The concept of information ecology was first proposed by Horton (35). Information ecology theory views information, information people, information technology, and the information environment as a whole. It emphasizes the balance and development of humans, environment, and technology in the information ecosystem, so as to realize the production, dissemination, and utilization of information (36). Within this theoretical framework, the information people is the subject, controlling the information activities within the system. The information itself is the object, existing independently of human will. Information technology serves as the medium for information transmission. The information environment is the field where the interaction between the subject and object occurs. The core issue of this theory is the information behavior generated by the information people and the interactions between the information people and other elements (37).

As a classic theory in the field of explaining user information behavior, information ecology theory has been widely applied in studies on influencing factors of OHCs users’ information seeking behavior (38), information adoption behavior (39), information interaction behavior (40), and information sharing behavior (41). Ji and Li (40) explored the willingness for emotional interaction among OHCs users from an information ecology perspective, in which the information dimension includes topic relevance and topic value, the information environment dimension includes group connection and group identity, the information technology dimension includes perceived platform usefulness and perceived function ease of use, and the information people dimension includes health status and health experience. Gao et al. (42) applied the information ecology theory to study the information processing of online health community users, in which the information dimension includes information quality, the information environment dimension includes emotional support and privacy issues, the information technology dimension includes system quality, and the information people dimension includes self-efficacy. Zhang et al. (36) explored the behavior of seeking medical information assistance from an information ecology perspective among OHCs users, where the information dimension includes information accuracy, information relevance, and information timeliness; the information environment dimension includes social acceptability and platform trust; the information technology dimension includes perceived ease of use and perceived usefulness; and the information people dimension includes self-efficacy, health information literacy, and assistance-seeking experience.

Users’ discernment of health information is an important part of online health community activities, affecting users’ health Decisions and usage experience. The complex influencing factors in health information discernment can also be explained systematically and holistically from the perspective of information ecology theory (43). In this study, from the viewpoint of information factors, users not only need to examine the quality of health information, but also consider whether the source of health information is credible and authoritative, which provides good information resources for their own health Decisions. From the viewpoint of information environment, rich information support and emotional support realms help to create an efficient, orderly, and sustainable user interaction environment, thus improving the efficiency of obtaining correct health information and promoting the quality of community information services. From the viewpoint of information technology, information technology connects users with the community environment. The safe, intelligent, and convenient technology can expedite the realization of platform functions and provide a basic guarantee for the good experience of users. From the viewpoint of information people, users are the most active elements in the process of information active. The driving force and the confidence in the discerning task are related to the flow of health information in the community. Therefore, this study explores the factors influencing the health information discernment abilities of online health community users from four aspects: information (information quality, information source), information environment (informational support, emotional support), information technology (technological security, technological facilitation), and information people (self-efficacy).

Mehrablan and Russell (44) proposed that the general human behavior model is the “stimulus-organism-response” model, which is used to explain and predict the impact of stimuli from different environments on individual cognition, individual behavior, and personal emotions. The model consists of three parts: stimulus, organism, and response. Stimulus is defined as some kind of object, event, or characteristic, which can be divided into external stimulus and internal stimulus. Organism, as an intermediary variable, is an individual’s psychological transformation mechanism, such as emotional or cognitive changes, which indicates that the behavior of the organism will be affected by the influence of internal factors. Response represents the individual outcome variable, which generally manifests as an approach or avoidance of things (45).

This model has been widely used in studies on individual offline and online behaviors, such as information participation behavior (46). Recently, it has been used to study the online behaviors of individuals in the face of online disinformation and rumors (47). Based on Stimulus-Organism-Response theory, Liu et al. (48) took rumor type and rumor ambiguity as stimulus and perceived authenticity, perceived importance and perceived trust as the organism’s perceptual states, and verified the influence mechanism of individual’s intention to verify rumor. Wu (49) considered informational dependency and social dependency as stimulus, and investigated how such stimulus led to positive and negative effects within organisms, which in turn affects the individual’s perception of misinformation sharing. Xie and Hu (50) regarded perceived credibility as an external stimulus, subjective norms, individual norms, and attitudes as organic states, and the rumor deletion intention and the rumor disproving sharing intention as a response, verifying the rumor correction behavior of social media users and its influence paths.

Additionally, users’ perception of health information services in OHCs is closely related to their ability to discern health information. Perceived value is the overall utility assessment of health information attributes by users (51). From the perspective of perceived value, when users perceive that the health information they obtain can bring value to themselves, they generate positive emotions, adopt the information, and further enhance their ability to discern information. Perceived risk is the subjective judgment by users of potential negative outcomes perceived in online health information services (52). Perceived risks include false health information, privacy concerns, privacy breaches, and financial security, among others, which are crucial factors influencing users’ use of online health information services (53). Based on the S-O-R theory, Cui (54) found that users’ perceived risks have a negative impact on the dissemination of false information on the internet. For health information with higher perceived risks, users need to invest more effort in systematic and rational thinking. In this study, perceived value and perceived risk represent the users’ organism perception states, which affect their health information discernment abilities.

It can be seen that the S-O-R theory is very suitable for research on health information in online health communities. Based on the S-O-R theory theoretical framework, this study examined the health information discernment abilities of OHCs users. In this study, when users exchange information in an online health community, they are not only stimulated by external factors, including the information itself, the environment of the community, and information technology; they are also stimulated by their own factors, such as self-efficacy. Therefore, the “stimulus” refers to the information factor (information quality, information source), information environment (informational support, emotional support), information technology (technological security, technological facilitation), and information people (self-efficacy). Under the double stimulation of internal and external factors, the psychological perception of the user will change. This “organism” perceptual state changes, mainly expressed in the size of the perceived risk and perceived value judgment. Finally, the user’s “reaction,” that is, health information screening ability.

Vuong and Napier (55) proposed a new information processing mechanism: Mindsponge Theory. It describes the psychological mechanism of which a person accepts or rejects new external information (56). First, an individual receives external information through books, media, the Internet, etc. Then, the individual will sift and filter the information according to established values, beliefs, and cognitive frameworks, and categorize the information into useful and useless, credible and non-credible categories. Later, the information that is considered useful and credible will be absorbed and integrated into the individual’s cognitive structure, which is similar to the process of a sponge sucking up water, and rejecting incompatible information. The individual May adapt and reconfigure the existing cognitive structure to fit the new information and environment. Ultimately, the reconfigured cognitive structure influences the individual’s feedback to the outside world.

This theory has been used to study vaccine production (57), suicidal behavior (58), and health information processing (57, 59). For example, Tanemura et al. (60) explored the differences in individuals’ trust levels of risk-only negative health information and benefit–risk negative health information based on the Mindsponge Theory. They found that individuals’ trust levels of risk-only negative health information were higher than those of benefit–risk negative health information. Mindsponge theory was also applicable in explaining that learning interventions for individuals that improve individuals’ ability in recognizing unreliable COVID-19 information (61).

In this study, when a user receives new health information from an online health community, the user goes through the brain’s filtering system to weigh its perceived value and perceived risk. When an individual’s perceived value of health information is greater than the risk, then the resulting high level of trust promotes health information uptake. Conversely, when risk outweighs value, i.e., when the perceived adverse consequences of new health information for an individual’s health Decisions outweigh its convenience, and health information is perceived to be unreliable, then health information is rejected. In conclusion, users’ health information screening activity is an uninterrupted and continuous process of selective information absorption and exclusion in order to maintain psychological and cognitive balance in order to adapt to the complex online health community information environment, thus realizing the improvement of health information discernment abilities.

In this study, information elements mainly included information quality and information sources. Information quality is a measure of the accuracy, truthfulness, and completeness of health information (62). The health information searched in OHCs includes advice and information on physical conditions, symptoms, and treatment options, which mainly guides users’ health Decisions. Obviously, the accuracy and authenticity of information are crucial. When persuasive and high-quality health information is searched for, it can enhance users’ perceived benefits from the information, thereby promoting perceived value and better satisfying their own information needs (63). On the contrary, the reliability of information is considered to reduce an individual’s perception of risk. Shah and Wei (64) conducted a survey on 630 users during the COVID-19 pandemic and found that source credibility and information quality are significantly positively correlated with perceived value and significantly negatively correlated with perceived risk. Sun and Wu (65) showed that the comprehensiveness and accuracy of COVID-19 information content can affect the public’s perceived risk. When health information is incomplete or lacks logical consistency, it can mislead users, increase their concerns, and elevate their psychological risk levels. Therefore, individuals with high perceived risk require more convincing arguments, such as higher quality health information, to reduce their risk perception. Liang et al. (66) drawing on rational choice theory, studied the online health information use of physically disabled individuals and found that information quality and system quality can increase perceived benefits and reduce perceived risks. Therefore, three hypotheses are proposed as follows:

The information source considers the credibility of the information source, independent of the information’s content. In OHCs, users can act as both creators and disseminators of health information, and their knowledge and experience vary. Hence, the source of information is an important factor in information behavior research (67). Shareef et al. (68) pointed out that when the source credibility is low, the recipients discount the value of the information content they receive. Information from authoritative and professional sources generally has a low perceived risk to users, while information from user-generated content and commercial sources May increase the perceived risk to users. Cha et al. (69) showed that users are more likely to have positive cognitive and affective responses to information obtained from Weibo that have a high level of credibility in its sources. Thompson (70) points out that perceived risk is negatively correlated with the trust of the Internet as an information source. This means that as people’s trust in these sources of information Decreases, their perception of risk to health information increases. At the same time, clear information sources are more conducive for users to judge the authenticity of health information. Therefore, three hypotheses are proposed as follows:

In this study, the elements of the information environment mainly included informational support and emotional support. Users engage in online discussions and interactions on the internet to obtain specific social support, including emotional, informational, companionship, and technical support (71). Informational support means that users obtain health and medical-related information through searching, consulting, answering, and discussing, including information on medications, healthcare measures, and treatment suggestions (72). Thus, OHCs facilitate the acquisition and sharing of health information and knowledge. According to the Mindsponge theory, when members of OHCs provide valuable advice or timely help to each other, users identify with these new values and assimilate and transform the health information into health knowledge, which is conducive to users’ reduction of informational uncertainty and making the right health Decisions, and facilitates their future self-health management. Hu et al. (73) used a case analysis method to conduct an in-depth analysis of the formation mechanism of patients’ perceived value, and found that sharing information among members is the key to improving patients’ perceived value. Therefore, the following hypotheses are made:

Emotional support is expressed as the respect, acceptance, or emotional comfort and encouragement received when an individual is in pain (74). Adequate emotional support is a key factor in creating a favorable atmosphere of interaction that promotes the sustainability of an online health community. Park et al. (75) stated that empathy and encouraging messages in OHCs can provide both emotional support and informational support to members. Emotional support can provide positive energy to other members, making community members feel more confident, in control, and empowered in order to address similar health issues. Receiving emotional support is also an important source of perceived value for users. Kanthawala and Peng (76) found a positive correlation between perceived emotional support and perceived message credibility. In addition, Sun (77) believes that emotional support, as a social resource, is a significant factor affecting the older adult’s health information discernment abilities. Therefore, the following hypotheses are made;

In this study, information technology elements mainly included technological security and technological facilitation. Technological security in OHCs refers to the security of hardware, software, and related data in online health communities. In the era of big data, the value of information and data is increasing day by day. The subsequent leakage of user privacy information or data loss has become more and more serious. Users have increased concerns about the security of information technology. Studies have shown that in OHCs, the higher the technological security, the more willing users are to share information (52). Internet users can help form a sense of belonging to websites with higher security, thereby further increasing perceived value (78). To enhance security levels from a technological standpoint, it is important to implement advanced data encryption measures, while deploying multi-layered access control mechanisms to ensure that only authorized users can access sensitive information, thereby reducing the risk of information misuse or unauthorized access (79). Additionally, establishing a scientifically efficient information review mechanism is essential to ensure the reliability of information sources, reducing the spread of false and misleading content, and thereby improving the overall quality of information. In such an environment, users can more quickly identify and filter out reliable health information, continuously learn, absorb, and increase their knowledge; ultimately enhancing their discernment capabilities. Some scholars believed that relevant departments must strengthen supervision of the quality of online health information to ensure the security of platform information, thus strengthening the ability of users to resist the interference of misinformation (80). Therefore, the following hypotheses are proposed:

The page layout, response speed, functional diversity, and convenience of OHCs all affect users’ usability experience and the ease of distinguishing health information. Technological facilitation emphasizes the promotion and support of information technology on people’s information behavior. For example, when an OHC has a user-friendly page layout and provides a clear and concise operation guide, it saves users’ learning cost and reduces difficulty in using it. Ease of use has a positive impact on users’ willingness to use a platform (81). Scholars have found that the difficult interoperability of the platform is one of the barriers to the public’s use of online platforms to obtain useful information (82). When constructing online health community information services, leveraging big data technology to deeply analyze users’ website browsing records, search habits, clicking patterns, and browsing durations can accurately capture users’ interests and health needs (83).This allows for intelligent and personalized health information recommendations, sparing users the trouble of sifting through vast amounts of information to find what they need. Additionally, providing users with other health information verification tools, such as recommending authoritative health information websites or fake health information features lists, to help users acquire, screen and validate health information more efficiently. Tang et al. (84) found that during the COVID-19 pandemic, the subjective and objective technical support conditions positively influenced the adoption behavior of government short videos. Improving OHCs functions and enriching user experience through technological means improves users’ understanding and judgment of health information in several ways. This enables users to more accurately discern valuable health information. Therefore, the following hypotheses are proposed:

The information people element in this study mainly consisted of self-efficacy. Warner and Schwarzer (85) argued that general self-efficacy is an overarching self-efficacy of users when facing various challenges or confronting new things. This self-efficacy competence will largely influence people’s presumptions about the ability to implement or organize activities. Stronger self-efficacy can better withstand the negative impact of uncertainty factors in the network environment, the easier it is to form a better quality perception and value recognition of the results, resulting in lower perceived risk (86). Wang et al. (87) suggested that the self-efficacy of older adult users who have been in different contexts of family, community, and education for a long time affects their health information needs. Older adult users with high self-efficacy were more adept at expressing differences in the efficacy of health information products, and had a certain degree of ability to discriminate health information products. Therefore, the following hypotheses are made:

User perception elements mainly included perceived risk and perceived value. Users assess the utility of health information when using OHCs, retaining information that is valuable for their health needs and refusing to access information that does not deal with their health risks (88). In this study, the perceived usefulness of health information, the emotional pleasure, and the establishment of good interpersonal relationships, as perceived by users, are all manifestations of perceived value. The perceived risk of online health information refers to the user’s perception and assessment of the risk level conveyed by the health information, and the subjective judgment and perception of the degree of threat to personal privacy information, physical health, or personal property. According to the Mindsponge theory point of view, the user’s processing of health information can be understood as a comparison of the perceived value and perceived risk of new information based on their own existing health knowledge and health experience. The higher the perceived value and the lower the perceived risk of health information, the more likely it is for the user to acquire and utilize. Liu et al. (89) used the S-O-R theory as a theoretical framework and constructed a model of the impact of mobile Medical App content presentation on user adoption willingness influence model. The results show that perceived value plays a mediating role, and the user’s perceived value of mobile medical APP content positively influences user adoption behavior. This indicates that health information with a higher value is more likely to capture users’ attention, stimulate their interest in in-depth exploration and learning, and motivate them to analyze and validate the information more carefully to enhance their trust in it. This process enhances users’ overall health literacy, thereby improving their ability to discern health information.

In contrast, when the information uncertainty is higher or the consequences are more unpredictable, users perceive a higher risk and become more vigilant. They scrutinize the source and content of health information more thoroughly. Zhang et al. (90) found that there was a significant positive correlation between the perceived riskiness of health information and the information avoidance behaviors during the COVID-19 pandemic. This indicates that users become more alert and cautious in the face of riskier information. At the same time, this motivates users to look for more evidence and support to distinguish the difference between misinformation, disinformation, rumors, and correct information. It helps avoid being misled by false or unreliable information. This vigilance helps users develop their discernment skills (91). Perceived value and perceived risk are relative concepts. Research showed that perceived costs and security risks have a great impact on customers’ perceived value (92). Therefore, the following assumptions are made:

The reliability of the scale was tested by SPSS25.0 of IBM, and the validity of the scale was analyzed by using Amos23.0. The results are shown in Tables 3, 4.

It can be seen that the Cronbach’s Alpha varied from 0.869 to 0.914 for each construct (Table 3), exceeding the cut-off point of 0.70, demonstrating good scale reliability (99). The standardized loading values of all variables ranging from 0.771 to 0.839, exceeded the threshold of 0.6; the average variance extracted (AVE) of all latent variables was between 0.620 and 0.680, which is higher than 0.5; and the composite reliability (CR) values of all latent variables exceeded 0.7; confirming the satisfactory convergent validity of the scale (94). In Table 4, the values bold on the diagonal are the arithmetic square roots of the corresponding variables AVE. The values outside the diagonal indicate the correlation coefficients among the latent variables. According to the results, the arithmetic square root of AVE of all latent variables is higher than the correlation coefficients of the rows and columns in which they are located, which means that the discriminant validity among the latent variables is good (100).

The collected sample data were imported into AMOS software, and a model diagram was constructed to the analysis of model fit. The criteria for evaluating each model fit were referred to the recommendations of Dou et al. (101). The results are shown in Table 5. The ratio of chi-square to degrees of freedom (X²/DF) = 1.125, which is less than the recommended threshold of 3. The values of RMSEA, CFI, IFI, NFI, and GFI are 0.014, 0.992, 0.992, 0.930, and 0.920. All the fit indices are within the range of the recommended range, which indicates that the research model fits well with the collected data.

The critical ratio (CR) value is the ratio of the parameter estimate to the parameter estimate standard error (SE). If the CR > 1.96, then the parameter estimate passes the test at the significance level α = 0.05.

According to the C.R value and p-value shown in Table 6, among the 20 hypotheses, 17 hypotheses were found to be significant at α = 0.05 level. While the remaining three hypotheses did not pass the test. The specific test results are as follows: The information source has a significant negative effect on perceived risk. The information quality, information source, informational support, emotional support, technological security, technological facilitation, and self-efficacy have a significant positive effect on perceived value. Information quality, information source, informational support, emotional support, technological security, technological facilitation, and self-efficacy have a significant positive effect on health information discernment abilities. Perceived value and perceived risk have a significant positive effect on health information discernment abilities. However, the hypothesis of the effect of information quality on perceived risk did not pass the test. The influence of self-efficacy on perceived risk is not significant. The influence of perceived risk on perceived value is also not significant (Figure 2).

To categorize the sample data into different set affiliations, a data calibration operation must be performed on the questionnaire data before the configuration analysis. In this study, a 5-point Likert scale was chosen for the design of the measurement questionnaire, which had to be transformed into an affiliation scale between 0 and 1 (102). By calculating the mean value of the measurement items as the original data of each variable. fsQCA4.1 software was used to calibrate the integrated data, with anchor points set at (0.95, 0.5, 0.05), which means that 95% of the set is fully affiliated, 50% is the maximum fuzzy affiliation point, and 5% is not affiliated at all (103).

A single condition necessity test was carried out on the outcome variables, so as to explore whether there is a single necessary condition variable that leads to the outcome. It is generally considered that consistency >0.9 can be regarded as a necessary condition. The results are shown in Table 7. It can be seen that the consistency of Information Quality (IQ), Information Source (ISE), informational support (IST), Emotional Support (ES), Technological Security (TS), Technological Facilitation (TF), Self-Efficacy (SE), Perceived Value (PV), and Perceived Risk (PR) was <0.9, which indicates that the explanatory power of these nine single elements on the outcome variables is weak. They did not have a dominant influence on health information discernment abilities. Further exploration is needed to examine the sufficiency of the effects of configurations on the outcome.

The truth table was constructed based on the calibrated fuzzy set data table. The consistency threshold is set to 0.80. The case threshold is set to 3. The PRI consistency threshold is set to 0.75 (104). Typically, three kinds of solutions are generated after analyzing through fsQCA4.1: complex solutions, intermediate solutions, and parsimonious solutions. The parsimonious solutions and the intermediate solutions can effectively balance the discrepancy between theory and facts (105). Thus, this study combined the parsimonious solutions and the intermediate solutions as the analysis results. The condition variables are distinguished into core conditions and peripheral conditions through qualitative comparative analysis. There are 10 specific condition variable combination paths, which are divided into four types, as shown in Table 8.

According to Table 8, the consistency of the configurations ranges from 0.991 to 0.995, and the solution consistency level of the health information discernment abilities is 0.976, with a solution coverage of 0.574. This indicates that the collection of these 10 paths has a strong explanatory power of the health information discernment abilities. It is able to explain 57.4% of the results of the sample. By categorizing configurations with the same core conditions, four types of configuration patterns for health information discernment can be identified as follows:

In summary, it can be seen from the 10 sub-configurations that while nine antecedent conditions all appear in the same configuration at the same time. These configurations contain at least two influencing factors. It shows that the health information discernment abilities of OHCs users are not caused by a single influencing factor but rather by the interdependence of multiple factors. There are multiple configurations that can lead to the target results. Traditional SEM studies cannot explain the complex relationship between these factors and outcomes.

The robustness testing is necessary to ensure that the findings are generalizable. Schneider and Wagemann (106) proposed feasible ways of robustness testing, such as by adjusting the data calibration thresholds and the consistency thresholds, and by changing the frequency of cases. In this study, the consistency threshold was changed from 0.8 to 0.85. It was found that the parameter adjustments did not result in any changes in the configuration paths as well as in the fit parameters for consistency and coverage. Thus, the analysis outcomes was found to be relatively robust.