Secondary Risk in COVID-19 Vaccination Intentions: A Theoretical Contribution to the Protection Motivation Paradigm

Abstract

This study presents an original model that features the emotion of fear of COVID-19 as a direct effect on vaccination intentions. A central research question addressed in the study is what roles do the emotion of fear of contracting COVID-19 and the threat posed by uptake of the COVID-19 vaccination play in levels of vaccination intention? The study used a structural equation model (SEM) and applied the SmartPLS 3.2.6 data analysis tool for model estimation and multivariate analysis variables. A key finding is that vaccination resistance is strongest when fear of COVID-19 is lower, and vaccination threat higher. Vaccination threat appraisal and vaccination intention were found to have a negative relationship. Response costs at higher levels lessen motivation for COVID-19 vaccination. Research implications include research-based targeting of differing segments by their primary fear, either fear of COVID-19 or of the preventative vaccine.

Keywords

protection motivation fear secondary threat appraisal vaccination intention coping appraisal vaccination threat

Introduction

As of October 13, 2022, globally, there have been 619,770,633 confirmed cases of COVID-19, including 6,539,058 deaths (World Health Organization 2022). By this time, more than 90 million were infected in the United States alone, and more than 1 million people died because of COVID-19 (Johns Hopkins University 2022). However, globally, only 68.2 percent of the world population has received at least one dose of a COVID-19 vaccine up to October 2022. For instance, the share of people who received at least one dose of COVID-19 was as follows: Brunei (97.6 percent); South Korea (87.1 percent); United States of America (78.6 percent); Hungary (65.7 percent); and Senegal (8.8 percent) (Randall et al. 2022). Although it is still unclear exactly how many people will need to be vaccinated in order to achieve herd immunity to COVID-19, experts have suggested that the vaccination rates in the 70 to 85 percent are needed (Maxouris and Vera 2021). With high vaccination rate requirements and ongoing global concerns regarding vaccine efficacy and safety, this study examines influential factors affecting COVID-19 vaccination intentions. These include primary threat appraisal (risk of the disease), secondary threat appraisal (risk of vaccination), coping appraisal, and fear associated with contracting the disease.

A recent study disclosed that vaccination motivations were far below optimal in the United States. About 40 percent of the U.S. population were either unlikely to get vaccinated or were hesitant about their vaccination intentions (Ruiz and Bell 2021). In another study, based in Hong Kong, similar results were obtained (Wong et al. 2021). In both studies, high perceived probabilities of susceptibility and severity of contracting COVID-19 disease were positive factors in the likelihood of taking the COVID-19 vaccination. These studies set the stage for the current research that employs protection motivation theory (PMT) as a framework. The essence of PMT is a fear-drive model where fear is a factor that motivates protective health behaviors. The model assumes that fear usually generates acceptance of recommended health behaviors. Protection motivation is a resultant of cognitive threat and coping appraisals. Therefore, the present study fills that gap with the strategic introduction of protection motivation model mediation by a perceived threat of self-protection strategies, specifically the threat posed by COVID-19 vaccination. The inclusion of the secondary risk of the vaccination into the model itself demonstrates a significant advance in PMT and its applications.

Theoretical Background and Hypotheses

Vaccine Hesitancy

The WHO Strategic Advisory Group on Experts (SAGE) on Immunization defined vaccine hesitancy as “a delay in acceptance or refusal of vaccination despite the availability of vaccination services” (MacDonald et al. 2015). However, vaccine hesitancy is not a new issue and is not unique to COVID-19. Previous research finds that vaccine hesitancy has influenced disease prevention of past pandemics, including Measles, Mumps, and Rubella (MMR), Severe Acute Respiratory Syndrome, Influenza A/H1N1, Middle East Respiratory Syndrome, and Ebola Virus Disease (DeStefano and Thompson 2004; Majid et al. 2020). For instance, some parents believed that the MMR vaccination could cause autism, which was later proven to be untrue (Taschner 2021). In addition, vaccine hesitancy can be triggered by fear of potential side effects (e.g., adolescent girls’ concern about possible reactions to HPV vaccination) (Herman et al. 2019; Karafillakis et al. 2019), including those associated with the new COVID-19 vaccines (Rosenbaum 2021).

Protection Motivation Theory

PMT has accounted for a significant amount of variance for intentions to vaccinate. That is to say that it predicts vaccination intentions quite well. In one study, all components of PMT except perceived costs of vaccination uniquely accounted for variance in intentions to vaccinate (Ling, Kothe, and Mullan 2019). PMT components included perceived susceptibility, severity, perceived benefits of not vaccinating, self-efficacy to vaccinate and perceived efficacy of vaccinating. A number of studies have explored only selected dimensions of PMT. These dimensions predicted vaccination intentions for a variety of vaccines (Freimuth et al. 2017; Morgan et al. 2010; Parsons, Newby, and French 2018; Weinstein et al. 2007). The next section discusses coping appraisal and threat appraisal which are distinct cognitive processes.

Coping Appraisal and Protective Motivation

Coping appraisal refers to one’s being able to handle threatening matters. Coping appraisal behaviors concern (1) appraisals of self-protective alternatives, and (2) assessment of behavioral options that are altruistic. Coping appraisal is comprised of response efficacy and self-efficacy. Response efficacy involves beliefs about the effectiveness of recommended protective behaviors. Self-efficacy involves beliefs about the ability to engage in the appropriate protective behaviors (Teasdale et al. 2012). Response efficacy has been linked to adaptive behaviors during pandemics. Understanding response efficacy implies a complex balancing of cognitions and motivations (Teasdale et al. 2012).

Primary Threat Appraisal

One source of information is persuasive messaging disseminated through the media. The assumption is that attitudes and beliefs as well as expectations of future events are major determinants of self-protective behaviors (Munro et al. 2007). The protection motivation model has been widely used to enhance the effectiveness of health-related appeals. One application is its use in fine-tuning fear appeals. For example, an ordered protection motivation model has been advanced. In this model, threat leads to fear when both severity of threat and probability of occurrence are perceived as high.

Fear arousal has been measured by mood arousal adjectives. These include being frightened, tense, nervous, anxious, uncomfortable, and nauseous. Another study outlined the importance of coping appraisal during the pandemic flu. Coping appraisal in the context of the pandemic flu found that perceived pandemic severity influenced threat and coping appraisals and vaccination intentions (Teasdale et al. 2012).

The possibility of high secondary risk perceptions led to a preliminary investigation in the search for a new model of PMT (Cummings, Rosenthal, and Kong 2021). This model was tested in the United States and Singapore, involving protective responses to four infectious diseases. These were dengue fever, chikungunya, bacterial meningitis, and cholera. The illnesses were described to respondents. COVID-19 was not mentioned in this study.

This was an experimental study, in which respondents indicated their intentions to take a vaccine if they were suffering from one of these diseases. Different groups of respondents were given different information about the likelihood and the severity of vaccination side effects. Given the particular experimental design, no information was issued on associations of variables within or between protection motivation or secondary threat evaluation domains. The study also neglected the impact of fear as a driving force. Study results showed similar predictions for the extended PMT when the likelihood and severity of side effects were both low or both high, which is theoretically counterintuitive.

The model advanced used secondary threat variables as an add-on to the traditional PMT. By simply adding independent variables to an existing model, better overall predictions of vaccination intention were obtained.

Individual and environmental factors explain why individuals engage in protective behaviors. These factors include perceptions, appraisals, motivations, and behaviors in response to environmental threats. PMT incorporates these factors. It helps to more fully understand and predict responses to threats to health and safety. A study by Schulz and Hartung (2021) concerned decisions about uptake of different vaccinations. It found that motivations to vaccinate are higher the more alarming the person’s threat appraisal and the more promising their coping appraisals. PMT was found to explain vaccination behaviors for tetanus, pertussis, measles, hepatitis B, meningitis, and two strains of influenza.

In another study (Ort and Fahr 2018), efficacy cues and threats were employed in persuasive vaccination communications. An important association was found for both threat and coping appraisals in facilitating recommended behavioral changes to protect against Ebola. Fear and danger control processes were found to influence vaccination intentions. These processes may be attributable to psychological vulnerability factors. Such factors include intolerance of uncertainty and susceptibility to a disease (Taylor 2019). Theory and related empirical evidence support the proposition that protection motivation explains intentions to adopt COVID-19 vaccination.

Moderators of Relations between Protective Motivation and Vaccination Intentions

A systematic review of barriers to vaccination (Schmid et al. 2017) found that low perceived risk of disease and doubts about vaccination efficacy gaps were the main barriers to influenza vaccination.

In another study based on PMT, prediction of influenza vaccination intentions was significantly affected by response costs (Ling et al. 2019). Vaccination for protection against COVID-19 was newly developed, made available for the first time, and conditionally released. This served to make the vaccination threat more salient in vaccination decisions than might have been the case heretofore. Cognitive appraisal theories are the mainstream framework that account for elicitation of fear (Arnold 1960; Ellsworth and Scherer 2003; Frijda 1986; Lazarus 1966; Scherer 2001). Cognitive dimensions in threat appraisal are based on relevant environmental encounters associated with a concrete sense of danger, a component of fear (Balzarotti and Cicero 2014). Cognitive threat leads to fear, especially under conditions of uncertainty (Smith and Ellsworth 1985). In their experimental study of secondary risk (Cummings et al. 2021), it was found that this type of risk was a significant predictor of vaccination intention. The discrepancy between the model and actual data was high when secondary risk was severe and likely to occur.

Hypotheses

Fear of disease was found to be an important predictor of high-risk behaviors (Baghiani-Moghadam et al. 2015). Fear is a mediator between perceived susceptibility, perceived severity, and protective motivation. Raising fear to higher levels results in stronger motivations to engage in protective behavior. Fear has served as a mediator in PMT predictions of behavioral intentions (Kim et al. 2021). It has been positively positioned as a function of threat, self-efficacy, and response efficacy (Rippetoe and Rogers 1987). In turn, fear results in a mindset of avoidance thinking and indirectly in intentions to engage in specific behaviors. Fear of contracting COVID-19 involves disgust, insensitivity, and other aspects of the broader emotional system. Fear stems in part from physical concerns, moderated by emotional reactions (McKay et al. 2020).

Hypothesis 1 (H1): Fear of developing COVID-19 positively affects vaccination intent.

Being informed or aware of recommendations for vaccination has been associated with vaccine receipt. Concerns about vaccination safety and the risk of side effects—vaccination threat—depress motivation to vaccinate (Gaygisiz et al. 2010). Relatedly, this includes perceptions of failure to receive a benefit from taking a vaccine. A study by Barr, Raphael, and Taylor (2008) on influenza pandemic found that the fear construct predicted a significant percentage of preventive behaviors. The threat of vaccination is associated with the protective motivation construct of response costs. Response costs involve negative outcomes to individual actions. Response costs have been shown to predict intentions of receiving seasonal influenza vaccine (Falato, Ricciardi, and Franco 2011; Freimuth et al. 2017; Weinstein et al. 2007).

Hypothesis 2 (H2): Vaccination threat negatively affects vaccination intent.

Threat appraisal is based on subjective estimates of (1) the severity posed by a health threat and (2) the felt vulnerability to that same threat (Floyd, Prentice-Dunn, and Rogers 2000). The PMT model includes sources of information, cognitive mediating processes, and coping modes. The emotional state of fear is theorized to raise attention and message believability. This leads to increasing the likelihood of coping appraisal (Tanner, Hunt, and Eppright 1991). Subjective probabilities of severity and vulnerability with respect to COVID-19 may generate fear arousal under PMT (Rippetoe and Rogers 1987). In short, threat appraisal is linked to fear arousal. The threat appraisal process elicits the sense of danger.

Hypothesis 3 (H3): Primary threat appraisal of COVID-19 positively affects fear of COVID-19.

Negative beliefs about the effectiveness the COVID-19 vaccine led to a negative attitude toward vaccination. Coping appraisal that reflects a negative vaccination attitude decreases the likelihood of taking the vaccine (Khoury and Salameh 2015; Myers and Goodman 2011). Without assimilation of information about vaccine safety and effectiveness, there is ambivalence about taking the vaccine. Perceived high risk of COVID-19 infection is a positive factor in threat appraisal and uptake of the vaccine. This stems from a sense of maximal susceptibility to COVID-19. With serious worries or concerns about contracting the disease, motivations to take the vaccine are heightened (Beattie et al. 2013; Villacorta and Sood 2015). Vaccine threat moderates relationships between coping appraisal, threat appraisal, and vaccination intent. This is because heightened vaccine threat, in itself, decreases vaccine intake (Podlesek, Rosrosdkar, and Komidar 2011; Redelings et al. 2012). Therefore, appraisal processes are linked with vaccination threat.

Hypothesis 4 (H4): Secondary threat coping appraisal negatively affects vaccination threat.

Perceived severity, perceived vulnerability, outcome efficaciousness, and self-efficacy have correlated positively with the frequency of health protective behaviors with respect to the COVID-19 virus (Kowalski and Black 2021). Protective behaviors include social distancing, hand washing, hand sanitizing, disinfecting, and self-quarantining. Willingness to take a vaccine has been found to be positively associated with a generalized sense of fear. Fear arousal was found to be a mediator between protection motivation and behaviors (Clubb and Hinkle 2015; Kim et al. 2021). From the perspective of PMT, fear is a resultant of the (1) perceived magnitude of the threat, (2) the probability of the threat’s occurrence, and (3) the efficacy of the coping response (Rogers 1975). Generalized fear is defined as excessive and unrestricted fear. It is reflected in a specific motivation, including willingness to take a vaccine in the interest of health protective behavior (Mesch and Schwirian 2019).

Hypothesis 5 (H5): Primary threat appraisal of COVID-19 positively affects secondary vaccination threat.

Methods

Data Collection Procedure and Sampling

After all recruitment materials and questionnaires were approved by Institutional Review Board (IRB) through the university research protection office, we obtained a total of 348 responses via U.S. Mechanical Turk (MTurk) sampling in exchange for a small monetary reward. MTurk is a clearinghouse for people who need to complete tasks and individual workers who offer their labor. Links are provided from the MTurk website to a survey or other data collection service (Mellis and Bickel 2020). MTurk is a nonprobability sampling frame platform. Therefore, there are some limitations that merit consideration. For the present study, MTurk is not vulnerable to self-selection bias. This study addresses the general adult population. Furthermore, there is no range restriction within this population. To further enhance the data quality in this research, we set restrictions to only include MTurk workers with high reputations (above 95 percent approval ratings), and with the number of HIT approved being greater than 500 (Peer, Vosgerau, and Acquisti 2014). Ten participants who did not meet our age criterion (age between 18 and 64) and 38 participants, who already received the Covid-19 vaccinations, were excluded from the analysis. This resulted in a final sample of 300 participants. The average age of participants was 39 (M_age = 38.68, SD = 10.49) and 50.7 percent were males. Approximately 64 percent of the sample had at least a bachelor’s degree, while another 25 percent of the respondents had some college education. About 66 percent of respondents had household incomes of less than $75,000. Appendix A displays the sample’s demographic profile.

Measures

All the measurement instruments were prepared in reference to existing literature. Four items for vaccination intention (α = .98) were adopted and modified from Fishbein and Ajzen (2010): (1) “I intend to take COVID-19 vaccine when it is available to me” (five-item scale from “definitely do not” to “definitely do”); (2) “I will take COVID-19 vaccine when it is available to me” (five-item scale from “extremely unlikely” to “extremely likely”); (3) “I am willing to take COVID-19 vaccine when it is available to me” (five-item scale from “false” to “true”); (4) “I plan to take COVID-19 vaccine when it is available to me” (five-item scale from “strongly disagree” to “strongly agree”). These four items served as indicators of the vaccination intent concept in the structural model.

Fear is measured by asking participants how the thoughts of developing COVID-19 make them feel, based on four items 1- (not at all frightened, not at all anxious, not at all worried, and not at all scared) to 5- (very frightened, very anxious, very worried, and very scared) point scales (α = .97) (Miline, Orbell, and Sheeran 2002).

Two dimensions of threat appraisal (α = .81) were assessed, as recommended (Brewer et al. 2007; Leppin and Aro 2009; Rimal and Real 2003): perceived susceptibility of getting COVID-19 and perceived severity of COVID-19. Participants were asked to answer the two questions on the perceived susceptibility and indicate on 5-point, Likert-type scales, ranging from 1 (extremely low) to 5 (extremely high): “Compared to most people my age, I understand that my risk of getting COVID-19 is” and “The likelihood of my getting COVID-19 is.” The perceived severity was assessed with two questions on a 5-point, Likert-type scale, ranging from 1 (strongly disagree) to 5 (strongly agree): “COVID-19 is a serious disease” and “COVID-19 is more deadly than most people realize.”

Adopted from Witte (1994), coping appraisal (α = .89) was measured by two constructs (self-efficacy and response efficacy) and was assessed using a four-item, 5-point, Likert-type scale, ranging from 1 (strongly disagree) to 5 (strongly agree). The specific statements were: “It is possible to carry out protective behavior by taking the COVID-19 vaccination,” “I am confident that I could take the COVID-19 vaccination, if I wished to have it,” “People can protect themselves from COVID-19 by taking the vaccine,” and “The COVID-19 vaccine is highly effective.” Coping is reflected in the second item, which refers to the ability to make use of available vaccination sources.

Finally, based on Camerini et al.’s (2018) measurement, vaccination threat (α = .90) was calculated based on perceived susceptibility and severity of having side effects from the COVID-19 vaccine. Participants were asked to answer the two questions on the perceived susceptibility of having side effects from the vaccine and indicate on 5-point, Likert-type scales, ranging from 1 (not at all strong) to 5 (very strong): “My chances of having immediate side effects from the COVID-19 vaccine are” and “My chances of developing unanticipated future health problems from the COVID-19 vaccine are.” Perceived severity was measured with two questions on a 5-point, Likert-type scale, ranging from 1 (strongly disagree) to 5 (strongly agree): “If I had side effects from the COVID-19 vaccine I would suffer severe symptoms” and “If I developed unanticipated future health problems it might cause a serious illness.” Refer to Appendix B for items used in the survey.

Heterogeneity and Segmentation

The study highlights the importance of heterogeneity and segmentation on the basis of motivations to uptake COVID-19 vaccinations. Furthermore, identification of demographic classifications associated with these segments is critical for practical applications. In sum, heterogeneity among potential vaccinators leads to a greater understanding of this type of mass behavior.

Data Analysis

All variables used in the model are latent variables with multiple items of measurement. These are measured by manifest indicators. All indicators in this study are reflective of latent variables. The study used a structural equation model (SEM) with the PLS (PLSc-SEM) algorithm approach and applied the SmartPLS 3.2.6 data analysis tool for model estimation and multivariate analysis variables (Dijkstra 2010; Dijkstra and Schermelleh-Engel 2014). SmartPLS is a variance-based partial least squares modeling algorithm. It affords measurement and structural models.

In this study, latent variables in the model were assessed for assesses for collinearity and discriminant validity. In addition, scale reliabilities were measured. Subsequently, structural modeling provided estimates of path coefficients and their significance levels. Heterogeneity of model components and segmentation analyses were conducted.

Unobserved heterogeneity was modeled by use of finite mixture partial least squares (FIMIX-PLS) and PS-PLS (Sarstedt and Ringle 2010). SmartPLS assumes the concepts can be measured as composites. PLS estimates path model relationships.

Results

Concepts were found to be appropriately measured, with all variance inflation factor (VIF) statistics well below the VIF threshold that signals collinearity. VIF is the variance inflation factor that represents the extent to which the presence of collinearity has increased the standard error. A VIF value of five or higher indicates a potential collinearity problem. All indicators of primary and secondary risk concepts in the model are well below this threshold. VIF values range from 1.535 to 4.621. Study constructs achieved satisfactory convergent validity and internal consistency reliability. A variety of criteria were satisfied, including tests by Cronbach, rho alpha, composite reliability, and average variance extracted (see Table 1).

Table 1.

Construct Reliability.

Variable	Cronbach’s alpha	Rho_A	Composite reliability	Average variance extracted (AVE)
Coping appraisal	0.889	0.932	0.924	0.754
Fear	0.922	0.922	0.962	0.927
Threat appraisal	0.812	0.835	0.875	0.637
Vaccination intention	0.984	0.985	0.988	0.954
Vaccination threat	0.896	0.920	0.927	0.762

Study constructs were tested for discriminant validity. This established that each construct is unique and captures phenomena other than those captured by other model constructs. Constructs were tested for discriminant validity. As a rule, the square root of each construct’s average variance extracted (AVE) should be greater than its highest correlation with any other construct. Table 2 shows that all study constructs meet this standard and are independent of each other. Consistent PLS (PLSc) was used to determine the goodness of fit of the overall model. The standardized root mean square residual (SMSR) was estimated at 0.021 and the root mean square covariance was 0.017. Both of these values indicate that the model is a good fit.

Table 2.

Discriminant Validity.

Variable	Coping appraisal	Fear	Threat appraisal	Vaccination intention	Vaccination threat
Coping appraisal	.868
Fear	.374	.963
Threat appraisal	.610	.619	.798
Vaccination intention	.871	.337	.552	.977
Vaccination threat	−.389	.053	.011	−.404	.873

Table 3 shows that path coefficients from the model are all significant with p < .01.

Table 3.

Path Coefficients.

Paths	Original sample (O)	Sample mean (M)	Standard deviation (SD)	T statistics (O/SD)	p values
Fear → Vaccination intention (H1)	0.359	0.358	0.554	6.646	.000
Vaccination threat → Vaccination intention (H2)	−0.423	−0.426	0.049	8.599	.000
Threat appraisal → Fear (H3)	0.619	0.623	0.042	14.765	.000
Coping appraisal → Vaccination threat (H4)	−0.630	−0.635	0.046	13.552	.000
Threat appraisal → Vaccination threat (H5)	0.395	0.395	0.074	5.345	.000

The magnitudes of the paths are relatively strong, as well.

All hypotheses were confirmed with path coefficients as shown: H1. Fear of developing COVID-19 positively affects vaccination intent (path coefficient = 0.359). H2. Vaccination threat negatively affects vaccination intent (path coefficient = −0.423). H3. Primary threat appraisal of COVID-19 positively affects fear of COVID-19 (path coefficient = 0.619). H4. Secondary threat coping appraisal negatively affects vaccination threat (path coefficient = −0.630). H5. Primary threat appraisal of COVID-19 positively affects secondary vaccination (path coefficient = 0.395). See Figure 1 to visualize the entire model.

Figure 1.

Structural equation model of COVID-19 vaccination intention.

Latent class modeling was applied to PLS structural equations results for the entire sample. This provided information about unobserved heterogeneity and led to the analysis of respondent clusters, together with their path estimates. FIT indices were ascertained for up to five segments, corresponding to a variety of information criteria. The appropriate number of segments was less than five, based on the analysis. Seven of the 11 criteria showed a better model fit for the two-segment relative to the three-segment solution (see Table 4). Examination of relative sample sizes within the two-segment framework showed that segment 1 is much larger than segment 2 (64.6 percent vs. 35.4 percent). Model fit was estimated by the R² values for each predicted variable. Every variable had an R² weighted average significantly exceeding the counterpart in the full data set. This indicated a very good overall data fit for segmentation based on unobserved heterogeneity (see Tables 5 and 6).

Table 4.

Fit Indices for a One- to Four-Segment Solution.

Criteria	Number of segments
Criteria	2	3	4	5
AIC	1960.72	2019.86	1968.37	1960.72
AIC3	2004.72	2045.86	2003.37	2004.72
AIC4	2048.72	2071.86	2038.37	2048.72
BIC	2123.69	2116.15	2098	2123.69
CAIC	2167.69	2142.15	2133	2167.69
HQ	2025.94	2058.4	2020.25	2025.94
MDL5	3127.55	2709.35	2896.53	3127.55
LnL	−936.36	−983.93	−949.18	−936.36
EN	0.65	0.61	0.62	0.65
NFI	0.6	0.62	0.6	0.6
NEC	102.49	114.3	113.11	102.49

Note. AIC = Akaike information criterion; BIC = Bayesian information criterion.

Table 5.

Relative Segment Sizes.

Number of segments	Relative segment sizes
Number of segments	Segment 1	Segment 2	Segment 3	Segment 4
2	0.646	0.354
3	0.577	0.25	0.173
4	0.321	0.245	0.227	0.208

Table 6.

FIMIX-PLS Values for the Two-Segment Solution.

	R ²
Variables	Full data set	FIMIX-PLS	FIMIX-PLS	FIMIX-PLS
Variables	Full data set	Segment 1	Segment 2	R²-weighted average
Fear	.383	0.294	0.638	.64
Vaccination intention	.292	0.491	0.788	.6
Vaccination threat	.249	0.514	0.335	.45

Note. FIMIX-PLS = finite mixture partial least squares.

Path coefficients for segments 1 and 2 exhibit stark differences (Table 7). Importantly, vaccination intentions are mainly driven by vaccine threat in segment 1. In contrast, vaccination intention for segment 2 is largely driven by fear associated with COVID-19. For this segment, fear is more likely to be generated by threat appraisal. Finally, vaccination threat is driven by threat appraisal for segment 2.

Table 7.

Path Coefficients for Original Sample and Segments.

Hypotheses	Segment 1	Segment 2
Fear → Vaccination intention (H1)	−0.151	0.866*
Vaccination threat → Vaccination intention (H2)	−0.768*	−0.041
Threat appraisal → Fear (H3)	0.453*	0.819*
Coping appraisal → Vaccination threat (H4)	−0.804*	−0.737*
Threat appraisal → Vaccination threat (H5)	0.067	0.707*

Path coefficients significant p < .05.

Logistic regression analyses were conducted to find demographic variables to better define segments 1 and 2. Age was the sole demographic that accomplished this end. It was found to be significant for segment 1, skewing older than individuals in segment 2 (see Table 8). Other demographic qualifications, including gender, income, and education, were statistically equivalent between the two segments.

Table 8.

Logistic Regression Fitted Model for FIMIX-PLS Segments and Age.

Source	Analysis of deviance
Source	Deviance	df	p value
Model	6.3516	1	.0117
Residual	391.094	298	.0002
Total (corr.)	397.445	299

Note: Percentage of deviance explained by the model = .59 (adjusted).

Discussion

This section is organized in relation to three key originating questions:

What roles do the emotion of fear of contracting COVID-19 and the threat posed by uptake of the COVID-19 vaccination play in levels of vaccination intention?

What underlies the relation between COVID-19 threat appraisal and fear of contracting COVID-19?

How do threat appraisal and coping appraisal determine the intensity of the vaccination threat?

Vaccination resistance is strongest when fear of COVID-19 is lower, and vaccination threat higher. Vaccination threat appraisal and vaccination intention were found to have a negative relationship, as hypothesized. Coping appraisals, when strongly internalized, induce higher confidence levels, supportive outcome calculations, facilitative convenience, and a heightened sense of urgency to vaccinate (Schmid et al. 2017). Response costs are beliefs about the utilization of resources in connection with uptake of a vaccination. Response costs at higher levels lessen motivation for COVID-19 vaccination.

The preceding is strongly supported by the segmentation analysis which shows two segments: segment 1, hesitators who are more recalcitrant about vaccine uptake than segment 2, amenables. Hesitators are less fearful of the primary threat of the COVID-19 disease, in comparison with amenables, but are more fearful of side effects of preventive vaccination.

Fear of contracting COVID-19, especially among those who are subjectively vulnerable to the virus, may activate an extensive search for self-protective modes of fear reduction. Evaluation of these modes will be based on information gleaned from internal and external search, normative influences, past vaccination experiences, and personal traits such as risk avoidance and reactance proneness. Alternative evaluations will engender positive behavioral decision-making about vaccination in the interest of fear reduction if fear levels exceed a certain emotional threshold.

Vaccination threat intensity can be a resultant of anticipatory regret. Impressions of the questionable efficacy of the vaccine, together with the uncertainty of coping with the aftermath of a decision to vaccinate, further promote the intensity of vaccination threat.

In sum, vaccination intentions are conditioned by predispositions to optimize or to satisfice protective motivational outcomes of decisions to vaccinate. Decision-making styles that emerge under psychological stress represent another important factor. Beliefs about the consequences of an action and about what other people think one should do are involved. Beliefs will affect levels of vigilance with respect to information processing and its impact on COVID-19 fear arousal, vaccination threat, and vaccination intentions.

Conclusion

The felt threat posed by the COVID-19 vaccination reflects cross-pressures between expectations about vaccination side effects and personal confidence and security based on accumulated knowledge, intuition, and modes of critical thinking. This stressful state of affairs can be coped with by the process of bolstering one side of the conflict, that is, maximizing arguments on the pro-side and minimizing those on the con-side. Through this research, the authors have made several significant contributions to the literature from both theoretical and practical perspectives.

Theoretical Contributions

PMT is one of several cognitive models of health behavior. These models address health behavior outcomes derived from thinking processes, perceived consequences, and self-efficacy (Norman and Conner 2005). There is considerable overlap between model constructs of PMT, health belief model and theory of planned behavior. Therefore, advances in the theory of protection motivation will also re-shape elements of other related theories of health behavior.

From a theoretical perspective, the primary contribution of the current study is that the concept of secondary self-protection threat is introduced. Secondary self-protection refers to protections taken to minimize the risks of actions to cope with a primary threat. Secondary self-protection is a mediator of consequence between protection motivation and protective intentions. This process is a significant theoretical contribution to the protection motivation paradigm. Protection motivation simply predicts protection motivation intentions and behaviors across many domains. It is based on a traditionally fixed general model that includes threat and coping appraisals. PMT is incomplete. It fails to consider secondary risk perceptions. The latter more fully explains the expected utility of individual motivation for self-protection (Cummings et al. 2021). The authors conclude that secondary risk theory presents a more granular and robust accounting of threat and response. It outperforms traditional PMT by itself.

Another major contribution is the centrality of fear as a mediator in protective behavior intentions (Kim et al. 2021). Studies that incorporate PMT make the assumption that self-protective decisions are based on thinking processes. Emotional aspects that come into play are simply not addressed. This study is a significant advance. It presents an original model that features fear of COVID-19 as a direct effect on vaccination intentions. Fear has been further compounded by social isolation, self-confinement, quarantines, restrictions on public facilities, and the like. This state of affairs has substantially increased the salience of the emotion of fear. Fear has become a primary factor in decision-making about protective strategies (see Barbisch, Koenig, and Shih 2015). This emotional dimension comes to the forefront in this study of vaccination decisions. Peril and contamination fears have generated maladaptive coping attempts. These may divert individuals from accepting public health recommendations (Taylor et al. 2020).

Practical Implications

Unobserved segments analyses invoke a communications strategy that steers resources toward the larger segment, segment 1 which consists of an older population than segment 2. The threat of vaccination is primary for segment 1. Therefore, communications to the older segment should convey information that lowers this particular threat. For example, vaccination side effects are unrelated to age and are rare in each case. On the other hand, the smaller segment 2 with a younger population will be motivated by communications that emphasized threats posed by the disease itself. This age group is likely to be persuaded to vaccinate by a message that informs them of their own extreme vulnerability to the disease. A message to this age group Persuasive communications can be delivered through media that attract audiences that favor appropriate age groups.

Communications messages may be more effective among the younger segment. This is because this segment is already more predisposed to seek vaccinations. They will be less inclined to counterargue with the communications message. However, the older segment is at greater risk if they are disinclined to seek vaccinations. Despite expectations that communications will have a lesser effect on this segment, this segment should be an important part of the total communications effort.

Other implications of these finding are that information about vaccination safety should be disseminated through communication campaigns, especially targeted to uninformed audiences. Information about vaccine efficacy, relative to other preventive measures, will promote positive vaccination behavior. Vaccination effectiveness perception, itself, has been found to augment vaccination intentions (Weinstein et al. 2007). Information from this study helps to design campaigns to meet COVID-19 protective needs of target groups. Willingness to vaccinate protectively against COVID-19 will be further enhanced by lowering response costs. Response costs may be addressed by messages minimizing the vaccination’s side effects and simplifying perceptions of the task of accessing the vaccination. Self-efficacy perceptions will be increased as a by-product of this approach, further lowering barriers to positive vaccination decisions.

Limitations and Future Research

The study model does not incorporate the impact of public communications about COVID-19 and the vaccination process. It fails to incorporate personal influence of primary group members. While this study applied the PMT framework, it omitted other variables that are predictors and mediators of vaccination intent. Omitted variables that may be invoked in future studies include awareness and knowledge of the COVID-19 disease, its prevalence and effects, prevention, and treatments. Personal health conditions that give pause or generate eagerness, and histories of previous vaccinations are also a key factor for inclusion in future studies of vaccination intention. Future studies of secondary risk should incorporate threat appraisals of medications to minimize severity of COVID-19, among those who have tested positive for COVID-19.

Footnotes

Appendix

Appendix B.

Measurement Items of the Survey Questionnaire (Five-Point Likert-Type Scales).

Construct	Item	Source
Fear	The thought of developing COVID-19 makes me feel: 1. Frightened 2. Anxious 3. Worried 4. Scared	Miline, Orbell, and Sheeran (2002)
Vaccination intention	If available, how do you feel about taking COVID-19 vaccine? 1. I intend to take COVID-19 vaccine when it is available to me. 2. I will take COVID-19 vaccine when it is available to me. 3. I am willing to take COVID-19 vaccine when it is available to me. 4. I plan to take COVID-19 vaccine when it is available to me.	Fishbein and Ajzen (2010)
Threat appraisal	We want you to think about personal concerns you may have during this COVID-19 pandemic.	Rimal and Real (2003)
COVID-19 Susceptibility	1. Compared with most people my age, I understand that my risk of getting COVID-19 is ____. 2. The likelihood of my getting COVID-19 is ____.
COVID-19 severity	1. COVID-19 is a serious disease. 2. COVID-19 is more deadly than most people realize.
Coping appraisal	Please complete the following questions about the COVID-19 vaccination itself.	Witte (1994)
Self-efficacy	1. It is possible to carry out protective behavior by taking the COVID-19 vaccination. 2. I am confident that I could take the COVID-19 vaccination, if I wished to have it.
Response efficacy	1. People can protect themselves from COVID-19 by taking the vaccine. 2. The COVID-19 vaccine is highly effective.
Vaccination threat		Camerini et al. (2018)
Vaccination susceptibility	1. My chances of having immediate side effects from the COVID-19 vaccine are ____.2. My chances of developing unanticipated future health problems from the COVID-19 vaccine are ____.
Vaccination severity	1. If I had side effects from the COVID-19 vaccine I would suffer severe symptoms.2. If I developed unanticipated future health problems, it might cause a serious illness.

Declaration of Conflicting Interests

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

Funding

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

ORCID iD

Melvin Prince

Author Biographies

Melvin Prince is a professor of marketing at the Southern Connecticut State University. He teaches marketing research, marketing planning, and consumer behavior at the graduate level. His publications have appeared in such journals as International Journal of Qualitative Market Research, Journal of Advertising, Journal of Advertising Research, and Academy of Marketing Science Review.

Young Kim is a tenured associate professor of marketing at the Southern Connecticut State University. He holds a PhD in marketing from the University of Iowa.

References

Arnold

M. B.

1960. Emotion and Personality. Vol. 2. New York: Columbia University Press.

Baghiani-Moghadam

M. H.

Seyedi-Andi

S. J.

Shokri-Shirvani

Khafri

Ghadimi

Parsian

2015. “Efficiency of Two Constructs Called ‘Fear of Disease’ and ‘Perceived Severity of Disease’ on the Prevention of Gastric Cancer: Application of PMT.” Caspian Journal of Internal Medicine 6(4):201–208.

Balzarotti

Cicero

M. R.

2014. “News Reports of Catastrophes and Viewers’ Fear: Threat Appraisal of Positively and Negatively Framed Events.” Media Psychology 17(4):357–77.

Barbisch

Koenig

K. L.

Shih

F. Y.

2015. “Is There a Case for Quarantine? Perspectives from SARS to Ebola.” Disaster Medicine and Public Health Preparedness 9(5):547–53.

Barr

Raphael

Taylor

2008. “Pandemic Influenza in Australia: Using Telephone Surveys to Measure Perceptions of Threat and Willingness to Comply.” BMC Infectious Disease 8:117.

Beattie

Palmer

Rees

Riddell

Roberts

Jordan

2013. “Factors Affecting the Acceptance of Pandemic Influenza a H1N1 Vaccine against Essential Service Providers: A Cross-sectional Study.” Vaccines 1(1):17–33.

Brewer

N. T.

Chapman

G. B.

Gibbons

F. X.

Gerrard

McCaul

K. D.

Weinstein

N. D.

2007. “Meta-analysis of the Relationship between Risk Perception and Health Behaviour: The Example of Vaccination.” Health Psychology 26:136–45.

Camerini

A.-L.

Diviani

Fadda

Schulz

P. J.

2018. “Using PMT to Predict Intention to Adhere to Official MMR Vaccination Recommendations in Switzerland.” SSM Population Health 7:1–11.

Clubb

A. C.

Hinkle

J. C.

2015. “PMT as a Theoretical Framework for Understanding the Use of Protective Measures.” Criminal Justice Studies 28(3):336–56.

10.

Cummings

C. L.

Rosenthal

Kong

W. Y.

2021. “Secondary Risk Theory: Validation of a Novel Model of Protection Motivation.” Risk Analysis 41(1):204–20.

11.

DeStefano

Thompson

W. W.

2004. “MMR Vaccine and Autism: An Update of the Scientific Evidence.” Expert Review of Vaccines 3:19–22.

12.

Dijkstra

T. K.

2010. “Latent Variables and Indices: Herman Wold’s Basic Design and Partial Least Squares.” Pp. 23–46 in Handbook of Partial Least Squares; Concepts Methods and Applications, edited by W. W.

Vesposito

Chin

Henseler

Wang

Berlin, Germany: Springer.

13.

Dijkstra

T. K.

Schermelleh-Engel

2014. “Consistent Partial Least Squares for Nonlinear Structural Equation Models.” Psychometrika 79(4):585–604.

14.

Ellsworth

P. C.

Scherer

K. R.

2003. “Appraisal Processes in Emotion.” Pp. 572–95 in Handbook of Affective Sciences, edited by R. J.

Davidson

Scherer

K. R.

Goldsmith

H. H.

New York: Oxford University Press.

15.

Falato

Ricciardi

Franco

2011. “Influenza Risk Perception and Vaccination Attitude in Medical and Nursing Students during the Vaccination Campaigns of 2007 and 2008.” La Medicina del Lavoro 102(2):208–15.

16.

Fishbein

Ajzen

2010. Predicting and Changing Behavior: The Reasoned Action Approach. New York: Psychology Press.

17.

Floyd

D. L.

Prentice-Dunn

Rogers

R. W.

2000. “A Meta-analysis of PMT.” Journal of Applied Social Psychology 30(2):407–29.

18.

Freimuth

V. S.

Jamison

Hancock

Musa

Hilyard

Quinn

S. C.

2017. “The Role of Risk Perception in Flu Vaccine Behavior among African-American and White Adults in the United States.” Risk Analysis 37(11):2150–62.

19.

Frijda

N. H.

1986. The Emotions. New York: Cambridge University Press.

20.

Gaygisiz

Ozkan

Lajunen

2010. “Why Were Turks Unwilling to Accept the H1N1 Influenza Pandemic Vaccination? People’s Beliefs and Perceptions about the Swine Flu Outbreak and Vaccine in the Latter Stage of the Epidemic.” Vaccine 29(2):329–33.

21.

Herman

McNutt

L. A.

Mehta

Salmon

D. A.

Bednarczyk

R. A.

Shaw

2019. “Vaccination Perspectives among Adolescents and Their Desired Role in the Making Process.” Human Vaccines & Immunotherapeutics 15:1752–59.

22.

Johns Hopkins University. 2022. “COVID-19 Dashboard by the Center for Systems Science and Engineering (CSSE) at Johns Hopkins University (JHU).” Retrieved February 23, 2023 (https://coronavirus.jhu.edu/map.html).

23.

Karafillakis

Simas

Jarrett

Verger

Peretti-Watel

Dib

De Angelis

Takacs

Ali

K. A.

Celentano

L. P.

Larson

2019. “HPV Vaccination in a Context of Public Mistrust and Uncertainty: A Systematic Literature Review of Determinants of HPV Vaccine Hesitancy in Europe.” Human Vaccines & Immunotherapeutics 15:1615–27.

24.

Khoury

G. E.

Salameh

2015. “Influenza Vaccination: A Cross-sectional Survey of Knowledge, Attitude and Practices among the Lebanese Adult Population.” International Journal of Environmental Research and Public Health 12(12):15486–97.

25.

Kim

Yang

Min

White

2021. “Hope, Fear, and Consumer Behavioral Change Amid COVID-19: Application of Protection Motivation Theory.” International Journal of Consumer Studies 46:558–74.

26.

Kowalski

R. M.

Black

K. J.

2021. “Protection Motivation and the COVID-19 Virus.” Health Communication 36(1):15–22.

27.

Lazarus

R. S.

1966. Psychological Stress and the Coping Process. New York: McGraw-Hill.

28.

Leppin

Aro

A. R.

2009. “Risk Perceptions Related to SARS and Avian Influenza: Theoretical Foundations of Current Empirical Research.” International Journal of Behavioral Medicine 16:7–29.

29.

Ling

Kothe

E. J.

Mullan

B. A.

2019. “Predicting Intention to Receive a Seasonal Influenza Vaccination Using PMT.” Social Science & Medicine 233:87–92.

30.

MacDonald

N. E.

Eskola

Liang

Chaudhuri

Dube

Gellin

2015. “Vaccine Hesitancy: Definition, Scope and Determinants.” Vaccine 33:4161–64.

31.

Majid

Wasim

Bakshi

Truong

2020. “Knowledge, (Mis-)Conceptions, Risk Perception, and Behavior Change during Pandemics: A Scoping Review of 149 Studies.” Public Understanding of Science 29:777–99.

32.

Maxouris

Vera

2021. “There’s No Need to Delay Second Dose of Covid-19 Vaccine So More Can Get the First, Fauci Says.” CNN, February 4. Retrieved February 23, 2023 (https://www.cnn.com/2021/02/03/health/us-coronavirus-wednesday/index.html).

33.

McKay

P. F.

Blakney

A. K.

Samnuan

Brown

J. C.

Penn

Zhou

Bouton

C. R.

Rogers

Polra

Lin

P. J. C.

Barbosa

Tam

Y. K.

Barclay

W. S.

Shattock

R. J.

2020. “Self-amplifying RNA SARS-CoV-2 Lipid Nanoparticle Vaccine Candidate Induces High Neutralizing Antibody Titers in Mice.” Nature Communications 11:3523.

34.

Mellis

A. M.

Bickel

W. K.

2020. “Mechanical Turk Data Collection in Addiction Research: Utility, Concerns and Best Practices.” Addiction Methods and Techniques 115:1960–68.

35.

Mesch

G. S.

Schwirian

K. P.

2019. “Vaccination Hesitancy: Fear, Trust, and Exposure Expectancy of an Ebola Outbreak.” Heliyon 5(7):1–5.

36.

Miline

Orbell

Sheeran

2002. “Combining Motivational and Volitional Interventions to Promote Exercise Participation: PMT and Implementation Intentions.” Health Psychology 7(2):163–84.

37.

Morgan

F. N.

McCabe

D. B.

Howley

M. J.

McCabe

Steward

M. D.

2010. “The Influence of Different Types of Cues-to-Action on Vaccination Behavior: An Exploratory Study.” Journal of Marketing Theory and Practice 18(2):191–208.

38.

Munro

Lewin

Swart

Volmink

2007. “A Review of Health Behavior Theories: How Useful Are These Developing Interventions to Promote Long-term Medication Adherence for TB and HIV/AIDS?” BNC Public Health 7:104.

39.

Myers

L. B.

Goodman

2011. “Determinants of Adults’ Intention to Vaccinate against Pandemic Swine Flu.” BMC Public Health 11:15.

40.

Norman

Conner

2005. “Predicting and Changing Health Behaviour.” Pp. 324–71 in Predicting Health Behaviour, edited by Mark

Conner

Norman

Paul

. 2nd ed. New York: Open University Press.

41.

Ort

Fahr

2018. “Using Efficacy Cues in Persuasive Health Communication Is More Effective That Employing Threats—An Experimental Study of a Vaccination Intervention against Ebola.” British Journal of Health Psychology 23(3):665–34.

42.

Parsons

J. E.

Newby

K. V.

French

D. P.

2018. “Do Interventions Containing Risk Messages Increase Risk Appraisal and the Subsequent Vaccination Intentions and Uptake—A Systematic Review and Meta-analysis.” British Journal of Health Psychology 23(4):1084–106.

43.

Peer

Vosgerau

Acquisti

2014. “Reputation as a Sufficient Condition for Data Quality on Amazon Mechanical Turk.” Behavior Research Methods 46(4):1023–31.

44.

Podlesek

Rosrosdkar

Komidar

2011. “Some Factors Affecting the Decision on Non-mandatory Vaccination in an Influenza Pandemic: Comparison of Pandemic H1N1 and Seasonal Influenza Vaccination.” Zdravstveno Varstvo 50:227–38.

45.

Randall

Sam

Tartar

Murray

Cannon

2022. “More Than 12.7 Billion Shots Given: Covid-19 Tracker.” Bloomberg, October 6. Retrieved February 23, 2023 (https://www.bloomberg.com/graphics/covid-vaccine-tracker-global-distribution).

46.

Redelings

M. D.

Piron

Smith

L. V.

Chan

Heingzerling

Sanchez

K. M.

Bedair

Ponce

Kuo

2012. “Knowledge, Attitudes, and Beliefs about Seasonal Influenza and H1N1 Vaccinations in a Low-income Public Health Clinic Population.” Vaccine 30(2):454–58.

47.

Rimal

R. N.

Real

2003. “Perceived Risk and Efficacy Beliefs as Motivators of Change: Use of the Risk Perception Attitude (RPA) Framework to Understand Health Behaviors.” Human Communication Research 29:370–99.

48.

Rippetoe

P. A.

Rogers

R. W.

1987. “Effects of Components of PMT on Adaptive and Maladaptive Coping with a Health Threat.” Journal of Personality and Social Psychology 52(3):596–604.

49.

Rogers

R. W.

1975. “A PMT of Fear Appeals and Attitude Change.” Journal of Psychology 91(1):93–114.

50.

Rosenbaum

2021. “Escaping Catch-22—Overcoming Covid Vaccine Hesitancy.” New England Journal of Medicine 384(14):1367–71.

51.

Ruiz

J. B.

Bell

R. A.

2021. “Predictors of Intention to Vaccinate against COVID-19: Results of a Nationwide Survey.” Vaccine 39(7):1080–86.

52.

Sarstedt

Ringle

C. M.

2010. “Treating Unobserved Heterogeneity in PLS Path Modeling: A Comparison of FIMIX-PLS with Different Data Analysis Strategies.” Journal of Applied Statistics 37(8):1299–318.

53.

Scherer

K. R.

2001. “Appraisal Considered as a Process of Multilevel Sequential Checking.” Pp. 92–120 in Appraisal Processes in Emotion: Theory, Methods, Research, edited by K. R

Scherer

Schorr

Johnstone

New York: Oxford University Press.

54.

Schmid

Rauber

Betsch

Lidolt

Denker

M. L.

2017. “Barriers of Influenza Vaccination Intention and Behavior: A Systematic Review of Influenza Vaccine Hesitancy, 2005-2016.” PLoS ONE 12:e0170550. doi:10.1371/journal.pone.0170550.

55.

Schulz

P. J.

Hartung

2021. “Unsusceptible to Social Communication? The Fixture of the Factors Predicting Decisions on Different Vaccinations.” Health Communications 36:1505–13.

56.

Smith

C. A.

Ellsworth

P. C.

1985. “Patterns of Cognitive Appraisal in Emotion.” Journal of Personality and Social Psychology 48(4):813–38.

57.

Tanner

J. F.

Jr. Hunt

J. B.

Eppright

D. R.

1991. “The Protection Motivation Model: A Normative Model of Fear Appeals.” Journal of Marketing 55(3):36–45.

58.

Taschner

N. P.

2021. “Vaccine Hesitancy: Old Story, Same Mistakes.” Journal Health NPEPS 6(2):1–6.

59.

Taylor

2019. The Psychology of Pandemics: Preparing for the Next Global Outbreak of Infectious Disease. Cambridge, England: Cambridge Scholars Publishing.

60.

Taylor

Landry

C. A.

Paluszek

M. M.

Rachor

G. S.

Asmundson

G. J. G.

2020. “Worry, Avoidance and Coping during the COVID-19 Pandemic: A Comprehensive Network Analysis.” Journal of Anxiety Disorders 76:102327.

61.

Teasdale

Yardley

Schlotz

Michie

2012. “The Importance of Coping Appraisal in Behavioral Responses to Pandemic Flu.” British Journal of Health Psychology 17(1):44–59.

62.

Villacorta

Sood

2015. “Determinants of Health Care Provider Recommendations for Influenza Vaccination.” Prevention Medicine Reports 2:355–70.

63.

Weinstein

N. D.

Kwitel

McCaul

K. D.

Magnan

R. E.

Gerrard

Gibbons

F. X.

2007. “Risk Perceptions: Assessment and Relationship to Influenza Vaccination.” Health Psychology 26(2):146–51.

64.

Witte

1994. “Fear Control and Danger Control: A Test of the Extended Parallel Process Model (EPPM).” Communication Monograph 61(2):113–34.

65.

Wong

M. C. S.

Wong

E. L. Y.

Huang

Cheung

A. W. L.

Law

Chong

M. K. C.

R. W. Y.

Lai

C. K. C.

Boon

S. S.

Lau

J. T. F.

Chen

Chan

P. K. S.

2021. “Acceptance of the COVID-19 Vaccine Based Health Belief Model: A Population-based Survey in Hong Kong.” Vaccine 39(17):1148–56.

66.

World Health Organization. 2022. WHO COVID-19 Dashboard. Geneva, Switzerland: World Health Organization. Retrieved February 23, 2023 (https://covid19.who.int/).