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 the Journal of Medical Internet Research, is properly cited. The complete bibliographic information, a link to the original publication on http://www.jmir.org/, as well as this copyright and license information must be included.
Rewarding health knowledge and health service contributors with money is one possible approach for the sustainable provision of health knowledge and health services in online health communities (OHCs); however, the reasons why consumers voluntarily reward free health knowledge and health service contributors are still underinvestigated.
This study aimed to address the abovementioned gap by exploring the factors influencing consumers’ voluntary rewarding behaviors (VRBs) toward contributors of free health services in OHCs.
On the basis of prior studies and the cognitive-experiential self-theory (CEST), we incorporated two health service content–related variables (ie, informational support and emotional support) and two interpersonal factors (ie, social norm compliance and social interaction) and built a proposed model. We crawled a dataset from a Chinese OHC for mental health, coded it, extracted nine variables, and tested the model with a negative binomial model.
The data sample included 2148 health-related questions and 12,133 answers. The empirical results indicated that the effects of informational support (β=.168;
Informational support, emotional support, social norm compliance, and social interaction positively influence consumers to voluntarily reward free online health service contributors. Social interaction enhances the effect of informational support but weakens the effect of emotional support. This study contributes to the literature on knowledge sharing in OHCs by exploring the factors influencing consumers’ VRBs toward free online health service contributors and contributes to the CEST literature by verifying that the effects of experiential and rational systems on individual behaviors can vary while external factors change.
With the development of information and communication technologies (ICTs), the sharing economy (SE) has emerged as a market for collaborative consumption in which peer communities gain access to a pool of shared knowledge and resources [
Similar to many other noncommercial web-based SE platforms, OHCs are facing the sustainability issue (ie, the provision of free health knowledge and health services) [
To keep the sustainable provision and sharing of free health knowledge and health services, some OHCs have designed a new feature that allows consumers to voluntarily reward free health service contributors. Such rewarding behavior is particularly important for OHCs to thrive, because the rewards act as monetary incentives that can stimulate health service providers to continuously contribute high-quality health knowledge and free health services [
What are the factors that motivate consumers to voluntarily reward free health service contributors in OHCs?
How do those factors motivate consumers to voluntarily reward free health service contributors in OHCs?
This study aimed to address the abovementioned questions. We adopted the cognitive-experiential self-theory (CEST) as the theoretical foundation and proposed seven hypotheses. We crawled an objective dataset from an OHC for mental health and verified most of the hypotheses. The empirical results indicate that informational support, emotional support, social norm compliance, and social interaction positively influence consumers to voluntarily reward free online health service contributors. Social interaction enhances the effect of informational support but weakens the effect of emotional support. These findings provided several important theoretical contributions and practical implications.
We reviewed two streams of related studies to address the research questions. Specifically, we reviewed the literature on free health services in OHCs to describe the characteristics of free online health services. We reviewed the literature on pay-what-you-want to understand the theories, variables, and models that were used to explain consumers’ voluntary rewarding behaviors (VRBs). In this section, we have summarized the implications of prior studies.
There are different types of OHCs (eg, peer communication for health care professionals, physician-patient interaction communities, and patient-patient interaction communities), and activities in different OHCs are organized differently [
Free web-based health services provide consumers many benefits. Consumers can conduct health-related activities in OHCs, such as health knowledge sharing and seeking (eg, recommending treatment plans and seeking health care suggestions) and health self-management [
The nature of free health services in OHCs can be treated as social support [
Voluntarily rewarding free health services belongs to an emerging business model that gives consumers full control in monetizing free web-based knowledge/goods/services [
Key constructs related to the pay-what-you-want behaviors in prior studies.
Sources | Contexts | Theory | Independent variables | Dependent variables |
Kim et al [ |
Restaurant, cinema, and delicatessen | Equity theory |
Fairness, altruism, satisfaction, and loyalty CVsa: price consciousness and income |
Final price paid |
Jang and Chu [ |
Experiments for consumers | Equity theory |
Fairness motives of individuals, self-signaling, and norm conformity |
Willing to pay |
León et al [ |
Travel company | Game theory |
Customer characteristics, the influence of subjective factors, and product characteristics |
Payments in El trato |
Hilbert and Suessmair [ |
A laboratory experiment about a travel mug | N/Ab |
Social interaction and social norm compliance |
Willing to pay |
Regner [ |
An online survey about the online music label/store, Magnatune | N/A |
Social preferences, reciprocity, guilt, social norms, altruism, fairness, and social image concerns |
Willing to pay |
Barone et al [ |
A leadership questionnaire | N/A |
Consumer power, perceived value, and perceived self-reliance |
Purchase intentions |
Dorn and Suessmair [ |
Survey in several countries under three hypothetical situations where a McDonald’s Big Mac was offered | N/A |
Satisfaction, income, price consciousness, reference price, high level of reputation, loyalty, altruism, fairness, social acceptance, and social norm compliance |
Willing to pay |
Narwal and Nayak [ |
Scenario-based online experimental approach on purchase intention | N/A |
Quality of product/services, satisfaction, types of products/services, self-image, and fairness perception Moderators: communication message, interaction, and reference prices |
Pay-what-you-want |
Viglia et al [ |
Service | Fairness theory |
Timing and uncertainty reduction |
Consumers’ chosen payments |
aCV: control variable.
bNot applicable.
We concluded three useful findings according to the literature review. First, pay-what-you-want is a result of consumers’ positive experiences with the services via direct interactions with service providers [
Specific to this study, we proposed that consumers’ VRBs toward free health service contributors are a result of consumers’ positive experiences with the services via direct interactions with service providers in OHCs [
Second, research focuses are shifting with time. As discussed above (please see the timeline of prior studies in
Finally, there is a lack of conceptual frameworks in analyzing consumers’ pay-what-you-want behaviors. Scholars tend to analyze this issue from a prosocial motivation perspective. They have adopted theories such as the equity theory and fairness theory to select influencing factors (see
As there is a lack of conceptual frameworks to explain consumers’ pay-what-you-want behaviors [
CEST is a psychological theory that argues that human beings operate with two systems: an experiential/intuitive system (hereafter referred to as the experiential system) and a rational/analytical system (hereafter referred to as the rational system) [
CEST is being widely used to explain consumers’ web-based behaviors, including their web purchase–related decisions. For example, consumers’ reactions to experiential information demonstrates a contagion effect: experiential information at the early stage can cause more similar information in the following stage, and normal consumers like to follow opinion leaders who post experiential information [
We built our research model based on the following logic.
According to CEST, the rational is a verbal reasoning system—it suggests that human behaviors are driven by logic inferences from the information or evidence received [
According to CEST, the experiential system is an affect-driven system—it suggests human behaviors are directed by pursuing positive feelings and avoiding negative feelings [
CEST also argues that the relative influence of both systems varies along a dimension of complete dominance by one system to complete dominance by the other [
Hypotheses and research model.
Informational support refers to the overall quality and usefulness of the information received in OHCs. According to CEST, the rational system is verbal and based on the information received, so users tend to rely on rational processing when receiving informational support. Service providers and consumers usually collaboratively generate health services in the form of question and answers in OHCs. Consumers post their questions and respondents address these questions. They discuss health-related issues and generate new health knowledge in OHCs. CEST also suggests that by rational processing, consumers behave based on the logical inference from information/evidence received [
H1: Informational support expressed in free health service threads positively influences consumers’ voluntary rewarding behaviors in OHCs.
Emotional support refers to sympathy, ie, perceiving, understanding, and reacting to others’ distress or needs [
H2: Emotional support expressed in free health service threads positively influences consumers’ voluntary rewarding behaviors in OHCs.
Social norm compliance refers to conformity to a set of norms that are accepted by a significant number of people in a social surrounding, community, or society [
H3: Social norm compliance positively influences consumers to voluntarily reward free health service contributors in OHCs.
Social interaction refers to the observed strength of relationships, the amount of time spent, or the communication frequency among health service providers and consumers in a health service thread [
H4: Social interaction between service providers and consumers motivates consumers to voluntarily reward online free health service contributors in OHCs.
CEST suggests that the extent to which people think or behave primarily according to the experiential system or rational system depends on the situation [
OHCs are web-based social networks in which health-related stakeholders with common interests, goals, or practices interact to share health information and knowledge, communicate health services, and engage in social interaction [
H5: Social interaction positively moderates the effect of informational support on consumers’ voluntary rewarding behaviors in OHCs.
As discussed earlier, both emotional support and social norm compliance are factors relating to the experiential system. According to CEST, because the relative influence of the experiential system and rational system varies from complete dominance by one to complete dominance by the other [
H6: Social interaction negatively moderates the effect of emotional support on consumers’ voluntary rewarding behaviors in OHCs.
H7: Social interaction negatively moderates the effect of social norm compliance on consumers’ voluntary rewarding behaviors in OHCs.
To test the hypothesized model, we crawled an objective dataset from the question and answer forum on a Chinese OHC for mental health (the question and answer forum on YiXinLi). YiXinLi is a leading web-based health community for mental health in China. We focused on mental health because without mental health there can be no true physical health [
YiXinLi was set up in 2011 and aims to promote mental health services in China. The question and answer forum on YiXinLi, which was launched in 2014, provides free mental health services for consumers. Consumers can post their health-related questions in the question and answer system and wait for free answers. However, with the emerging trend of knowledge monetizing [
We used a spider program (named Locoy Spide) and crawled all the threads on the YiXinLi question and answer forum on January 12, 2019. We treated a question and answer thread (ie, a question and its answers) as the basic analysis unit. We cleaned the data by deleting 12 inconsistent threads—the threads in which the actual number of answers was less than the number shown on the web page because one or more answers were deleted by the providers (the number of answers displayed on the web page includes all the answers that have been provided. However, if a provider deletes his or her answer, the number shown on the web page does not change, but the actual number of answers we crawled would be less than the number shown on the web page). After cleaning the data, we had 2148 data samples, including 2148 questions and 12,133 answers.
As shown in
A sample of a question.
A sample of an answer.
As shown in
We coded nine variables that were used for data analysis. We treated consumers’
The descriptive statistical results of different variables are shown in
Variables and measurement.
Variable | Value, mean (SD) | Measurement |
VRBa | 2.141 (3.334) |
The VRB is measured by the rewarding times of a thread received. For example, the answers of the sample thread in We did not use the sum of real money that all answers received. In fact, we cannot capture the actual sum of rewarded money in a thread |
ISb | 4.375 (3.991) |
On YiXinLi, consumers can evaluate the answer quality with the feature, For example, the answer in |
ESc | 3.274 (1.467) |
On YiXinLi, providers and other consumers can use the feature, We thus use the volume of Although |
SNCd | 0.536 (0.61) |
SNC is measured by the percentage of people interested in the question who finally reward the question. Such a measurement reflects the peer pressure the consumers feel when they find that others have rewarded the thread they viewed. We designed this measurement according to industrial practice and prior studies. Previous literature suggests that other consumers’ purchase behavior (number of goods purchased) acting as social norms influences a focal consumer’s intention [ Specifically, VRB refers to the number of rewarding. The volume of For example, there are five favorites in |
SIe | 8.75 (8.757) |
SI is measured by the interaction frequency between service providers and consumers in a thread. On YiXinLi, providers can respond to a question by posting their answers. Providers and consumers can also discuss a particular answer via the feature For example, there are three answers and 0 comments in |
ALf | 188.4 (120.866) |
AL refers to the average text length of all answers in a thread. We calculated the character numbers of all answers and then divided the volume of answers in a thread For example, there are three answers in a thread. The first one has 200 characters, the second one has 300 characters, and the last one has 400 characters. Thus, the value of AL is 300 (ie, (200+300+400)/3=300) |
DoEg | 73.17 (135.115) |
DoE is measured by comparing the time a question is posted with the time we crawled the dataset |
PVh | 647.985 (1918.211) |
PV refers to how many times a thread is read. For example, the thread in |
PRi | 0.835 (0.193) |
On YiXinLi, there are 3 rank levels for a service provider, ie, normal provider, higher-rank provider, and top provider. The rank level is related to how many times their answers were set as best answers. We used the rate of higher rank/top providers of all providers in a thread to measure the PR For example, the three providers in a thread include one normal provider, one higher-rank provider, and one top provider. Thus, the value of PR is 0.667 (ie, 2/3=0.667). |
aVRB: voluntary rewarding behavior.
bIS: informational support.
cES: emotional support.
dSNC: social norm compliance.
eSI: social interaction.
fAL: answer length.
gDoE: date of exposure.
hPV: page view.
iPR: provider reputation.
Results of descriptive statistics and the covariance matrix.
Variables | Value, mean (SD) | Min | Max | VRBa | ALb | PVc | DoEd | PRe | ISf | ESg | SNCh | SIi |
VRB | 2.141 (3.334) | 0 | 37 | 1 | 0.025 | 0.216j | −0.019 | 0.013 | 0.562j | 0.376j | 0.464j | 0.490j |
AL | 188.46 (120.866) | 9 | 894 | 0.025 | 1 | −0.025 | 0.015 | −0.071k | 0.081j | 0.001 | 0.029 | 0.104j |
PV | 647.985 (1918.211) | 17 | 46,173 | 0.216j | −0.025 | 1 | 0.350j | −0.136j | 0.374j | 0.110j | 0.026 | 0.302j |
DoE | 73.170 (135.115) | 0 | 2457 | −0.019 | 0.015 | 0.350j | 1 | −0.234j | 0.113j | −0.052l | −0.031 | 0.164j |
PR | .835 (.193) | 0 | 1 | 0.013 | −0.071k | −0.136j | −0.234j | 1 | −0.189j | −0.002 | 0.094j | −0.230j |
IS | 3.274 (1.467) | 1 | 54 | 0.562j | 0.081j | 0.374j | 0.113j | −0.189j | 1 | 0.357j | 0.052l | 0.135j |
ES | 4.375 (3.991) | 0 | 14.5 | 0.376j | 0.001 | 0.110j | −0.052l | −0.002 | 0.357j | 1 | 0.055l | 0.116j |
SNC | .536 (.610) | 0 | 5.333 | 0.464j | 0.029 | 0.026 | −0.031 | 0.094j | 0.052l | 0.055l | 1 | 0.589j |
SI | 8.75 (8.757) | 1 | 88 | 0.490j | 0.104j | 0.302j | 0.164j | −0.230j | 0.135j | 0.116j | 0.589j | 1 |
aVRB: voluntary rewarding behavior.
bAL: answer length.
cPV: page view.
dDoE: date of exposure.
ePR: provider reputation.
fIS: informational support.
gES: emotional support.
hSNC: social norm compliance.
iSI: social interaction.
j
g
j
As our dependent variable (ie,
We ran the NB model with the volume of
Results of the negative binomial model (N=2148).
Indicesa,b,c | Results | |||
|
Coefficient | SE | Z test | |
Constant | 0.367d | 0.021 | 17.180 | <.001 |
Response length | −0.033e | 0.019 | −1.780 | .07 |
Page view | 0.072d | 0.017 | 4.220 | <.001 |
Date of exposure | −0.050f | 0.022 | −2.250 | .02 |
Provider reputation | 0.135d | 0.023 | 5.960 | <.001 |
Informational support | 0.168d | 0.020 | 8.540 | <.001 |
Emotional support | 0.463d | 0.023 | 20.490 | <.001 |
Social norm compliance | 0.510d | 0.018 | 28.150 | <.001 |
Social interaction | 0.281d | 0.021 | 13.230 | <.001 |
Social interaction×informational support | 0.032f | 0.013 | 2.410 | .02 |
Social interaction×emotional support | −0.086d | 0.006 | −13.600 | <.001 |
Social interaction×social norm compliance | 0.014g | 0.016 | 0.880 | .38 |
aLog likelihood=−3130.778.
bLikelihood ratio211=2178.5 (
cPseudo R2=0.258.
d
e
f
gNonsignificant.
As shown in
Although we proposed that social interaction negatively moderates the effect of social norm compliance on consumers’ VRBs, our results did not support this hypothesis. This may be because although CEST indicates such a negative moderating effect [
On the basis of prior related studies and grounding our research in CEST, this study has identified two health service content–related factors and two interpersonal factors and explored how these factors influence consumers’ VRBs toward free health service contributors in OHCs. Our empirical findings have demonstrated that informational support, emotional support, social norm compliance, and social interaction positively influence consumers to voluntarily reward free health service contributors. In addition, social interaction enhances the effect of informational support but weakens the effect of emotional support on consumers’ VRBs toward free health service contributors in OHCs.
This paper makes two theoretical contributions. First, we contribute to the literature on knowledge sharing in OHCs. As noncommercial web-based SE platforms are becoming increasingly popular, scholars have begun to examine health care professionals’ or consumers’ health knowledge–sharing behaviors [
Second, our research is based on CEST and also contributes to CEST. Specifically, CEST mentioned that the extent to which individuals behave primarily according to one of the systems varies based on situations or the person himself or herself [
This paper has identified and verified the effects of four main variables on consumers’ VRBs on free health services in OHCs. We contributed to noncommercial web-based SE platforms by providing these platform operators strategies on how to motivate consumers to voluntarily reward free service contributors.
First, platform operators could optimize their platform feature design. They can optimize the platform communication features and encourage service providers and consumers to interact with each other. In addition, they can design and implement new rewarding systems. For example, they can display the rewarding messages such as “consumer XX just rewarded provider YY some money.” These rewarding messages might cause more consumers to comply with others and choose to reward free service contributors.
Second, platform operators should encourage service providers to contribute professional knowledge and generate high-quality services. They can invite more professionals or experts to use their platforms. They can help enthusiastic consumers to improve professional capabilities. The engagement of professionals and enthusiastic consumers can guarantee the quality of services on noncommercial SE platforms and can in turn attract more consumers to use their platforms and reward free service contributors.
We address two potential limitations. First, we did not test the effects of consumers’ sociodemographic variables and consumer characteristics. As the dataset was crawled in a public community, we could not obtain consumers’ sociodemographic information and their characteristics. In addition, we measured all variables with the objective data, namely an indirect measurement approach. Second, different from prior studies that use the actual volume of money as dependent variables, we used the number of times a thread is being rewarded as the dependent variables. We are not sure whether these points undermine our conclusions or not. We appeal that more studies be conducted through the econometric modeling approach and also suggest a mixed method approach of combining objective data and subjective data in future studies.
cognitive-experiential self-theory
information and communication technology
negative binomial
online health community
sharing economy
voluntary rewarding behavior
This work was supported by the National Natural Science Foundation of China (71501062), Key Projects of Philosophy and Social Sciences Research of Chinese Ministry of Education (grant number 19JZD021), Guangdong Provincial Science and Technology Research Project (grant number 2019A101002110), and Shantou University Scientific Research Initiation Grant (STF18011). Full control of all primary data is with the authors, and the data will be available if requested.
None declared.