemic risk level was assigned by the Chinese Center for Disease Control.

3.3　Data Analysis

　　According to the suggestions of Sarstedt et al.^[49-50], SmartPLS 3.0 software was used to conduct partial least squares structural equation modeling (PLS-SEM) for the path model. First, the measurement model was evaluated. Then, the structural relationships of constructs were examined. The bootstrapping method was used to test the significance of the path coefficient and the mediating effect^[51], with a confidence of 95%.

04

Results

4.1　Summary of vaccination intentions

Currently, the Chinese public’s perceived risk is low (39.4 out of 100), and their intentions to be vaccinated is generally high (85.64 out of 100). Also, 85.6% of the samples expressed their willingness to be vaccinated if available (a score of 5-7 on a scale of 1-7), 8% expressed uncertainty (5 on a scale of 1-7) and 6.4% expressed unwillingness (a score of 1-3 on a scale of 1-7) (Figure 2). If the government requires vaccination, willingness increased to 88.2%. A total of 62.1% expressed the willingness to be vaccinated immediately (within one month) if available, and 37.9% expressed a desire to delay vaccination (at least one month later), indicating some vaccine hesitation. The main reasons for respondents, unwillingness to get vaccinated include concerns about the side effects of the vaccine, doubts about its effectiveness, and the absence of local cases that makes vaccination seem unnecessary.

Figure 2  Vaccination intention

4.2　Evaluation of measurement model

The measurement model was evaluated with reliability and validity analysis. Cronbach’s alpha, combined reliability (CR), and average variance extracted (AVE) of the measurement model are shown in Table 2.The factor loadings of all items exceeded 0.5 and were significant. All Cronbach’s alpha and CR values of each construct exceeded 0.7, and all AVE values exceeded 0.5. Therefore, this scale had high convergent validity and combination validity.

Table 2  The statistics of measurement model

The discriminant validity was tested by comparing the square root value of AVE with the correlation coefficient between the indicators, as shown in Table 3. It can be seen that correlation coefficients of all indicators were less than the square root of AVE, which indicates that the questionnaire had good discriminant validity. In general, the questionnaire in this paper had high reliability and validity.

Table 3 Construct correlations and square roots of AVE

4.3　Path model estimation

To answer RQ1 and RQ2, the PLS-SEM analysis results of the path model and the significance test results of the bootstrapping are shown in Figure 3. The results showed that perceived risks (β=0.059,P=0.02), perceived benefits (β=0.265, P<0.001), perceived barriers (β=-0.137, P<0.001), and perceived social norms (β=0.475, P<0.001) were significant predictors of vaccination intention, accounting for 51.1% of variance. Hence, H1, H2, H3 and H4 were supported. Among them, perceived social norms (0.475) were the largest predictor, followed by perceived benefits (0.265), perceived barriers (-0.137), and perceived risk (0.59).

Figure3. PLS-SEM estimation of the path model.

Notes:*p<0.05;**p<0.01;***p<0.001; 

dashed lines represent non-significant relationships

The results also showed that formal media usage significantly increased perceived risks (β=0.094,P=0.02), perceived benefits (β=0.372, P<0.001), and perceived social norms (β=0.328, P<0.001) and decreased perceived barriers (β=-0.164, P<0.001). Hence, H5, H7, H9 and H11 were supported. Social media interaction significantly increased users’ perceived risks (β=0.379, P<0.001), perceived barriers (β=0.330, P<0.001), and perceived social norms (β=0.082, P=0.03) and only had a weak effect on perceived benefits (β=0.057, P=0.09). Thus, H6b, H12b were supported, while H8b was weakly supported and H10b was the opposite. Social media usage significantly increased perceived social norms (β=0.125, P=0.007), but had no significant direct effect on the three constructs of health beliefs. Since social media usage was significantly correlated with social media interactions (β=0.590, P<0.001), it affected health beliefs and perceived social norms through the mediator of social media interaction. We compared the total effects to better compare the effects of formal and social media usage on health beliefs and perceived social norms (Table 4). The results of the total effects showed social media usage significantly increased perceived risks (β=0.217, P<0.001), perceived barriers (β=0.221, P<0.001), and perceived social norms (β=0.173, P<0.001) and only had a weak effect on perceived benefits (β=0.076, P=0.06). Thus, H6a, H12a were supported, while H8a was weakly supported and H10a was the opposite. In contrast to formal media usage (β=-0.164, P<0.001), social media usage (β=0.221, P<0.001) and interaction (β=0.330, P<0.001) significantly increased perceived barriers.

Table4  Total effects of media usage on health beliefs and perceived social norms

4.4　Mediation effects of media usage

To answer RQ3,following the mediation test in PLS proposed by Zhao et al^[52], Nitzl et al.^[51] and Shiu et al.^[53], we further tested the mediation effects of formal media usage, social media usage, and social media interaction on intentions to get vaccinated against COVID-19 through four cognitive factors (perceived risks, perceived benefits, perceived barriers, and perceived social norms). The results are demonstrated in Table 5. For the effect of formal media usage on vaccination intention, the direct effect was significant (β=0.082, P=0.02), and the three specific indirect effects of perceived benefits (β=0.090, P<0.001), perceived barriers (β=0.024, P=0.001), and perceived social norms (β=0.151, P<0.001) were also significant. The complete indirect effect size was .268 (P<0.001), accounting for 76.6% of the total effect (VIF=0.766). Therefore, the complementary partial multiple mediation was confirmed.

Table 5  The indirect effects of media usage on vaccination intention

For the effect of social media usage on vaccination intention, the direct effect was not significant (β=-0.022, P=0.55), but the specific indirect effects of perceived social norms (β=0.056, P=0.007) and the parallel chain indirect effect of social media interaction and perceived barriers (β=-0.028, P<0.001), and perceived social norms (β=0.021, P=0.02) were significant. The two indirect effects involving perceived social norms were positive, while the indirect effect involving perceived barriers was negative, and the complete indirect effect (β=0.098, P=0.001) was positive, so it belonged to the inconsistent full mediation model. In addition, social media interaction affected vaccination intention through perceived barriers and perceived social norms, but the sign was opposite, and the complete indirect effect was not significant (β=0.014, P=0.49). Since the direct effect was not significant, it also belonged to the inconsistent full mediation model.

4.5　Moderation effects of media dependence

To further clarify the boundaries of the effects of media usage and interaction on vaccination intentions (RQ4), we tested the moderating effects of media dependence. The ratio of formal media usage to social media usage was used to describe the degree of media dependence and served as the basis for grouping. If the ratio was less than 0.8, there was social-media-dependence (Group 1; 76 samples). If the ratio was between 0.8-1.2, there was non-media-dependence (Group 2; 596 samples). If the ratio was greater than 1.2, there was formal-media-dependence (Group 3; 358 samples).

Multi-group analysis (MGA) was used to conduct path analysis for the three groups, and the Welch-Satterthwaite Test was used to test the difference of effects between groups.The results are shown in Table 6. The results showed that the positive effect of formal media usage on vaccination intention was significant in group 1 (β =0.437, P<0.01) and group 3 (β =0.222, P<0.001), but not in group 2 (β =0.064, P >0.05). The effect size of formal media usage in group 1 (β1-β2 =0.373, P<0.05) and group 3 (β3-β2 =0.158, P<0.1) was significantly larger than that in group 2. Therefore, H13 was supported.

Table 6  Results of Multi-group Analysis

The results also showed that the positive effects of social media usage (β=0.342, P<0.001) and social media interaction (β=0.073, P<0.05) on vaccination intention were significant in group 2, but not in group 1 (β=0.038, P>0.05)and group 3 (β=-0.008, P>0.05), and the effect size in group 2 was significantly larger than that in group 1(β2-β1 =0.304, P<0.1) and group 3 (β3-β2 =0.350, P<0.001). Therefore, H14 was supported. Similarly, the positive effect of social media interaction on vaccination intention was significant in group 2 (β=0.073, P<0.05), but not in group 1 (β=0.045, P>0.05) and group 3 (β=-0.036, P>0.05). The effect size of social media interaction in group 2 was significantly larger than that in group 3 (β2-β3 =0.108, P<0.05), but not significant different from group 1(β2-β1 =0.027, P>0.05). Therefore, H15b was supported, but H15a was not.

In addition, the results suggested an interesting phenomenon: in the non-media-dependent group, social media usage and interaction had a significant impact on health beliefs, perceived social norms, and vaccination intentions, while formal media usage did not.

05

Discussion

5.1　Principal findings

To the best of our knowledge, this study is the first to comprehensively examine the psychological, social, and informational factors that influence the public’s intention to be vaccinated against COVID-19 in low-risk China and how these factors work together. Previous studies have examined the influence of social demographics and public health beliefs^{[11,13-14,20]}. This study also examines the combined effects of perceived social norms and media usage in terms of frequency, interaction, and dependence. Below, we discuss the main findings of this study.

First, the public’s intention to be vaccinated against COVID-19 was high (85.63%), but it continued to decline compared with 91.9%^[13] in March 2020 and 88.6%^[11] in November and December 2020. More than one-third of people have delayed vaccination. This finding indicates that there is a degree of vaccine hesitancy and a worsening trend in low-risk China, which is in line with previous studies showing a decline in vaccination intentions^[11-12] and vaccine hesitancy in China^[11,14]. The main reason for reluctance to get vaccinated is concerns about the safety and efficacy of the COVID-19 vaccine, which has been shown in previous studies^[8,54].

Second,our results demonstrate that health beliefs (perceived risks, perceived benefits, and perceived barriers) are predictors of vaccination intentions for COVID-19, which is consistent with previous studies^[14-15]. HBM is suitable for the prediction of vaccination intentions. Based on the effect size of the health beliefs, the largest one is perceived benefits, followed by perceived barriers, and the smallest is perceived risks. The perceived benefits are 4.6 times the perceived risks. This suggests that under a low perceived risk (39.4 points out of 100) situation, risk factors are not the main factors influencing public vaccination intentions. The outcome expectations of vaccination, including possible benefits and possible adverse consequences, are the most important factors, which is consistent with the findings of Lin et al.^[14] and Carpenter^[23].

Third, our results indicate that perceived social norms are the strongest predictor of COVID-19 vaccination intentions, with an effect size much larger than health beliefs, which is similar to but different from Gouin’s^[33] view that perceived social norms (as well as perceived benefits) are the strongest predictors of adherence to social distancing guidelines. The reason for the difference may be that China is collectivist and social norms have a greater influence on people. This finding suggests that the public may get vaccinated not because they want to, but more as a social responsibility. That is, if they do not get vaccinated, they may be condemned by others. Drury et al.^[31] and Neville et al.^[48] explained this with the social identity model, arguing that the sense of shared destiny during the COVID-19 pandemic will facilitate an emergent shared social identity and collective action. Group norms (i.e., social norms) may form to constrain group members, especially in collectivist cultural countries (such as China).

Fourth, in terms of information factors, both formal media (68.5 points) and social media (58.5 points) are important information sources for the public to obtain information about COVID-19 and vaccines, but they have different effects on health beliefs, perceived social norms, and vaccination intentions. The frequency of formal media usage can not only improve the public’s perceived benefits and perceived social norms but also reduce the perceived barriers, thus indirectly promoting the public’s vaccination intentions. Although the frequency of social media usage and interaction can impact the public's perceived norms and perceived risks, it can also increase perceived barriers, thus weakening the positive impact on vaccination intentions with only one-fourth of the effect size of formal media (0.07 vs. 0.28). These findings are in line with Allington’s and Romer’s findings that the vaccination intention is positively correlated with traditional media usage^[16-17,40], but are at odds with the view that social media usage has no effect^[16] or even has a negative effect^[17,40]. The possible reason for this is that the Chinese public has a collectivist culture and is more sensitive to social norms,and the sociality of social media can just improve the public’ s perceived social norms to promote vaccination. Another reason may be that social media supervision in China is stricter. As a result, there is relatively less misinformation about COVID-19 in China. In addition, our study reveals that the perceived barriers to vaccination are caused by social media interaction rather than social media usage, and therefore social media interaction does not promote vaccination intentions. Our study highlights the positive role of access to COVID-19 epidemic and vaccine information through different media outlets (formal media or social media) in promoting vaccination, but the negative impact of social media interaction on vaccination should be noted.

Fifth, our results also indicate that media dependence is a moderator of the relationship between media usage and interaction with vaccination intentions. With formal or social media dependence, formal media usage can increase vaccination intentions, while social media usage and interaction cannot. This is in accordance with Allington’s view that social media use has no positive effect^[16], but differs from our previous results that did not distinguish media dependence. However, without media dependence, social media usage and interaction significantly positively influenced vaccination intentions, while the effect of formal media usage was not significant. This is inconsistent with previous reports^[16-17,40] and also beyond our expectations. The possible reason for this is that when a user accesses COVID-19 information on formal media and social media at the same time, the information in the two media forms complements each other and provides a basis for the formation of correct beliefs about health and social norms. But the social nature (such as social network, social interaction, and social engagement) of social media will further amplify the publics’ perceived risks, benefits, and social norms, thus more significantly affecting vaccination intentions.

5.2　Theoretical and practical implications

5.2.1　Theoretical implications

The results of this study provide some implications for theory. First,although previous studies have investigated the impact of individual health beliefs on vaccination intentions, few studies have focused on the impact of social norm beliefs on them. However, this study shows that it is the most important predictor of vaccination during the period of low perceived risk, thus expanding the cognitive mediating dimension of the protection motivation theory.

Second, although previous studies have investigated the impact of social media usage on vaccination intentions, there are differences in results, lack of Chinese evidence, and no in-depth analysis of its mechanism and reasons for differences. This study complements the evidence in China, and for the first time, by distinguishing between social media usage and social media interaction, and considering the mediating factors of health beliefs and social norm beliefs, analyzes the mechanism of their effects on vaccination intentions, and explains the difference of effects in eastern and western contexts. The results reveal the positive mediating effects of social media usage and interaction through perceived social norms, whereas previous studies did not. Meanwhile, it also reveals the masking mediating effects of social media usage and interaction through perceived risks and perceived barriers, thus promoting further understanding of the mechanism of social media on vaccination intentions.

The results better explain the differences in perceived social norms and intentions of COVID-19 vaccination caused by cultural differences between East and West (collectivism VS individualism). Finally, we considered the moderating effects of media dependence to better understand the effect boundary of media usage on vaccination intentions, which was seldom considered in previous studies.

5.2.2　Practical implications

Our findings provide some insights into health communications promoting COVID-19 vaccination intentions in the low-risk phase. First, the results show that perceived social norms are the strongest predictor of vaccination intentions, followed by perceived efficacy of vaccines (perceived benefits and perceived barriers) and perceived risks of COVID-19. Thus, persuasive messages should highlight the social norms of COVID-19 vaccination, such as the descriptions that everyone is willing to or has been vaccinated, as well as the social responsibility of vaccination and the social harm of not getting vaccinated. The messages should also emphasize the effectiveness of vaccines and the benefits to individuals and society, while dispelling concerns about the negative effects of vaccines. In addition, the public's perception of risk is generally low during the well-contained phase, that’s why how to enhance people's risk awareness is the focus of health risk communication.

Second, the results show that more exposure to COVID-19 information can increase vaccination intentions, and formal media has played a greater role than social media. Therefore, correct and scientific information and knowledge should be frequently and promptly communicated to the public through various media. Formal media should play a leading role in this information dissemination, while social networking functions of social media should also be executed to improve the arrival rate of information and enhance awareness of social norms. Moreover, the results show that social media usage and interaction can significantly increase individuals' awareness of risks, including not only perceived risks of COVID-19 but also perceived barriers to vaccination (risks). Consequently, it is important to add the supply of correct and scientific vaccine knowledge on social media, as well as strengthen the monitoring and management of misinformation.

In addition, different strategies should be adopted depending on whether the public has media dependence. For those with media dependence, the access to formal media should be increased as much as possible, while for those without media dependence, the access to social media should be increased as much as possible.

5.3　Limitations

This study has several limitations. First, although our sample is in line with the social-demographic characteristics of China, the sample data is very limited relative to the total population. The online survey method has limitations. Second, this is a cross-sectional survey conducted during the well-contained phase in China, so caution should be taken when extending this data to high-risk areas or countries with individualistic cultures.

06

 Conclusions

This study investigates how media usage and interaction, health and social norm beliefs influence vaccination intentions. The results show that both health beliefs and perceived social norms are direct predictors of vaccination intentions. Media usage and interaction indirectly affect vaccination intentions through health and social norm beliefs, and formal media plays a greater role than social media. The results also suggest that attention should be paid to the adverse effects of social media usage and interaction on perceived barriers and to the differences between media dependence groups. This study provides insights into health crisis communication and intervention.

参 考 文 献

作者简介

郭路生(通讯作者)，硕士，副教授，研究方向为应急信息管理与传播，Email: [e-mail]；

周瑶瑶，硕士生，研究方向为应急信息管理与传播；

周金凤，本科生，研究方向信息管理与传播。

*原文载于《信息资源管理学报》2022年第3期，欢迎个人转发，公众号转载请联系后台。

* 引用格式

郭路生，周瑶瑶，周金凤.理解公众常态化防疫期的新冠疫苗接种意愿——从媒体接触与交互、健康与社会规范信念的视角[J].信息资源管理学报，2022,12(3):100-117.

往期 · 推荐

制版编辑 | 王伊杨

审核 | 于阿媛

长按识别二维码关注我们

信息资源管理学报

微信号

xxzyglxb

分享、在看与点赞

只要你点，我们就是朋友😊