Article information

J Health Info Stat. 2025;50(3):292-303

Publication date (electronic) : 2025 August 31

doi : https://doi.org/10.21032/jhis.2025.50.3.292

Kyu Eun Lee, Yunsoo Kim

Professor, Department of Nursing, Catholic Kwandong University, Gangneung, Korea

이규은, 김윤수

가톨릭관동대학교 간호학과 교수

Corresponding author: Yunsoo Kim. 24 Beomil-ro 579beon-gil, Gangneung 25601, Korea Tel: +82-33-649-7614, E-mail: doxapram@naver.com

No potential conflict of interest relevant to this article was reported.

Received 2025 July 17; Accepted 2025 August 27.

Abstract

Trans Abstract

INTRODUCTION

Novel infectious disease is a new disease or disease that has existed since before but is prevalent in different aspects and regions than in the past [1]. In late 2019, the new epidemic was named 2019-nCOVn after the World Health Organization (WHO) reported on pneumonia patients in Wuhan, China [2]. Coronavirus disease 2019 (COVID-19) is an acute respiratory infectious disease that has become a global epidemic, creating a tremendous impact on global health and severely disrupting society and worldwide economics [3]. The ongoing COVID-19 epidemic affected every aspect of daily life, and individuals had to strive to live under quarantine [4]. The global number of confirmed and killed COVID-19 was more than 200 million and 4.2 million, respectively, and more than 200,000 and 2,000 in South Korea [5].

COVID-19 is thought to spread from person-to-person, most commonly through close contact mainly. The virus travels through respiratory droplets released in the air when an infected person coughs, sneezes, or talks [6]. Various strategies for COVID-19 management have been im-plemented globally, focusing on minimizing the transmission and spread of infection and providing protective care for infected individuals [3].

As a result, South Korean citizens are more concerned about health than ever before, and they are trying to follow preventive measures to protect themselves from the solid spreading power of COVID-19. The U.S. Centers for Disease Control and Prevention (CDC) recommends preventive measures such as handwashing, using disinfectants, wearing masks, and social distancing to reduce the risk of spreading the virus, even after one has been fully vaccinated against COVID-19 [7]. The South Korean government also continues to promote disease prevention guide-lines; wearing masks in public, washing hands, 2-meter social distancing, covering mouth and nose when coughing or sneezing, periodic ventilation, avoiding contact with people with fever or respiratory symptoms, refraining from traveling, to prevent COVID-19 [8].

Health Belief Model (HBM) [9] is a widely used theory developed to predict individual health behavior based on behavioral assessments related to perceived risks exposed to them. This model may explain why individuals accept or reject preventive health services or adopt healthy behaviors [9]. The model presently includes the following constructs: perceived susceptibility, perceived severity, perceived benefits, perceived barriers, cue to action, and health motivation [10]. According to HBM, one is more likely to engage in a health behavior when one perceives themselves as susceptible to a disease, when one senses the consequence of infection to be grave, when one perceives health behaviors to be beneficial, or when one perceives health behaviors lowers the degree of im-pairment [11].

Self-efficacy means personal effectiveness in using personal health beliefs and behaviors to achieve health goals in daily life. People high in self-efficacy generally maintain the belief that, through their effects, they can achieve desired health results in an effective manner [12].

Infection is associated with public health-related stigma [13]. Stigma-tized persons are subject to active avoidance from society and have a social and psychological consensus that they may be treated unfriendly [14]. In COVID-19, the higher the level of social stigma awareness, the more likely the people are to take preventive measures to escape negative perception [15].

There have been researching on HBM and COVID-19 in Korea: a study of preventive behavior for the general public [16] and a study on preventive behaviors of restaurant diners [17]. Prior studies related to HBM and COVID-19 have not been made to make outpatient appoint-ments due to COVID-19 [18]. There was a study of the correlation between Knowledge and HBM variables for nursing students [12], a study of vaccinations for the public [19], an effect of an educational interaction study for nursing students [20], and a comparison study of vaccine acceptance between COVID-19 and seasonal influenza [3].

This study was conducted using HBM, which has accumulated useful empirical results in studies focusing on early detection and preventive behaviors of diseases in the healthcare field. The health beliefs, self-effi-cacy, and preventive actions associated with COVID-19 are critically important to reducing the mortality rate and maintaining health and qual-ity of life [12]. Therefore, in the long-term proliferation of COVID-19, the purpose of this study was to establish a hypothetical model of the factors influencing preventive behaviors of South Korean COVID-19 based on Becker's health belief model to identify the relationships among these factors, and then finally to test the model fit. This research is expected to serve as the basis for building COVID-19 prevention strategies to help prevent the further spread of COVID-19 and end it early.

This study was based on Becker's HBM extracted conceptual framework for use in research analysis. The concept of configuring a health belief model is a perceived susceptibility, perceived severity, health motivation, perceived benefits, perceived barriers, cues to action, and self-effi-cacy. Originally, Becker's model had cues to action. Still, because the prior study determined that stigma was associated with infectious diseases and affected health behaviors, the cues to action was removed. The stigma was considered as a variable of COVID-19.

Accordingly, the hypothetical model consists of two exogenous variables and seven endogenous variables describing the perceived susceptibility, perceived severity, health motivation, perceived benefits, perceived barriers, stigma, self-efficacy, preventive behaviors for COVID-19. A hypothetical path constructed on this basis is shown in Figure 1.

Figure 1.

Structural equation model of COVID-19 preventive behaviors. (A) Hypothetical structural equation model of COVID-19 preventive behaviors. (B) Modified structural equation model of COVID-19 preventive behaviors. COVID-19, Coronavirus diseases-2019.

METHODS

Setting and sample

The participants of this study were 322 adults aged 20 years and older residing across Korea. The administrative districts were divided into the metropolitan area, including Seoul, and five other regions (Chungcheong, Gangwon, Gyeongsang, Jeolla, and Jeju), and sampling was conducted for each area. The demographic composition of the study population was based on the 2020 Population Census of the South Korea, ensuring that the distribution of participants reflected the national population structure. Participants were not sampled from a complete enumeration list; rather, they were recruited through online communities using a snow-ball sampling method. Details of the extraction rate and regional distribution of participants are presented in Table 1.

Table 1.

Calculating the number of subject samples

The number of participants in this study was calculated using G*Power 3.1.9 (Heinrich-Heine-Universität Düsseldorf, Germany). In the multiple regression analysis, the required sample number was calculated based on a significance level of 0.05, 0.95 power, effect size 0.15 (medium effect size), and 19 predictors were calculated with 217.

Online questionnaires were assigned to the population by region following the 2018 Population Statistics of Korea based on 330 people because of the return rate of 20%. The surveys were conducted online, with 322 surveys were used in the analysis, and incorrect responses and missing values were excluded. Evidence suggests that the proper sample size is around 150-400 based on a classification of 5 factors that may influence the sample size [19,20], a sample size of 330 participants was deemed reasonable for this study.

Online surveys were conducted with people who understood the purpose and methods of the study and agreed to participate. Exclusion cri-teria for the study participants are as follows:

• - Those who did not agree to participate in the study.

• - Those who experienced infections with COVID-19.

• - Those with severe illnesses (e.g., cancer, stroke).

RESULTS

General characteristics of participants

Finally, the data from 322 participants were analyzed in this study, evenly assigned to each region (Table 1). The general characteristics of the study participants are shown in Table 2. The average age of the participants was 45.4 years, with 62.1% of women. College graduates com-posed 45.3% of the sample population, 40.7% had no religion, and the monthly household income was the highest at KRW 5,000,000 to KRW 8,000,000. In terms of occupation, 24.1% were unavailable, 21.7% were professionals and related workers, and 20.2% were office workers. The mean score for perceived health and perceived health compared to a year ago were 3.77 and 3.52, respectively.

Table 2.

General characteristics of participants (n=322)

Descriptive statistics and confirmatory factor analysis

As shown in Table 3, the skewness (-1.33 to 0.55) and kurtosis (-0.99 to 0.27) of the single measurement variables satisfied the assumptions of univariate normality. Confirmatory factor analysis was conducted to evaluate the validity of the factors. In this analysis, standardized factor loadings (λ) were examined to determine the contribution of each item to its latent construct. A loading of 0.50 or higher is generally considered the minimum acceptable threshold to indicate that an item adequately represents the underlying construct; therefore, three items with λ values below 0.50 were excluded from further analysis. For the remaining items, composite reliability (CR), which reflects the internal consistency of the construct, was calculated. A CR value above 0.80 is typically regarded as evidence of satisfactory construct validity. In addition, the average vari-ance extracted (AVE) exceeded 0.50 for all constructs, indicating no issues with convergent validity.

Table 3.

Descriptive statistics of observed variables (n=322)

The correlations between the correlation matrix and the AVE were examined to verify discriminant validity. Specifically, the AVE and multi-correlation coefficients of the key variables showed that the value of the multi-correlation coefficient was appropriately small to ensure the facto-rial discriminant validity. This result is because the correlation coefficients ranged between 0.07 and 0.57, with absolute values less than 0.85.

Goodness of fit testing of the model

Table 4 presents the model fit of the hypothetical and modified models, summarizes the modified model, and details the direct, indirect, and total effects of the model.

Table 4.

Direct, indirect, and total effects of the structural equation model of the South Koreans’ COVID-19 preventive behaviors (n=322)

Initially, it appeared that the hypothetical model would be rejected, as χ² =356.64, Normed χ² =3.43 (p <0.001), χ² =296.61 (p <0.001), RMR, RMSEA values were not a good fit. Thus, the personal attributes of the affecting factors on COVID-19 preventive behaviors were removed and a modified model was set, which included perceived susceptibility, perceived severity, health motivation, perceived benefits, perceived barriers, stigma, and self-efficacy. The following values were obtained from goodness of fit testing of this model: χ² =252.07, Normed χ² =2.86 (p <0.001), GFI=0.91, AGFI=0.85, CFI=0.90, RMR=0.05, TLI=0.86, NFI=0.86, RMSEA=0.07. Therefore, the goodness of fit of the modified model is acceptable.

DISCUSSION

Based on Becker's HBM, this study developed a hypothetical model describing preventive behaviors for COVID-19 through a structural equation model within a theoretical framework of factors that influence the COVID-19 preventive behaviors.

Variables that affect COVID-19 preventive behaviors are self-efficacy, perceived benefits, and perceived susceptibility, which accounted for 53.7% of the overall preventive behaviors. Self-efficacy directly impacted preventive behaviors against COVID-19 with the highest critical factor. This result is consistent with Tao et al. [3] study on behavioral intention and the HBM, where self-efficacy was seen as the most strongly associated variable. Self-efficacy is important in the performance of healthy behaviors because it is essential in overcoming all the barriers they may confront when trying to reach their own health goals [13]. Furthermore, it is the prerequisite and most critical factor in one's changing from the consideration phase to the behavioral decision-making phase during behavioral change [21]. Self-efficacy is a cornerstone of health-seeking behavior that is very important not only in preventive behavior of infectious diseases but also in successful managing the chronic diseases such as hyper-tension, diabetes, osteoarthritis [22].

Perceived benefits and susceptibility act an indirect effect on preventive behaviors with self-efficacy. It implies that the more people perceive many benefits of COVID-19 preventive behaviors, the greater the risk of getting COVID-19. Thus, the more people have self-efficacy, and the more COVID-19 prevention behaviors are performed. In addition, this result implies that it takes confidence to perform such preventive behaviors.

Although previous studies and meta-analyses of the HBM have con-sistently shown that perceived benefits and barriers were the strongest predictors of preventive health behaviors, while perceived susceptibility and severity had limited impact [23], the findings of this study differed in several respects. Specifically, this study demonstrated that perceived susceptibility was significantly associated with stigma, while perceived severity and perceived benefits had a direct influence on both self-effica-cy and preventive behaviors.

These differences may be attributed to several factors. First, the context of COVID-19 as an emerging infectious disease with high uncer-tainty and social fear may have heightened individuals’ perceptions of severity, thereby strengthening its role in shaping self-efficacy and behavior. Similar to the findings of Kim [24], who reported that self-effica-cy was the primary predictor of preventive intentions, our results emphasize the importance of self-efficacy but also highlight the additional influence of perceived severity and benefits. Second, the sociocultural environment, including heightened public discourse on infection, stigma, and preventive practices, may have influenced how susceptibility was linked with social stigma rather than with preventive behavior itself, consistent with previous research noting the role of stigma in health-related behaviors [25]. Third, variations in study populations, measurement tools, and timing of data collection could have contributed to the discrepancies with previous findings [23].

In this study, perceived benefits and susceptibility affected preventive behaviors, while the perceived severity and barriers did not affect these behaviors. This result is presumed to be due to the sporadic trend of COVID-19 outbreaks in Korea. Local residents with an increase in CO-VID-19 infection are more anxious than those who do not because the infection route is often not clearly identified [26]. In addition, during the data collection in this study, COVID-19 infections occurred only in one region (south-east region of Korea). Relatively, people outside of that region did not feel the seriousness.

In South Korea, the cost for COVID-19 testing was free for people suspected of being infected, resulting in fewer perceived barriers caused by the economic burden. or discomfort associated with testing. According to the HBM, the higher the perceived barriers, the more likely preventive behaviors are performed. In our study, it is assumed that the participants did not have significant perceived barriers, which did not affect preventive behaviors for COVID-19. Hence, A well-designed, planned studies in the different regions under pandemic situations are necessary to determine the relevance between perceived severity, perceived barriers, and preventive behaviors. Research on COVID-19 prevention behavior and affecting factors should continue for policies on preventive behavior of infectious diseases.

The perceived severity in this study did not affect COVID-19 preventive behaviors. Contrastingly, Tao et al. [3] study showed that differences in vaccination rates based on the severity of the illness were different. Based on previous studies, it is expected that perceived severity may be related to the high mortality rate, the speed of spread of infectious diseases, and the reproduction rate of infectious diseases. Our result may be the reason why this study was conducted in the early of the COVID-19 outbreak in Korea, that is, the early pandemic in which many people lacked understanding and awareness of COVID-19 infection. Further research with different timing is needed depending on the spread of infection. However, as a result of this study itself, we suggest that perceived severity does not significantly affect infection prevention behavior because of the lack of awareness of the general public in the early stages of infectious disease pandemics.

Health motivation averaged at 3.55 out of 5 points and did not affect preventive behaviors. According to Conner and Norman [27], health motivations possess an inclusive definition and can be used as various modified psychological variables depending on the different populations, behavior patterns, and contexts. For this reason, health motivation is in-terpreted in a variety of meanings depending on the subject can lead to meaningless consequences in preventive behaviors. Health motivation is a more influential health-seeking behavior factor in chronic diseases than in acute infectious diseases such as COVID-19. Of course, we will have to build up knowledge through further research on this, as there is no historical case of a long-lasting infectious disease beyond expectations by creating new variations over time, such as the current COVID-19 pan-demic.

This study found that demographic variables barely directly or indi-rectly affected South Koreans’ COVID-19 preventive health behaviors. In addition, the explanatory power of the modified model without demographic characteristics was higher than that of the hypothetical model with demographic variables. However, the relevance between age, health status, gender, and preventive behaviors of COVID-19 in Souliotis et al. [28] study and the relevance between age and vaccination rates in Tao et al. [3] study differed from this study. Future studies need to identify the relevance between sociodemographic variables and COVID-19 preventive behaviors, develop educational programs appropriate for target groups, and use them for infectious disease education.

Health stigma averaged 2.82 out of 5 points and did not affect preventive behaviors. These results contrast with stigma in the COVID-19 pan-demic influenced by preventive behavior intentions [16]. In unusual and unpredictable COVID-19 pandemic situation, health stigma is sometimes an individual problem and sometimes a national stigma. As mentioned above, further studies at different times are helpful to find out more about the relationship between health stigma and infection prevention behavior through.

The model results revealed a significant pathway from stigma to self-efficacy and, subsequently, to preventive behavior, indicating a positive association. While stigma has traditionally been described as a barrier to health-seeking behaviors— such as in cases of leprosy, epilepsy, mental illness, cancer, HIV, and obesity [17]— our findings suggest a more com-plex role of stigma in the context of COVID-19. Specifically, perceived severity was positively related to stigma, meaning that the more seriously individuals perceived COVID-19, the more they tended to view infected persons as deviant or harmful. This negative social perception may con-tribute to exclusionary attitudes; however, it also appears to stimulate self-protective motivations, thereby increasing individuals’ confidence in adopting preventive behaviors.

In this regard, stigma functions not only as a potential barrier but also as a driver of self-efficacy in preventive contexts. This dual role under-scores the need for nuanced interpretations: while stigmatization can harm infected individuals through social exclusion, it may simultane-ously strengthen preventive practices among the uninfected population by heightening their sense of self-efficacy. Therefore, public campaigns should aim to emphasize accurate information about causes, transmission, and preventive behaviors in order to reduce harmful stigmatization of infected persons. At the same time, sustained social attention and support are necessary to address anxiety about reinfection and to prevent post-traumatic stress disorders rooted in fear of stigma among CO-VID-19 survivors [29].

This study extends prior research by integrating stigma into the Health Belief Model. Unlike Park and Oh [30], who found that adolescents’ preventive behaviors were largely driven by intention, and Baek et al. [31], who emphasized susceptibility, self-efficacy, and information trust among adults, our findings revealed a different pathway. Perceived susceptibility was linked to stigma, but stigma did not predict self-efficacy or behavior. Instead, perceived severity and benefits significantly influenced both self-efficacy and preventive behaviors, with self-efficacy showing the strongest effect. These results highlight the importance of strengthening self-efficacy while framing messages to emphasize severity and benefits without reinforcing stigma.

In summary, perceived severity affects stigma, and perceived susceptibility and perceived benefits affect self-efficacy. Thus, perceived susceptibility, perceived benefits, self-efficacy greatly affect the COVID-19 preventive behaviors of South Koreans. This study is meaningful because it pro-vides the primary data necessary for COVID-19 prevention policymaking by identifying such factors. In particular, as self-efficacy is seen to have the most substantial influence on the prevention of COVID-19, more ways to encourage individuals to have confidence and reach their goals in preventing COVID-19 contraction and spread are needed. It is also recom-mended to include stigma as a construct in the HBM because it address-es the social variable of diseases and is based on individual beliefs.

CONCLUSION

This study confirmed that the modified HBM created in this study is a suitable model for describing the COVID-19 preventive behaviors of South Koreans. Following Becker's HBM, this study identified causal relationships among factors influencing South Koreans’ preventive behavior for COVID-19, established a hypothetical model, and tested the hypothesis and the suitability of a structural equation model. The study proposes further research into increasing the explaining power of COV-ID-19 preventive behaviors by incorporating other theories to the HBM.

However, the study was limited to data collection from those with online access. Furthermore, it did not reflect the local and timely change of COVID-19. Thus, the reliability of some survey questions was low. It will be necessary to study the shift of the COVID-19 pandemic through merged online and offline data collection and regional and periodic surveys.

Article information Continued