This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR Medical Informatics, is properly cited. The complete bibliographic information, a link to the original publication on http://medinform.jmir.org/, as well as this copyright and license information must be included.
Social question-and-answer communities play an increasingly important role in the dissemination of health information. It is important to identify influencing factors of user willingness to share health information to improve public health literacy.
This study explored influencing factors of social question-and-answer community users who share health information to provide reference for the construction of a high-quality health information sharing community.
A cross-sectional study was conducted through snowball sampling of 185 participants who are Zhihu users in China. A structural equation analysis was used to verify the interaction and influence of the strength between variables in the model. Hierarchical regression was also used to test the mediating effect in the model.
Altruism (β=.264, P<.001), intrinsic reward (β=.260, P=.03), self-efficacy (β=.468, P<.001), and community influence (β=.277, P=.003) had a positive effect on users’ willingness to share health information (WSHI). By contrast, extrinsic reward (β=−0.351, P<.001) had a negative effect. Self-efficacy also had a mediating effect (β=.147, 29.15%, 0.147/0.505) between community influence and WSHI.
The findings suggest that users’ WSHI is influenced by many factors including altruism, self-efficacy, community influence, and intrinsic reward. Improving the social atmosphere of the platform is an effective method of encouraging users to share health information.
Social question-and-answer (Q&A) communities collect a large amount of high-quality health information based on the informal and collaborative method of information generation. Therefore, they have become an important means for the public to obtain health information. They also play an increasingly important role in promoting public health literacy. Zhihu is one of the most representative Q&A communities in China. In 2019, Zhihu had over 220 million registered users, over 28 million questions, and 130 million answers. On this platform, 750,000 questions were health-related, and nearly 21 million people followed the topic of health. In the Healthy China 2030 plan, the Chinese government requires news media to strengthen the publicity of health science knowledge. Moreover, news media is required to actively use social networks for health education. Therefore, exploring the influencing factors of users’ willingness to share health information (WSHI) based on the Q&A community is necessary and meaningful.
Users in a social Q&A community gather considerable amounts of high-quality health information through the sharing mechanism of question-answer-feedback. This topic has become one of the hot spots in medical informatics research to encourage more users to participate in producing health information. Empirically, Zhao et al [
As public platforms, social Q&A communities were established on social networking sites (SNSs) for internet users to seek and share knowledge, experiences, and other information [
Thus far, the concept of WSHI has no unified definition. According to the self-determination theory proposed by Ryan et al [
The aim of this study is to establish a user WSHI model based on the social Q&A community environment. The study also seeks to explore factors that influence the sharing of health information among such users. There are many classical models in the area of research on health information sharing such as the social cognition theory [
Altruism is usually understood as an individual’s behavior of offering help to others at the expense of their own interests [
H1: Altruism positively affects WSHI.
Intrinsic and extrinsic rewards are two of the most important concepts in social exchange theory [
H2: Intrinsic reward positively affects WSHI.
H3: Extrinsic reward positively affects WSHI.
Self-efficacy refers to the subjective judgment of whether an individual can successfully implement a certain behavior and achieve the expected results in a specific environment and state [
H4: Self-efficacy positively affects WSHI.
The characteristics of the community platform, such as design (Q&A format, agree/disagree mechanism, the
H5a: Community influence positively affects user self-efficacy.
H5b: Community influence positively affects WSHI.
In this study, Zhihu users with a history of health information sharing were selected as the research objects. The sampling started with students at a medical university and extended by snowball sampling of an online questionnaire using WeChat and other message tools. This was done to maximize the population representation.
Zhihu users are mainly concentrated in the high knowledge-level population or people with a higher level educational background. Additionally, medical university students have high health literacy and will be the main health information disseminators in the future. Therefore, selecting medical university students as the starting point of snowball sampling is meaningful and necessary.
This study explores the influencing factors of WSHI within the environment of a social Q&A community based on the interpretation of the above variables and related assumptions.
Willingness to share health information path model of the social question and answer community users.
The observation indexes of the 6 variables in the model were screened according to the literature. A small-scale presurvey was conducted. Based on the results, the observation indexes of variables increased or decreased. Experts were asked to determine the questionnaire content in the form of a 5-point Likert scale.
Variables, indexes, and index descriptions in the model.
Variable, reference, and index description | |
|
|
|
AL1: I like to share my health information with other users on Zhihu. |
|
AL2: I think sharing health information on Zhihu can help others. |
|
AL3: I enjoy the process of helping others by sharing health knowledge on Zhihu. |
|
AL4: In my opinion, sharing health information on Zhihu is a manifestation of one’s social value. |
|
|
|
IR1: I think by sharing health information on Zhihu, we can gain others’ respect. |
|
IR2: I think by sharing health information on Zhihu, I can gain praise and recognition from others. |
|
IR3: In my opinion, sharing health information on Zhihu can help me gain a more positive and confident attitude toward life. |
|
|
|
ER1: I think sharing health information can result in more followers. |
|
ER2: I think sharing health information on Zhihu can bring money or other material benefits. |
|
|
|
CI1: I think the Zhihu platform has high credibility in solving health problems. |
|
CI2: I think Zhihu is an important platform for me to obtain health information. |
|
CI3: I think the public image of Zhihu can promote users to share health information. |
|
CI4: I think Zhihu has certain security measures for sharers and the information they share. |
|
CI5: In my opinion, the platform design of Zhihu (question-and-answer format, agree/disagree mechanism, the |
|
|
|
SE1: I believe that the health information I have released on Zhihu is scientific and accurate. |
|
SE2: I can express my opinions on a topic on Zhihu with confidence. |
|
SE3: I can share new ideas and concepts about health information with others on Zhihu. |
|
SE4: I can provide rich content in other aspects for a certain health problem on Zhihu. |
|
SE5: I can accurately address the relevant issues and discuss them on Zhihu. |
|
|
|
WSHI1: I am willing to share the health information I know on Zhihu. |
|
WSHI2: I would like to continue the practice of sharing health information. |
|
WSHI3: I will find more effective ways to share health information on Zhihu. |
|
WSHI4: I would like to participate in the discussion of health information content and express my views. |
|
WSHI5: I am willing to spend time to improve my knowledge system to provide others with better health information content. |
Zhihu users with a history of health information sharing were selected as the research objects. In this study, a history of health information–sharing behavior (screening criteria) was defined as follows:
Publishing health-related information (including asking questions, answering questions, posting articles or ideas)
Commenting on health-related information (20 words or more)
Sharing or forwarding health-related information (eg, to WeChat, microblog, and other platforms)
The online questionnaire was issued from June 5 to June 19, 2020 (14 days). At the end of the period, 921 Zhihu user responses were collected. Among them, 210 users had previously shared health information. After eliminating responses with missing values, 185 valid responses were obtained. This number accounts for an effective rate of 88.10% (185/210).
Preprocessing, such as data filtering, was completed using Excel (Microsoft Corp) before importing information to the database. SPSS Statistics version 24.0 (IBM Corp) with AMOS version 24.0 and PROCESS [
The Cronbach alpha coefficient of the questionnaire was .961. It was well above .60 for each variable [
Factor load, Cronbach alpha, average variance extracted, and composite reliability values of each variable.
Variable and index | Factor load | Cronbach alpha | AVEa | CRb | |||||
|
—c | .875 | .650 | .881 | |||||
|
AL1 | .754 | — | — | — | ||||
|
AL2 | .792 | — | — | — | ||||
|
AL3 | .906 | — | — | — | ||||
|
AL4 | .764 | — | — | — | ||||
|
— | .894 | .622 | .892 | |||||
|
CI1 | .806 | — | — | — | ||||
|
CI2 | .776 | — | — | — | ||||
|
CI3 | .797 | — | — | — | ||||
|
CI4 | .772 | — | — | — | ||||
|
CI5 | .792 | — | — | — | ||||
|
— | .688 | .638 | .727 | |||||
|
ER1 | .842 | — | — | — | ||||
|
ER2 | .672 | — | — | — | ||||
|
— | .785 | .755 | .902 | |||||
|
IR1 | .918 | — | — | — | ||||
|
IR2 | .922 | — | — | — | ||||
|
IR3 | .756 | — | — | — | ||||
|
— | .892 | .628 | .893 | |||||
|
SE1 | .648 | — | — | — | ||||
|
SE2 | .848 | — | — | — | ||||
|
SE3 | .853 | — | — | — | ||||
|
SE4 | .776 | — | — | — | ||||
|
SE5 | .819 | — | — | — | ||||
|
— | .928 | .724 | .929 | |||||
|
WS1 | .787 | — | — | — | ||||
|
WS2 | .877 | — | — | — | ||||
|
WS3 | .850 | — | — | — | ||||
|
WS4 | .910 | — | — | — | ||||
|
WS5 | .824 | — | — | — |
aAVE: Average variance extracted.
bCR: Critical ratio.
cNot applicable.
Content validity reflects the degree to which the description of measurement items affects the survey results. The measurement items in the questionnaire were mainly taken from the published literature. The self-designed indexes are obtained through expert discussion. They were then combined with the characteristics of the research object. Therefore, we believe that the scale has a good content validity. Structural validity includes both convergent and discriminant validities. The main measurement indexes are the factor load and average variance extracted (AVE) [
Discriminant validity matrix.
Variable | ALa | CIb | ERc | IRd | SEe | WSHIf |
AL | 0.650 | —g | — | — | — | — |
CI | 0.420 | 0.623 | — | — | — | — |
ER | 0.619 | 0.628 | 0.638 | — | — | — |
IR | 0.637 | 0.454 | 0.655 | 0.754 | — | — |
SE | 0.562 | 0.261 | 0.505 | 0.721 | 0.628 | — |
WSHI | 0.688 | 0.324 | 0.654 | 0.735 | 0.628 | 0.724 |
AVEh | 0.806 | 0.789 | 0.799 | 0.868 | 0.793 | 0.851 |
aAL: Altruism.
bCI: Community influence.
cER: Extrinsic reward.
dIR: Intrinsic reward.
eSE: Self-efficacy.
fWSHI: Willingness to share health information.
gNot applicable.
hAVE: Average variance extracted.
Demographic characteristics of participants.
Variable | Value, n (%) | |
|
||
|
Male | 54 (29.2) |
|
Female | 131 (70.8) |
|
||
|
≤18 | 9 (4.9) |
|
19-38 | 168 (90.8) |
|
39-58 | 8 (4.3) |
|
||
|
Senior high school and below | 3 (1.6) |
|
Junior college | 3 (1.6) |
|
Undergraduate | 143 (77.3) |
|
Master and above | 36 (19.5) |
|
||
|
Medical science or related | 139 (75.1) |
|
Nonmedical-related | 46 (24.9) |
|
||
|
Student | 150 (81.1) |
|
Government personnel | 11 (5.9) |
|
Professional technical personnel | 11 (5.9) |
|
Business and service personnel | 4 (2.2) |
|
Other | 9 (4.9) |
Model fitting test index values.
Index | Value | Standard | Fitting |
1.959 | <5 | acceptable | |
|
|
<3 | ideal |
RMSEAb | 0.072 | <0.08 | acceptable |
|
|
<0.05 | ideal |
IFIc | 0.934 | >0.9 | ideal |
CFId | 0.933 | >0.9 | ideal |
a
bRMSEA: Root mean square error of approximation.
cIFI: Incremental fit index.
dCFI: Cumulative fit index.
Structural equation model path diagram. AL: altruism; CI: community influence; IR intrinsic reward; ER: extrinsic reward; SE: self-efficacy; WS: willingness to share.
Path testing.
Path | Unstandardized coefficient | Standardized coefficient | Standard error | CRa | P value |
CIb→SEc | 0.814 | .764 | 0.092 | 8.818 | <.001 |
SE→WSHId | 0.520 | .468 | 0.095 | 5.313 | <.001 |
ALe→WSHI | 0.285 | .264 | 0.087 | 3.446 | <.001 |
CI→WSHI | 0.328 | .277 | 0.125 | 2.963 | .003 |
IRf→WSHI | 0.291 | .260 | 0.153 | 2.157 | .03 |
ERg→WSHI | –0.347 | –.351 | 0.101 | –3.433 | <.001 |
aCR: Critical ratio.
bCI: Community influence.
cSE: Self-efficacy.
dWSHI: Willingness to share health information.
eAL: Altruism.
fIR: Intrinsic reward.
gER: Extrinsic reward.
Based on the PROCESS [
Model for testing the mediating effect of self-efficacy.
Variable | SEa | WSHI1b, c | WSHI2d | ||||
|
P value | P value | P value | ||||
ALe | 2.387 | .02 | 5.169 | <.001 | 4.506 | <.001 | |
ERf | 0.609 | .54 | –2.298 | .02 | –2.811 | .006 | |
IRg | 3.102 | .002 | 3.883 | <.001 | 2.754 | .007 | |
CIh | 4.680 | <.001 | 6.804 | <.001 | 5.012 | <.001 | |
SE | —i | — | — | — | 6.261 | <.001 |
aSE: Self-efficacy.
bWSHI: Willingness to share health information.
cNo Moderating variables.
dSelf-efficacy was introduced as a moderating variable.
eAL: Altruism.
fER: Extrinsic reward.
gIR: Intrinsic reward.
hCI: Community influence.
iNot applicable.
Proportion of the mediating effect.
Mediating effect | β | Boot SE | Boot upper | Boot lower | % |
Total | .505 | 0.088 | 0.327 | 0.670 | — |
Direction | .358 | 0.082 | 0.198 | 0.517 | 70.83 |
Mediation | .147 | 0.047 | 0.063 | 0.245 | 29.15 |
Altruism has a positive effect on the user’s WSHI. From the perspective of social norms, altruism is a moral requirement and standard for individual social values based on one’s ability and social influence. In other words, this is the self-perception of “with great power comes great responsibility.” In a social Q&A community, users tend to exert certain moral requirements and restrictions on themselves based on their own cognitive ability and knowledge. These include the sharing of the health information they know and grasping to help other community users. Raj et al [
In the context of a social Q&A community, intrinsic reward has a positive effect on WSHI, but extrinsic reward has a negative impact. Users typically respect, praise, and thank the information sharers when users improve their health with the help of information shared by others. Intrinsic reward, such as respect and reputation, can promote a sense of satisfaction, pleasure, and fulfillment among health information sharers. In turn, this state of mind can continue to generate their WSHI. Thus, intrinsic reward (ie, reputation) can positively affect users’ willingness in health information sharing [
The conclusion that extrinsic reward negatively affects users’ willingness in health information sharing differs from that of existing research. This finding may be caused by the demographic distribution characteristics of the study’s current sample. The participants are mainly college students with their family or parents as their main sources of income. After the lower economic pressure is mapped to these users and their WSHI, it was found that intrinsic reward (ie, reputation and respect) can have a greater influence than extrinsic reward. Conversely, inappropriate extrinsic reward may cause user aversion or resistance. The users may feel that their sharing behavior is controlled by the organization if extrinsic reward is the intention of sharing health information. Just like the imposition of punishment from the organization, material reward is another mechanism of controlling individual behavior. This can cause an individual to lose interest and enthusiasm in sharing health information [
Community influence and self-efficacy have a positive effect on users’ WSHI. Moreover, self-efficacy has a mediating effect similar to the way that community influence, an objective variable in the environment, affects WSHI. This objective fact includes various aspects such as community development philosophy, platform design, information protection, and user group influence. These factors interact with each other, and they not only have a direct impact on the WSHI but also exert further influence by positively affecting users’ sense of self-efficacy. The perception and evaluation of factors such as self-ability and environmental conditions are necessary steps before users share health information. Users tend to be satisfied with their information-sharing behavior in perceiving that their own knowledge can help other users [
Many scholars have conducted in-depth research in the field of knowledge sharing. They proposed different knowledge sharing models for different types of information or communication environments [
This study also has certain limitations. First, only Zhihu users were taken as the objects of this research. Thus, not all social Q&A community users are covered. Follow-up research should further improve the coverage of social Q&A community users. Second, the samples mainly comprised ordinary Zhihu users, and key users with a large number of followers are underrepresented. Third, although the sample population, mainly comprising medical college students, can better represent the information-sharing behavior of the general population, the snowball sampling method used in this study still has systematic errors. Fourth, the method of using an online questionnaire may introduce bias of demographic characteristics. Finally, the data obtained were formed by subjective reports provided by participants. Overly conservative or exaggerated choices can lead to a certain degree of bias in the statistical results. More scientific experimental designs can be adopted to avoid the biases caused by these deficiencies in follow-up studies. Despite these shortcomings, this study presents novel ideas, and the results provide new insights into the promotion of WSHI.
Promoting the dissemination of high-quality health information is important for guiding users of a social Q&A community to actively participate in health information sharing. Compared with relevant research, this study introduces the variable community influence into the model based on the characteristics of a social Q&A community. Additionally, combining the variables intrinsic reward, extrinsic reward, altruism, and self-efficacy, WSHI’s SEM in a social Q&A community is constructed. The results indicate that intrinsic reward, altruism, and self-efficacy have a positive effect on WSHI. By contrast, extrinsic reward has a negative effect. Self-efficacy has a mediating effect on the relationship between community influence and WSHI. The generation of WSHI may be promoted by paying more attention to the social atmosphere of the community, optimizing the gratitude feedback mechanism, and striving to build good social relations among users. The results can provide a theoretical and practical reference for social Q&A community operators, health education and promotion, and other aspects.
Questionnaire.
average variance extracted
cumulative fit index
incremental fit index
question-and-answer
root mean square error of approximation
structural equation model
social network site
willingness to share health information
chi-square/degree of freedom
This study was supported by the Western China Social Sciences Program entitled Research on Health Information Transmission Path and Diffusion Model Based on Mobile Internet (Item No. 16XTQ012). We would like to thank all the participants who took part in the questionnaire survey. Furthermore, we are also very grateful to all the reviewers for their careful examination of this paper.
None declared.