1. Introduction
Online health services are a type of service that is built on a telemedicine system in which doctors provide online services through the mobile Internet to achieve cross-territory diagnosis and treatment without the restrictions of time and place [
1]. In the context of the novel coronavirus epidemic, many offline medical services find it difficult to operate normally, while there is a temporary shortage of medical resources in some areas where there is a severe epidemic, and online health communities (OHCs) have become an important way for patients to receive an initial consultation and relevant health information support [
2]. Health information is considered the “key to quality care”, and users’ health information search behavior can help them make better care decisions and prevent unhealthy behavior promptly [
3]. In recent years, with the development of social networks and increasing citizen health awareness, OHCs have played an important role in users’ healthcare knowledge-seeking and disease management experience sharing [
4]. For example, PatientsLikeMe, WebMD, and MedHelp provide an Internet-based platform that includes not only patients but also doctors [
5], where community members can share health knowledge through interactive modules such as private messages or discussion forums. OHCs provide an open platform for users to access medical resources and share knowledge, experiences, and emotions, which provide effective social support for users, encourage them to prevent diseases in advance, and relieve patients’ stress and anxiety [
6]. Online health services can improve the quality of care by maintaining contact with patients before clinical examinations, allowing patients to compare the services of different doctors, thus increasing patient satisfaction [
7]. Therefore, it is of great practical significance to further explore the inner mechanism of doctor-patient information exchange in health communities both from a macroscopic perspective for the harmonious coexistence of doctor–patient relationships and for the effective utilization of medical resources from a microscopic perspective for the satisfaction of both doctors and patients in online health community information exchange.
Extensive research has been conducted on various antecedents that influence knowledge-sharing among online community users [
8,
9,
10,
11]. However, little research has been conducted to investigate the influence of different personality traits on the users’ knowledge-sharing intentions in online communities from the perspective of their personality traits. At the root, personality traits drive individuals’ complex behaviors; individual motivation and social capital are related to personality traits, and the effective transfer of knowledge through the behaviors of individuals with different personalities affects their judgment of whether to adopt knowledge-sharing [
12]. Second, although we have learned a lot about the antecedents of users’ knowledge-sharing intention in online health communities, these factors revolve around a single user’s subjective feelings, such as a sense of self-worth, members’ perceptions of social support and reputation enhancement [
13], users’ attitudes [
14], and the lack of a systematic mechanistic framework. Finally, previous studies on knowledge-sharing in online health communities have focused on patients or physicians [
15], and fewer studies have combined physicians and patients for comparative analysis. To fill the above research gaps, based on the personality trait theory, this paper explores the mechanisms of different personality traits on users’ knowledge-sharing intentions in online health communities. At the same time, this paper considers the network relationship benefits of social capital, introduces social capital theory into the study of the mechanisms affecting the users’ knowledge-sharing intention in the field of online health communities, constructs a theoretical model of how individuals with different personality traits in online health communities stimulate their knowledge-sharing intention through different social capital and try to elucidate how users with different personality traits stimulate their knowledge-sharing intention in the process of sharing information on online health communities in the structural dimension and relational dimension. The relationship between structural and relational capital on the knowledge-sharing intention of online health communities is examined. Finally, this paper examines the differences in the knowledge-sharing willingness between doctors and patients to bridge the gap between these two types of user differences.
Based on the personality trait theory and social capital theory, this paper establishes a new framework on the influence path of users’ knowledge-sharing intention in online health communities, which takes the users’ introversion-extraversion personality traits as independent variables, their knowledge-sharing intention as dependent variables, and structural dimensional capital-interaction, relational dimensional capital in social capital theory-trust, and reciprocity as mediating variables, to verify the influence of users with different personality traits on their knowledge-sharing willingness and the intermediate mechanism of action when opening the black box between users’ different personality traits and their knowledge-sharing willingness in online health platforms. At the same time, this paper introduces social capital theory into the field of online health community information sharing and further validates the explanatory logic of the social capital theory. Therefore, this paper has important theoretical and practical significance for enriching the research on knowledge-sharing in online health communities, grasping the mechanisms of action and factors that affect the knowledge-sharing of online health community users, stimulating users’ knowledge-sharing intentions, and helping to promote online health community operators to clarify their operation mechanisms.
This paper has the following innovative points: first, this paper introduces social capital theory into the field of online health community users’ knowledge-sharing research, explains the factors affecting the online health community users’ knowledge-sharing willingness from the perspective of social capital theory, and establishes a systematic mechanistic framework. Second, this paper integrates and compares two types of users (doctors and patients) in online health communities and examines the differences in the knowledge-sharing willingness between doctors and patients with different personality traits, bridging the gap between the two types of users in online health communities, which is rarely studied in concert. Third, based on the personality trait theory, this paper considers the users of online health communities as individuals with both introversion-extraversion personality traits and explores the influence of different personality traits on knowledge-sharing willingness from the root, enriching the research on knowledge-sharing in online health communities.
4. Data Analysis
4.1. Reliability and Validity
The data analysis followed a two-stage approach: a measurement model and a structural model. First, the validity and reliability of the six elements of the measurement model (i.e., inside-out personality traits, interaction, trust, reciprocity, and knowledge-sharing intention) were assessed, followed by a validating factor analysis through structural equations, which focused on examining the structural relationships between underlying variables.
Based on Fornell et al.’s two-step approach, this study examined the internal validity and reliability of the constructs measuring the model [
61]. First, the internal consistency of the constructs was assessed using two measures.
Table 2 shows the results for the validated factor analysis of the measurement model, with Cronbach’s alpha for each of the resulting constructs appearing above 0.796, exceeding the recommended 0.7. Therefore, the reliability of the constructs in this study was good.
Second, the convergent validity and discriminant validity of the constructs were tested using two measures.
Table 2 shows that the AVEs of the measured variables were all exceedingly higher than 0.571, which exceeded the recommended value of 0.5. Therefore, this study has good convergent validity. We verified the discriminant validity by testing whether the correlation between the constructs was less than the square root of the AVE. In
Table 3, the main diagonal value is the square root of the AVE, and the non-main diagonal is the correlation coefficient between the constructs, with all diagonal values exceeding the correlation between any pair of constructs. This result indicates that the measurement model also had sufficient discriminant validity.
4.2. Confirmatory Factor Analysis
To further test the discriminant validity of the variable measures, confirmatory factor analysis was conducted using AMOS 22 software for INT, EXT, INA, TRU, REC, and KSI. The fit of the six-factor model for the physician group (χ2/df = 1.338, RMSEA = 0.037, GFI = 0.917, IFI = 0.980, CFI = 0.980, NFI = 0.926) was significantly better than the other models, and the fit of the six-factor model for the patient group (χ2/df = 1.263, RMSEA = 0.028, GFI = 0.942, IFI = 0.988, CFI = 0.987, NFI = 0.943) was also significantly better than the other models, which suggests that the variables have good discriminant validity.
4.3. Common Method Variance Test
To avoid the problem of common method bias from influencing the findings of this study, a common method bias was controlled and tested in terms of both procedural and statistical methods. In terms of procedural design, the study draws on the following measures: developing clear and concise questions, using anonymous questionnaires for collection, and varying how variables are obtained and measured to minimize respondent guesswork about the purpose of the measurement. For statistical testing, this study utilized Harman’s one-way test to verify the extent of the homogeneity error, which is an exploratory factor analysis of the full set of constructs, and if the variance explained by the first factor exceeded 50%, this indicated a high common method bias in the data. The results of this study, calculated using SPSS 22.0, showed that the variance explained by the first factor in the healthy physician group was 40.05%, and the variance explained by the first factor in the patient group was 23.07%, with both less than 50 %. Therefore, there was no serious problem of common method bias in the results of this study.
4.4. Hypotheses Testing and Multi-Group Analysis
4.4.1. Direct Effects Test
In this paper, we used AMOS to test the path coefficients of the hypothesized model for the effect of the introversion-extraversion personality traits of doctors and patients on their knowledge-sharing intention, and the results are shown in
Table 4. Among them, the positive effects of introversion personality traits on interactions (β = 0.089,
p = 0.174) and knowledge-sharing intentions (β = 0.004,
p = 0.952) were not significant, and the positive effect of trust on knowledge-sharing intentions (β = 0.065,
p = 0.274) was not significant in the patient group, while the rest of the paths were significant. Thus H1c, H2c, H3a, H5a, H5c, and H6c are supported and H1a, H2a, and H6a are partially supported (physician part).
4.4.2. Indirect Effect Test
To further analyze the mediating role of interaction, trust, and reciprocity between the introversion-extraversion personality traits and knowledge-sharing intention, this paper used the Bootstrap method to test the mediating effect, and if the 95% confidence interval did not contain zero, the results were statistically significant, indicating the existence of the mediating effect. The results are shown in
Table 5. The results of the mediating effect test, both in the patient group and the physician group, showed that the mediation effects of the interaction between extraversion personality traits and knowledge-sharing intention and the mediation effects of reciprocity between interaction and knowledge-sharing intentions did not include zero. Therefore, interactions mediate the relationship between extraversion personality traits and knowledge-sharing intentions, and reciprocity mediates the relationship between interaction and knowledge-sharing intentions. H4b and H7b are supported. In the physician group, the mediation effects of the interaction between introversion personality traits and knowledge-sharing intentions and the mediation effects of trust between interaction and knowledge-sharing intentions did not include zero. However, in the patient group, the mediation effects of the interaction between introversion personality traits and knowledge-sharing intention and the mediation effects of trust between interaction and knowledge-sharing intention included zero. Therefore, interactions mediated between extraversion personality traits and knowledge-sharing intentions, and trust mediated between interactions and knowledge-sharing intentions in the physician group, whereas interactions did not mediate between extraversion personality traits and knowledge-sharing intentions, and trust did not mediate between interactions and knowledge-sharing intentions in the patient group; therefore, H4a and H7a were partially supported (physician part).
4.4.3. Multi-Group Analysis
To compare the pathway differences between physicians and patients, a multi-group structural equation model analysis using AMOS was conducted in this paper. The results of the multi-group analysis showed (see
Table 3) that the standardized path coefficient differences between physicians’ and patients’ introversion personality traits and interaction on knowledge-sharing intentions were −0.096 and −0.103, respectively, with p-values of 0.159 and 0.138. It indicated that the differences between the physicians’ and patients’ introversion personality traits and interaction of knowledge-sharing intentions were not significant, and the differences in the remaining seven direct effects were all significant. H1b, H2b, H2d, H5b, H5d, H6b, and H6d were supported and H1d and H3b failed to be supported.
5. Discussion
The novel coronavirus epidemic has caused a global shortage of offline medical resources, and people are increasingly inclined to pre-search health information from the Internet. Research related to online health communities has attracted the attention of many scholars. However, systematic research on users’ knowledge-sharing willingness in health communities and comparative research on different users are still inadequate, so this paper uses well-known Chinese online health communities as a data source, takes the knowledge-sharing intention of different users in online health communities as a research object, and analyzes the proposed hypothesis model. The results are as follows.
5.1. Theoretical Implications
From the perspectives of personality trait theory and social capital theory, this study expands the research on knowledge-sharing in OHCs by comparing the knowledge-sharing paths of two types of key users [
47]. From the perspective of the introverted personality traits of users in OHCs, this paper examines the issues of personality traits and knowledge-sharing in OHCs that have hardly been investigated in previous studies [
13,
62], enriching the understanding of knowledge-sharing.
This study uses social capital theory to explain how doctors and patients with two personality traits can better stimulate their knowledge-sharing in OHCs. Interaction and reciprocity are the driving factors that explain users’ participation in knowledge-sharing in online communities [
54], and trust is a key factor that causes patients to reduce their knowledge-sharing intention [
63]. Social capital is the mechanism that facilitates individual collaboration in OHCs: a concept that explains an individual’s potential or real capital with friends or strangers [
20]. Members in a social network can obtain different resources according to their positions in the social relationship structure and can also invest in other resources, expecting to obtain future benefits [
21]. The social capital theory can also explain the expectations and reality of users in the health community [
36]. Patients acquire health knowledge and share experiences in communication with doctors or other patients. Doctors guide patients in scientific treatment and cultivate doctor–patient relationships [
15,
24]. The desire to continue to profit in the community has stimulated the knowledge-sharing intention of both types of users.
This paper develops a study on knowledge-sharing between physicians and patients by building a model of comparative mechanisms. The study found that doctors with introverted personality traits were more willing to share knowledge than patients, and doctors with both introversion and extraversion personality traits were more willing to participate in community interactions. For doctors, interactions can generate more trust than patients [
57], and further, trust can stimulate more knowledge-sharing than patients [
46,
50,
56]. Additionally, for patients, more reciprocity results in more knowledge-sharing than for doctors.
5.2. Practical Implications
OHCs should divide their users into doctors and patients and should conduct differentiated management according to the interests and needs of the two parties.
OHCs should set incentives for active users, especially for patients. To encourage highly introverted patients to participate in interactions and to share their doubts and experiences, OHCs can use the platform database to observe the number and content of posts, can frequently ask patients to provide more information on healthcare or disease prevention, and can provide patients with emotional support to create an atmosphere of emotional exchange.
OHCs should create new sections to provide interaction and communication opportunities for introverted patients and establish a mutually beneficial mechanism that is more focused on the needs of patients. Meanwhile, OHCs should reward doctors for their attitude and behavior.
OHCs could establish reputation-scoring systems that allow patients to evaluate services after seeking treatment. Then, the OHCs could archive and store electronic medical records, establishing a database for patients to prompt patients to undergo their next treatment, thus enhancing the patient’s trust in the platform and promoting the sustainable development of OHCs.
OHCs should also establish trust and reciprocal emotional channels to create a warm emotional atmosphere. In such an atmosphere, patients are more likely to communicate more with doctors to obtain more targeted treatment. Meanwhile, OHCs should establish a mutually beneficial mechanism. When providing services, it should be more targeted to the needs of patients and enhance patients’ benefits and rewards in many respects.
5.3. Limitations and Future Research
This study has some limitations. First, the sample size is not large. This study only collected data from three OHCs, and the sample size has not been able to represent all users of the OHCs. In the later stage, the sample size needs to be increased to further expand the collection scope. Second, this study only investigated introversion-extraversion personality traits, and whether other personalities affect the knowledge-sharing of these two types of users in the OHCs still needs to be explored. Third, this study only incorporated the structural dimension and relational dimension of social capital into the model and did not examine the role of the cognitive dimension. Future research should consider the impact of the cognitive dimension of capital. Fourth, this study uses interactions to represent the structural dimension and trust and reciprocity to represent the relational dimension. However, there are other factors, such as network centrality, comments, etc., that can better enrich the explanation of each dimension of social capital.
6. Conclusions
This study explains the mechanisms by which both physicians’ and patients’ introversion-extraversion personality traits influence their knowledge-sharing intentions. First, the effect of patients’ introverted personality traits on knowledge-sharing intentions and interactions is insignificant, while physicians’ introverted personality traits positively influence their interaction and knowledge-sharing intention in online health communities. Meanwhile, compared to patients, doctors’ introversion personality traits positively influence interaction and knowledge-sharing intentions more strongly, which is related to doctors’ professional ethics and emphasis on professional altruism.
Second, extraversion personality traits positively influence both physicians’ and patients’ interaction and knowledge-sharing intentions in online health communities. Additionally, compared to patients, the positive effect of extraversion personality traits on interactions is stronger for physicians, but the difference in the effect on the knowledge-sharing intention was not significant.
Third, interaction in online health communities positively affects doctor-patient trust and reciprocity. The more interaction there is between doctors and patients and between patients and patients, the closer the relationship is and the more likely it is to generate a sense of trust and reciprocity. Meanwhile, physician interaction and reciprocity positively affect their knowledge-sharing intention. Patients’ sense of reciprocity positively influences their knowledge-sharing intention, while the effect of interaction on their knowledge-sharing intention is not significant. In addition, among these four path relationships, physicians have a stronger influence relationship compared to patients, which is related to the dominance of physicians in the knowledge sharing of online health communities.
Fourth, interactions mediate between extraversion personality traits and knowledge-sharing intentions, and reciprocity mediates between interaction and knowledge-sharing intentions. This suggests that extroverted doctors and patients are more willing to share knowledge and information by communicating with others on online health platforms, and that information sharing between doctors and patients is based on the premise of mutual benefit and reciprocity.