Sage Journals: Discover world-class research

Abstract

An important reason for the slowdown of online Q&A (Question and Answer) communities is that the number of users with knowledge input behavior is relatively small and the intensity of input behavior is decreasing, which is precisely the key to the sustainable development of communities. Therefore, it is of great significance to explore the factors affecting knowledge contribution behavior to maintain the sustainable development of the community. Based on the social exchange theory from the perspective of configuration, the theoretical model of the factors influencing knowledge contribution behavior is constructed by sorting out nine dimensions of variables from users’ self-improvement and social identity, writing a distributed crawler program to capture a total of 5,000 sample data from Zhihu community, using fuzzy set qualitative comparative analysis for empirical research. We found that users’ knowledge contribution behavior is influenced by multiple prefactors, there may be three different types of constitutive paths that influence users’ knowledge contribution behavior: self-improvement, social identity, and both. At the same time, users’ self-improvement and social identity play a linkage role and jointly influence knowledge contribution behaviors. From the perspective of configuration, the platform can find multiple influencing factors in various ways, so as to enhance the enthusiasm of users for knowledge contribution and promote the sustainable development of the online Q & A community. In addition, this paper further deepen the study of the factors influencing knowledge contribution behavior and develop new research perspectives for the study.

Keywords

Q&A community knowledge contribution behavior qualitative comparative analysis social exchange theory

Introduction

With the booming development of online Q&A communities, people’s growing knowledge needs have been greatly met here, making the continuous satisfaction of knowledge becomes a key element in the rapid development of online Q&A communities (Jin & Xu, 2020). As a network platform, the main function of online Q&A communities is to promote knowledge sharing among users.

Before users use the online Q&A community to ask questions to other netizens, they can first search for answers. The platform will provide a similar list of answered questions. When users are still not satisfied with these questions, they can ask questions, describe them in detail, and submit them to the system. The questions will be posted in the community, waiting for other users to answer, and ultimately the questioner will obtain answers from the answers. All Q&A information will eventually enter the knowledge base for users to ask and search, and complete knowledge sharing. This is the basic operating mechanism of online Q&A communities.

They break through the time and space limitations of traditional knowledge sharing, are characterized by convenience, openness, and low cost, and are widely accepted by users. With the continuous development of community technologies and applications, many users are accustomed to acquiring knowledge through online Q&A communities to solve the problems they face or improve their insights, thus making online Q&A communities a new platform for knowledge acquisition. However, in recent years, the activity level of users has gradually decreased and the development of online Q&A communities has gradually slowed down. According to research, most online Q&A communities have a typical 90-9-1 phenomenon, that is, 90% of users never contribute content, 9% contribute content occasionally, and 1% contribute the vast majority of content in the community. The key to the development of online Q&A communities lies in the continuous knowledge contribution behavior of users. Therefore, it is a problem that platform operators need to consider to study the influencing factors of users’ knowledge contribution behavior, take targeted measures to enhance users’ knowledge contribution behavior, and maintain the stable operation of the community.

In recent years, on the issue of factors influencing knowledge contribution behavior in online Q&A communities, different scholars have focused on different aspects, which can be roughly divided into two levels: scholars focusing on individuals believe that individuals’ internal and external motivations have a greater influence, such as factors like personal honor enhancement, self-efficacy, helpfulness, and perceived pleasure. Isabel and Minhyung (2021) used self-determination theory to construct a structural equation model from a new perspective, and explore the influence of gamification on the internal and external motivation of users to contribute knowledge; scholars focusing on the environment emphasize the influence of the external environment on users’ knowledge contribution behavior, such as elements like social identity, social influence, and group cohesion. X. H. Chen et al. (2021) found that the environmental characteristics of knowledge dissemination and the richness of knowledge content have a positive impact on the effect of knowledge dissemination.

Literature Review

Research on Knowledge Contribution Behavior in Online Q&A Communities

The improvement of community value and activity relies on the knowledge contribution behavior of users (Guan et al., 2018), which is a key factor for the sustainable development of online Q&A communities. At present, researches on knowledge contribution behavior are mainly focused on the study of its influencing factors and behavioral mechanisms. In terms of subject areas, recent studies are mainly based on psychology, management, communication, information ecology, sociology, organizational behavior, etc. In addition, some scholars also try to introduce theories from the health field into the research; in terms of research theories, relevant studies are mostly based on social cognitive theory, theory of planned behavior, social exchange theory, incentive theory, identity communication theory and social capital theory, etc. There are many studies on the influencing factors of user knowledge contribution behavior: One kind is research on non-configuration perspective. For example, Zhou and Wang (2020) used structural equation modeling to explore the influence of social influence mechanisms on user behavior; X. R. Zhang et al. (2020) argued that the external institutional environment has a positive impact on knowledge accumulation in organizations; the other category is to study multiple factors based on the configuration perspective. For example, Lu et al. (2022) explained the influencing factors of user knowledge contribution behavior at both individual and environmental levels with the help of qualitative comparative analysis methods. W. W. Zhang et al. (2021) explored the influence mechanism of divers’ and contributors’ knowledge contribution behaviors based on social learning theory; R. Liu and Yu (2022) introduced health domain theory to explore the complex causality of each variable on the results; Hong et al. (2020) explored the adoption mechanism of knowledge contribution behaviors from the perspective of information ecology; Zhou et al. (2020) analyzed the influence elements of users’ willingness to contribute knowledge with the help of social capital theory. In recent years, the discussion on the relationship between users’ willingness to contribute knowledge and behavior has also gradually become the focus of research in this field. Isabel and Minhyung (2021) used self-determination theory to construct a structural equation model from a new perspective to explore the influence of gamification on the internal and external motivation of users’ knowledge contribution behavior; Imlawi and Gregg (2020) used expectation value theory to explore the influence of incentives on practitioners’ continuous knowledge contribution in online health communities; Shi et al. (2021) explored the influence of learning on users’ knowledge contribution behavior with the help of social cognitive theory; N. Wang et al. (2022) investigated the influence of textual and non-textual feedback on users’ knowledge contribution behavior; Cai et al. (2022) predicted users’ willingness to share knowledge in online Q&A communities by using decision tree analysis; J. J. Li and Guo (2022) analyze the influencing factors of users’ knowledge contribution behavior through rooting theory; S. L. Gao et al. (2020) used negative binomial regression method to explore the influencing factors of knowledge contribution behavior.

In addition, there is also some literature devoted to other aspects of user knowledge contribution behavior. Among them, for knowledge dissemination: based on social exchange theory and convergence theory, X. H. Chen et al. (2021) found that the environmental characteristics of knowledge dissemination and the richness of knowledge content have a positive impact on the effect of knowledge dissemination. For the research on user answers: the answer quality evaluation index system was constructed by Shen et al. (2020) using a support vector machine model; Z. Chen et al. (2021) constructed an answer quality prediction model using FM algorithm, and the accuracy of answer selection was improved compared to the deep learning model; Guo and Bu (2021) used weighted gray correlation analysis to investigate the ranking method of answer relevance. In terms of user types and their relationships: X. H. Chen et al. (2020) constructed a network of respondent-focused relationships to study the impact of peer effects on users’ knowledge contribution behaviors; J. Q. Yang and Ye (2020) explored the relationship between users’ social connections and their information production using Lotka’s law. For different types of users, W. Z. Tang et al. (2020) distinguish whether answering users are experts or not based on Gaussian-Gamma Mixture Model (GGMM) clustering algorithm; Agouti (2022) used a graph-based association rule mining method to identify influential knowledge disseminators.

In addition, many scholars have conducted relevant research on user self-improvement and social identity. For example, H. Liu (2022) believe that self-interested motivation directly affects users’ willingness to contribute knowledge on social network platforms. In C. Chen and Hung’s (2010) study, self-efficacy factors, such as users’ knowledge reserves and expression ability, affect users’ knowledge contribution behaviors, and the higher the user’s self-efficacy, the more they are inclined to make knowledge contributions; Scholars, such as Wask et al. (2004), have proved through empirical research that reward, personal reputation, and reciprocity are all important factors in promoting users’ knowledge contributions; Bai (2022) revealed the formation mechanism of pro-environmental behavior from the perspective of social identity process, and proposed specific measures to promote people’s pro-environmental behavior based on theory, which shows that knowledge contribution behavior has a certain relationship with social identity.

At present, the existing literature research on user behavior in knowledge online communities mainly focuses on knowledge sharing and user-generated content, and there is less research on user knowledge contribution behavior. Based on social cognition theory, this study uses the QCA method to conduct empirical research on the basis of previous research, analyzes the influencing factors of knowledge contribution in online Q&A communities from a new theoretical perspective, and supplements the research content of influencing factors of knowledge contribution behavior of users in knowledge online communities. According to the conclusions, targeted promotion strategies are proposed to further contribute to the operation and management of the knowledge online community. To a certain extent, it fills the gap in the existing literature on the influencing factors of the knowledge contribution of individual online learners in the knowledge online Q&A community.

Research Review and Innovation Points

In a comprehensive analysis of the existing studies, there are many achievements in the research of knowledge contribution behavior at home and abroad, and the applied theories have become more and more extensive in the disciplinary fields, and the research on the influencing factors of knowledge contribution behavior has become more and more thorough, among which, the research based on the social cognitive field is particularly in-depth. This study makes improvements based on various previous studies and has the following advantages. First, since fuzzy set qualitative comparative analysis is a method commonly used in the study of community user behavior, it can be used to compare the differences and characteristics of different users in terms of knowledge contribution and to investigate the complex causal relationships between the groupings formed by different variables on the outcomes from a holistic perspective. Second, objective user data are used for the study, which avoids the differential relationship between user intention and behavior. Objective user data can make the data not affected by subjective factors, reflect the real information behavior of users, and also make research and decision-making more scientific, accurate, and reliable, avoiding the interference of subjective assumptions and personal bias. Therefore, this paper chooses to crawl 5,000 pieces of user-related data, which is a reasonable scientific range, which is not only large enough to objectively reflect the behavior of many target groups, but also moderate enough to avoid too much data. Third, Zhihu, one of the representatives of Chinese Internet online Q&A communities, was selected to crawl the data and has a large amount of user data, which has some universality. According to CIC, Zhihu is the largest Q&A-style online community in China. The user group on Zhihu is very wide, covering almost all industries and fields, so this study chose Zhihu as a research platform. Fourth, social exchange theory is used to simultaneously analyze the relationship between different variables and outcomes in the individual and the environment, this is because social exchange theory is a reinforcing principle advocated by behaviorist psychology that examines the behavior of social interaction between people in terms of reward and value. To sum up, this paper takes Zhihu users as the research object, crawls 5,000 pieces of data related to Zhihu users, and explores the multiple motives of users’ knowledge contribution behavior in online Q&A communities based on social exchange theory and with the help of fsQCA method. The research results can provide a reference for community operators to stimulate online learners’ willingness to participate in knowledge contribution and achieve accurate knowledge delivery and operation, improve the long tail effect of online learners’ knowledge contribution, promote the development of community collaborative learning functions, disciplinary tools, and online education model innovation, and provide a basis for community learning and knowledge contribution in designing and optimizing online learning incentive mechanisms.

Knowledge contribution behavior is the behavior of users discovering problems and giving solutions, and the process of users choosing answers to solve problems is called knowledge adoption behavior (H. Li & Zhang, 2014). The knowledge contribution behavior of users in the social Q&A community has attracted attention from all walks of life and has produced a large number of research results (Liao, 2020), but in the past, researchers rarely explored the factors affecting knowledge contribution behavior and its mechanism from the perspective of configuration. The main source of Q&A community resources is the user’s knowledge contribution (L. Wang, 2022), and the content of the Q&A community basically comes from the user’s input, and it is with the user’s knowledge contribution that these high-quality and high-content contents can attract other users to access and use the Q&A community. The contribution of users to knowledge determines the degree of development of the Q&A community (Z. F. Zhang & Li, 2010). This study combines the qualitative comparative analysis of fuzzy sets, discusses from the perspective of social exchange theory, analyzes the impact of different factors on users’ knowledge contribution behavior in the Q&A community, and puts forward suggestions for improvement based on the actual situation, which can guide the operators and designers of the Q&A community to optimize the platform, provide methods for the knowledge aggregation process of the Q&A community, put forward favorable conditions for the development of the platform, and lay the foundation for the sustainable development of the Q&A community.

In summary, this study uses individual online learners from the social Q&A community Zhihu as the study population, the fuzzy set qualitative comparative analysis method is selected to study and look for multiple equivalent paths of high and low user knowledge contribution behavior, so as to reveal the systematic influence of user self-improvement and social identity on user knowledge contribution behavior from the perspective of configuration. Users ‘self-improvement and social identity play a linkage role, and both affect users’ knowledge contribution behavior. Among them, users’ sense of self-improvement has a greater impact on knowledge contribution behavior. The impact of this study on this field will be reflected in the following aspects: First, it is conducive to finding incentive mechanisms to improve users’ knowledge contribution behavior. Second, it is conducive to the industry field to provide deep and personalized knowledge services based on this research. Third, it is conducive to the platform operators to create a good community interaction environment based on the configuration perspective and promote the sustainable knowledge contribution behavior of users. This will greatly improve users ‘participation in knowledge contribution, help the industry to build good community identity evaluation indicators, and promote users’ sustainable knowledge contribution behavior.

Theoretical Foundation and Research Design

Social Exchange Theory

The social exchange theory was proposed by the American sociologist Homans G C, and modified and supplemented by Blau P. Social exchange theory suggests that exchange is the act of people interacting with each other and making a return (Bierstedt & Blau, 1965). Social interaction is the process of exchange (Johannisson, 1987). When a person provides the most rewards to the user, that person creates the greatest attraction to the user. At the same time, as an exchange, the user also needs to pay a certain price. Social exchange theory holds that people are primarily concerned with the total outcome of the exchange process, that is, whether, in the aggregate, the relationship results in more reward than cost, or more cost than reward.

Analyzed from the perspective of social exchange theory, users’ knowledge contribution behavior belongs to the payment of time and energy, which is the price of exchange by users; and the reward received by users can be roughly divided into two categories: self-improvement of users and social recognition obtained by users. Users’ knowledge contribution behavior is affected by the amount of rewards they receive, and different types of rewards do not measure the same for different types of users. Therefore, this paper classifies users according to different antecedent configurations of high and low knowledge contribution behavior and proposes targeted measures to improve the knowledge contribution behavior of different types of users.

Knowledge Contribution Behavior

Knowledge contribution behavior refers to the behavioral process of knowledge creators to provide and create knowledge (Jing et al., 2015). The knowledge contribution behavior of users is a key element for the sustainability of online Q&A communities. In Zhihu, both professional knowledge and entertainment knowledge are provided. Through the acts of writing, asking questions, and reading, users accumulate their knowledge and provide the community with knowledge resources for sustainable development.

Research Objectives

This paper focuses on the combined influence of antecedent and causal variables of users’ knowledge contribution behavior in online Q&A communities, and tries to answer the question “Given the existence of many knowledge acquisition channels, how do self-improvement and social identity combine to influence users’ knowledge contribution behavior for different types of users?.” By answering this question, this paper hopes to propose targeted measures to enhance the knowledge contribution behavior of different types of users to maintain the intensity of their contribution behavior and ensure the sustainable development of online Q&A communities.

Research Methods and Steps

Research Methods

This paper intends to study the influencing factors of users’ knowledge contribution behavior in online Q&A communities. In the Q&A communities, the users’ knowledge contribution behavior is affected by the synergy of many factors such as self-improvement and social identity and other factors and the influencing factors of high and low knowledge contribution behavior are not a simple symmetrical relationship (Ji et al., 2022). Knowledge contribution behavior is a complicated decision-making process with global, complexity and causal asymmetry. The qualitative comparative analysis method, namely QCA, usually studies the combination of possible causes of occurrence from the perspective of the whole. Configuration refers to the logical combination of a set of causal conditions (Y. Liu et al., 2017). QCA is mostly used to study the configuration relationship, which is suitable for analyzing the complexity of antecedent configuration. It is an important tool to study the complexity, global, and causal asymmetry of user behavior in a digital environment (Sun, 2019). Therefore, compared with other methods, QCA has more advantages in solving the problem of users’ knowledge contribution behavior.

According to the type of variable, QCA can be divided into csQCA (clear set), mvQCA (multi-value set), and fsQCA (fuzzy set), which has certain advantages in dealing with fixed ratio variables and fixed distance variables compared to the other two analysis methods (Du & Jia, 2017). The core problem of fsQCA is the adequacy and necessity of determining the condition variable and the result variable, if a combination is a sufficient condition for the result, then this combination condition is one of the reasons for the result, that is, the result can be achieved through this combination of conditions (W. Gao et al., 2018). The fsQCA is now gradually being addressed in the study of user behavior, for example, Lu et al. (2020) studied the factors influencing users’ willingness to pay for knowledge in professional virtual communities, but these studies were mostly based on questionnaires (Lu et al., 2020). Therefore, according to the characteristics of the study, we choose the method of qualitative comparison of fsQCA fuzzy sets, and due to some differences between users’ willingness to contribute knowledge and their behaviors, this paper improves on previous research by using a Python web crawler to crawl Zhihu users’ data and constructing a model based on objective user data, thus making the research results more convincing.

For the above reasons: at the theoretical level, this paper uses social exchange theory as a research lens to construct a research framework on the constructs of factors influencing the knowledge contribution behavior of users of online Q&A communities; at the practical level, the data were processed using fsQCA3.1b software to investigate the constructs of factors influencing user knowledge contribution behavior in Zhihu. The conclusions drawn are useful for online Q&A communities to analyze the knowledge contribution behavior of different types of users, to further enrich theories related to user behavior in online Q&A communities, and to provide targeted reference strategies for enhancing users’ knowledge contribution behavior.

Research Steps

The research steps are as follows: First, the antecedent variables and outcome variables are determined by the theoretical analysis of social exchange theory. The antecedent variables are 9 variables, including the number of Likes, Thanks, Questions, Favorites, Followers, Idols, Topics followed, Columns followed, and Questions followed of Zhihu users. These nine variables are represented by L, T, Q, FA, FO, I, TF, CF, and QF respectively, and they are divided into two categories: self-improvement (SIM) and social identity (SID). The outcome variables are user’s knowledge contribution behaviors, which are recorded as the sum of the number of answers, articles, columns, and ideas. The results are shown in Figure 1. Second, collect data and calibrate the sample data. Third, the necessity of a single antecedent variable is tested. Fourth, construct the truth table, analyze the multiple concurrent causalities between variables, further analyze the configuration of knowledge contribution behavior and classify it.

Figure 1.

Variable causal model.

Data Collection and Data Calibration

Variable Measurement

This chapter mainly explains the specific meaning of proper nouns in the Zhihu community and their meanings.

(1) The number of likes is the number of answers posted by a user that are approved by an unspecified majority of other users and correspondingly “liked,” from which the user obtains feedback from others in the community about the level of approval.

(2) The number of thanked is the number of answers posted by a user that other unspecified users consider valuable and “thank,” from which the user can obtain feedback from other users in the community on the degree of value of the answer.

(3) The number of questions is the number of users who think about a certain issue, ask questions, and seek answers from the community. Users learn that they do not understand a certain issue and ask questions to others to get corresponding answers, thus achieving self-improvement.

(4) The number of favorites refers to the number of answers of the communicator that have a certain value to the receiver of the information and receive the corresponding “favorite” operation from the receiver of the information.

(5) The number of followers refers to the number of other users who follow the communicator user, which represents the first dissemination range of the user’s published information in the society, from which the user can obtain the recognition from other users in the society and the user’s ability to spread information and influence others.

(6) Number of followers refers to the number of other users a user follows. After browsing the content of the knowledge sharing platform, users mark the high-quality users who post in it to prevent missing relevant content that is beneficial to self-improvement.

(7) The number of topics followed is the number of topics that a user follows for the operation of the topic he or she is interested in. A higher number of topics followed largely indicates that the user has a greater willingness to answer questions.

(8) The number of columns followed is the number of columns that the user follows for the column he is interested in; the number of questions followed is the number of questions that the user follows for a specific issue in the topic or column.

(9) The number of topics followed, the number of columns followed and the number of questions followed are important indicators of user engagement, which can directly reflect users’ participation in and contribution to the Zhihu community.

Data Calibration

In this paper, samples of the data related to Zhihu users were crawled and 5,000 of them were randomly selected as the study case of QCA. The data distribution of the sample is wide, and the affiliation range of fsQCA is 0 to 1, so the original data of the sample needs to be transformed into fuzzy set data in the range of 0 to 1 (Ragin et al., 2008). Meanwhile, since the direct choice of 1, 0.5, and 0 as thresholds may yield meaningless results, this paper uses percentile calibration of the data based on the results of Ragin and Strand (2005), which is referenced in a large number of studies. The criteria for applying percentile calibration vary from study to study, where most studies use the method of upper and lower quartiles proposed by Fiss (2011) for fully affiliated and fully unaffiliated thresholds to determine them, and the intersection point has two main choices of mean and median.

After fully considering the distribution characteristics of the sample data, the upper and lower quartiles were chosen as the calibration thresholds for complete affiliation and complete disaffiliation in this study with reference to the research results of Fiss. Meanwhile, due to the uneven distribution of the sample data and the mean value being influenced by the extreme value, the median is used as the crossover point in this paper. The calibration calculation is completed by using the calibrate function in fsQCA. The calibration thresholds of the variables are shown in Table 1.

Table 1.

Variable Calibration Threshold.

Variables	Fully unaffiliated	Intersections	Fully affiliated
L	0	8	89.25
T	0	2	19
Q	0	1	2
FA	0	2	19
FO	0	3	19
I	6	21	65
TF	2	10	25
CF	0	1	3
QF	8	30	101
KC	3	12	39

Note. L = likes; T = thanks; Q = questions; FA = favorites; FO = followers; I = idols; TF = topics followed; CF = columns followed; QF = questions followed; KC = knowledge contribution.

Analysis and Discussion

Necessity Analysis

This chapter mainly explains the specific meaning of proper nouns in the Zhihu community and their meanings. To analyze fuzzy sets, it is first necessary to conduct a necessity analysis of all independent antecedents, in order to determine whether each antecedent variable or the non-set of each antecedent variable is a necessary condition that affects the user’s knowledge contribution behavior. Necessity analysis measures the consistency and coverage of the antecedent variables with the outcome variable in terms of whether each antecedent variable can individually cause that outcome to occur, where consistency is used to reflect the extent to which each antecedent variable explains the results. Coverage is used to reflect the necessity of what percentage of cases can explain the corresponding antecedent variables. The formula for calculating the second indicator is expressed by equations (1) and (2), respectively:

C onsis tency (Xi < < Yi) = \sum min (Xi, Yi) / \sum Xi

(1)

C overage (Xi < < Yi) = \sum min (Xi, Yi) / Yi

(2)

If the consistency detected by necessity is greater than or equal to 0.9, this antecedent variable is necessary for the outcome; If the detected consistency is less than 0.9 but not less than 0.8, that antecedent variable is found to be nearly necessary for the outcome (Ragin, 2000). The results of the necessity test are shown in Table 2. In this table, the consistency between the presence and absence of the antecedent variable is less than 0.9. It shows that user knowledge contribution behaviors are not the result of a single cause. For high-degree knowledge contribution behaviors, it is a near necessity that the consistency of the number of likes and thanks by the user is greater than 0.8 and less than 0.9. For low-level knowledge contribution behavior, it is almost a necessity that the number of likes, thanks, favorites and questions followed by users is greater than 0.8 and less than 0.9. This result fully illustrates the influence of social identity on users’ knowledge contribution behavior. From the necessity test results, the extraction of antecedent conformations by conditional combination analysis is necessary.

Table 2.

The Need for a Single Antecedent Variable.

Antecedent variables	Outcome variable: knowledge contribution behavior		Outcome variable: ~knowledge contribution behavior
Antecedent variables	Consistency	Coverage	Consistency	Coverage
L	0.803378	0.816240	0.290982	0.324057
~L	0.334733	0.301050	0.835009	0.823172
T	0.806096	0.820935	0.283517	0.316489
~T	0.328876	0.295169	0.839590	0.825969
Q	0.640181	0.679170	0.359778	0.418378
~Q	0.451798	0.391666	0.724110	0.688073
FA	0.752065	0.787993	0.296555	0.340589
~FA	0.370690	0.324677	0.815404	0.782838
FO	0.774439	0.748222	0.355675	0.376665
~FO	0.354860	0.334428	0.762273	0.787433
I	0.649771	0.640852	0.422057	0.456274
~I	0.448707	0.414625	0.667791	0.676382
TF	0.607198	0.586021	0.476013	0.503570
~TF	0.485625	0.458146	0.608680	0.629434
CF	0.587669	0.626865	0.405289	0.473875
~CF	0.506788	0.437398	0.680876	0.644135
QF	0.790441	0.783393	0.316839	0.344197
~QF	0.338296	0.311184	0.800619	0.807244

Conditional Portfolio Analysis

This paper uses conditional portfolio analysis to analyze whether the combination of conditions is the core condition in the user’s knowledge contribution behavior according to its impact on the results under the combination of different conditions. Conditional Portfolio Analysis, measured by Consistency and Coverage, can be used to analyze the effect of different combinations of conditions on results. By running the “fuzzy truth table algorithm” command in fsQCA, the simple, intermediate, and complex solutions of the resultant variables were obtained. Since complex solutions are not considered and not observed, intermediate solutions contain “logical remainders” where theory and substantive knowledge are consistent. In this paper, the intermediate solution for analysis is chosen, combining simple solutions with intermediate solutions to analyze whether the condition is a core condition. The specific results are shown in Tables 3 and 4. Among them, we classify the combination of paths such as A1 and B1 according to the impact of different core conditions on the user’s high-level knowledge contribution behavior and low-level knowledge contribution behavior paths. To make the research results more visual and intuitive, the symbolic expression form proposed by Ragin and Fiss is chosen in this paper, that is, “•” indicates the presence of the core condition, “●” indicates the presence of the edge condition, “⊗” indicates the lack of the core condition, “•” indicates the lack of the edge condition, and blank indicates that the condition is dispensable.

According to the results of the analysis in Table 3 and 4, it is clear that:

Table 3.

Antecedent Constructs of High-Level Knowledge Contribution.

Table 4.

Antecedent Constructs of Low-Level Knowledge Contribution.

Under the combined influence of users’ sense of self-improvement and social identity, there are 10 conditional combination paths for users’ high level of knowledge contribution behavior, respectively:

1.L*T*FA*FO*QF;2.L*T*Q*~TF*~CF*QF;3.L*T*Q*~I*~CF*QF;4.L*T*Q*FA*I*QF;5.L*T*Q*FO*I*QF;6.L*T*Q*FA*TF*QF;7.T*Q*~FO*~I*~TF*~CF*QF;8.L*T*FA*~I*~TF*~CF*QF;9.T*Q*FA*~I*~TF*~CF*QF;10.L*T*Q*I*TF*CF*QF. A1 indicates that the core conditions of L, T, FA, FO, and QF exist as one of the 10 paths for users’ high-level knowledge contribution behavior. There are seven conditional combination paths for user’s low-level knowledge contribution behavior, respectively:

1.L*T*Q*FA*QF;2.L*T*FA*FO*I*QF;3.L*T*Q*FO*TF*QF;4.L*T*Q*FO*I*CF*QF;5.L*T*Q*FO*~I*~CF*QF;6.T*Q*FA*FO*I*~TF*~CF*QF;7.L*T*Q*FA*FO*~I*~TF*~CF. B1 represents one of the seven paths of user-level knowledge contribution behavior, which does not exist the core conditions of L, T, Q, F.

By running the “fuzzy truth table algorithm” command in fsQCA, several different paths of users’ knowledge contribution behaviors are obtained, that is, the user’s knowledge contribution behavior is affected by a combination of many different antecedent variables. But all show the “different paths” of the construct perspective. It also demonstrates that for different types of users, different types of rewards have different effects on knowledge contribution behavior. The results show that the consistency scores of all the constructs for both knowledge contribution behaviors reached 0.9, which is above the acceptable threshold of 0.75, indicating the good reliability of these constructs. Knowledge contribution behavior is indeed greatly influenced by these antecedent variables. Meanwhile, the total coverage was 0.633016 and 0.644996, respectively, indicating that the combination of conditions explained the results strongly. Can more comprehensively reflect which combinations of antecedent variables are affected by the user’s knowledge contribution behavior.

Antecedent Construct Analysis of Knowledge Contribution Behavior

Considering the comparative advantages of fsQCA in solving complex relationships or partial affiliation problems regarding different variables to outcomes, fsQCA was selected as a specific analytical tool in this paper. fsQCA places variable values within the interval of [0,1], enabling more precise assignment of continuously changing condition variables (Hao et al., 2023). The data were imported into fsQCA3.1b software, and the antecedent constructs of user knowledge contribution behavior were obtained by fuzzy set necessity detection and truth table analysis, and the constructs revealed multiple equivalent motivational paths of high and low user knowledge contribution behavior. The results show that the core conditions that lead to high and low user knowledge contribution behaviors are not the same, and the combinations of antecedent configurations are not completely opposed, which reflects the asymmetry of user knowledge contribution behaviors and further illustrates the necessity of qualitative comparative analysis.

Antecedent Constructs of High Degree Knowledge Contributing Behavior

The antecedent constructs of high degree knowledge contribution behaviors are shown in Table 3, and the results show that there are 10 different paths to obtain high degree knowledge contribution behaviors. It can be roughly divided into the following three configurations: First, the user self-improvement-oriented configuration, which consists only of the A10 path. In such configurations, the self-improvement element of users is more often used as a core causal condition for knowledge contribution behavior, suggesting that gaining knowledge is more important to such users than gaining recognition as a reward. Second, the user’s social identity-based conformation. It consists of five paths, A1, A4, A6, A8, and A9. In this construct, social identity elements such as the number of likes and thanks appear more as core causality conditions, and for such users, the honorary rewards received as rewards are of greater importance. Third, the two-component type. This configuration consists of four pathways, A2, A3, A5, and A7. In the opinion of such users, self-improvement and social recognition are of roughly equal importance as rewards for the cost of the act of contributing knowledge. From the perspective of social exchange theory, the identification among users drives the dissemination of knowledge among them; therefore, this result is consistent with social exchange theory.

Meanwhile, from the results of the combination of antecedent constructs with higher knowledge contribution behavior, the number of concern questions is a core causality condition common to 10 constructs, the number of liked and thanked is a core causality condition present in 8 constructs, the number of questions is a core causality condition present in 7 constructs, and the number of answers favorited is a core causality condition present in 5 constructs. This reflects that when users get self-improvement and social identity enhancement, they tend to make more knowledge contribution behaviors to the platform in exchange, which is consistent with the content of social exchange theory. In contrast, when users receive more knowledge as payment, they will generally increase their knowledge contribution behavior, which reflects that the increase in users’ sense of self-improvement will increase their knowledge contribution behavior to a greater extent compared to their sense of social recognition. This is similar to the findings of X. Y. Yang et al. (2020).

At the same time, in users’ sense of self-improvement, the number of topics and columns is less often a core causality condition than the number of questions and issues raised, and in some paths, it is even missing as a core causality condition, which reflects that users prefer to read fragmented information and start from a specific issue detail in a certain field than to understand a certain field and topic as a whole. In this regard, it has been shown that the attractiveness of information content has a significant impact on the effectiveness of information dissemination in large-scale social networks (Han et al., 2014). In terms of social recognition, the number of likes and thanks received from user responses has a greater impact on user knowledge contribution behavior than the number of fans (Cao et al., 2023; Khansa et al., 2015).

Antecedent Constructs of Low-Degree Knowledge Contribution Behavior

The antecedent constructs of low-degree knowledge contribution behavior are shown in Table 4, and the results show that there are seven different paths to obtain high-degree knowledge contribution behavior. They can be roughly divided into the following three configurations: First, the user self-improvement-oriented configuration, which consists of two paths B3 and B4. In this type of configuration, the user’s self-improvement element is more taken as the core causal condition for knowledge contribution behavior, indicating that for this type of users, when the acquired knowledge is not enough as a reward, the users tend not to engage in knowledge contribution behavior even if they have received some social recognition. Second, the user’s social identity-based construct. It consists of three paths, B1, B2, and B7. In this configuration, social identity elements such as the number of likes and thanks appear more as core causality conditions. For such users, the honorary rewards received as rewards are more important, and users do not tend to engage in knowledge contribution when their social identity is not satisfied. Third, the two-component type. This configuration consists of two paths, B5 and B6. In this type of users’ view, self-improvement and social identity are both rewards for the cost of knowledge contribution behavior, and either one is indispensable, and their importance is roughly equal.

From the results of the combination of antecedent constructs with lower knowledge contribution behavior, the core absence of the number of thanks is common to 7 constructs, the core absence of the number of questions asked and the number of questions followed is common to 6 constructs, the absence of the number of likes, favorites, followers, and followers is common to at least 4 constructs, and the number of topics followed and the number of columns followed exist as core causality conditions in some paths. This reflects that when users do not receive sufficient self-improvement and social recognition, they do not contribute knowledge.

The results show that there are 10 acquisition paths for high-degree knowledge contribution behavior, which are divided into three configurations according to different core conditions. There are 7 acquisition paths for low-level knowledge contribution behavior, which are also divided into 3 configurations according to different core conditions. It can be seen that the user’s knowledge contribution behavior is affected by multiple factors at the same time, resulting in a constitutive path of knowledge contribution behavior. The experimental results tend to be the main influence of user social identity, and from the conclusion, the influence of user self-improvement and improvement is more important than that of social identity.

Conclusion and Insight

Research Findings

Users’ knowledge contribution behavior is the key to the development of online Q&A platforms. This paper selects the fuzzy set qualitative comparative analysis method to study the complex influence of each combination of antecedent variables on users’ knowledge contribution behavior, finds multiple equivalent paths of high and low user knowledge contribution behavior, and reveals the systematic influence of user self-improvement and social identity combined on users’ knowledge contribution behavior from a constitutive perspective.

From the above study, it is clear that users’ knowledge contribution behaviors are influenced by multiple antecedents concurrently, and there are probably three different types of constitutive paths that influence users’ knowledge contribution behaviors: self-improvement, social identity, and both. For different types of users, the importance of each element of self-improvement and social identity as rewards varies. For self-improvement users, gaining knowledge as a reward is more influential than social identity; for social identity users, the social recognition gained from the answer is more motivating for users to engage in knowledge contribution behavior. As for social recognition, the number of thanks and likes received by users’ answers is more important to users than the number of macro followers; for users who have both, gaining knowledge and social recognition are both motivating for their knowledge contribution behaviors.

From an overall perspective, users’ self-improvement and social identity play a linkage role, and both influence users’ knowledge contribution behavior simultaneously, such that the negative effect of the phenomenon of less self-improvement on users’ knowledge contribution behavior is reduced by the effect of honor reward. Compared with social identity, users’ sense of self-improvement has a greater impact on knowledge contribution behavior.

Online Q&A communities are important platforms for information exchange, transmission, and sharing, allowing users to connect and share information with others (Jia et al., 2022; Y. Zhang et al., 2019). The in-depth study of user knowledge contribution behavior in the online Q&A community is conducive to clarifying the influencing factors of user knowledge contribution behavior, objectively and scientifically guiding network managers to improve the website, providing a reference for community operators to stimulate users’ willingness to participate in knowledge contribution, improving the life cycle of users, and maintaining the sustainable prosperity and development of the online Q&A community.

Practice Inspiration

From the research results, the following suggestions are proposed to enhance users’ knowledge contribution behaviors: first, diversified incentive mechanisms are designed to enhance users’ knowledge contribution behaviors in terms of both material and spiritual incentives. On the one hand, a knowledge payment mechanism can be introduced to provide material incentives to users. On the other hand, we can adopt a gamification model and design point acquisition methods such as check-in, tasks, and quizzes to enhance users’ rank and provide spiritual incentives to users. However, it should be noted that the difficulty of points acquisition should be set appropriately; no difficulty will easily lead to the proliferation of indiscriminate answers in the community, while too much difficulty will discourage users from contributing knowledge. Second, pay attention to the motivation of users’ knowledge contribution behavior, and provide in-depth and personalized knowledge services. Multiple levels of users’ knowledge need to require communities to improve their software and hardware conditions to varying degrees (X. X. Zhang et al., 2018). Communities should take proactive measures to improve the quality of users’ answers to questions. For example, communities can encourage users to publish high-quality responses by rewarding them with different levels of quality; on the other hand, they can filter answers to avoid valuable information being overshadowed by a large number of invalid information. The platform should strengthen the accuracy of information retrieval and provide personalized services to different users. For example, the platform can accurately refine question tags, automatically index and semantically expand questions, and carry out multi-label classification of questions and sentences (X. B. Tang & Liu, 2021), transform information services into knowledge services, and push relevant answers to user concerns. Third, create a good community interaction environment. Due to the anonymity of the network platform, users dare to express their views on the Internet, while enriching the accumulation of community knowledge, personalized views may also lead to the occurrence of quarrels, in the face of this situation, managers should improve the interactive content screening system, reduce low-quality interactive behavior, can take measures to gradually open up interactive permissions and set tasks to improve the level according to different levels, stimulate the positive emotional expression of community members, and promote the user’s knowledge contribution behavior.

Research Limitations and Prospects

In terms of data collection, the script written in this paper uses crawlers to capture 5,000 sample data of Zhihu community for empirical research, and does not use other forms of data collection, which may lead to missing variables and unable to control some factors that may affect user knowledge contribution, such as platform service factors, community environment factors, etc. Although this paper uses a fuzzy set qualitative comparative analysis method and uses objective user data, there are still certain shortcomings. In follow-up studies, data can be collected in different forms by combining secondary data with questionnaire data or experimental data to make the data more convincing.

In terms of research samples, this paper selects users in the high-quality online Q&A community Zhihu of the Chinese Internet as the research subjects, and whether the research results are also applicable to other platforms remains to be verified. There may be certain differences in user preferences in different online Q&A communities, and user characteristics and behavior orientation will be different. Future research can consider exploring different types of Q&A communities and different cultural backgrounds, increase the sample size and scale, and improve the level of sample quality based on the premise of sufficient material conditions and time.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The authors acknowledge the financial support from the Humanities and Social Science Research Project of Hebei Education Department (JCZX2023005)

Ethical Approval

The authors declare that the research was conducted in the absence of any animal and human studies.

ORCID iD

Xin Feng

Data Availability Statement

Data sharing not applicable to this article as no datasets were generated or analyzed during the current study.

References

Agouti

(2022). Graph-based modeling using association rule mining to detect influential users in social networks. Expert Systems with Applications, 202, 117436.

Bai

X. L.

(2022). Pro-environmental behavior in groups: Based on the perspective of social identity. Psychological Science, 45(2), 439–445.

Bierstedt

Blau

P. M.

(1965). Exchange and power in social life. American Journal of Sociology, 30(5), 789.

Cai

Yang

Y. Y.

Shi

W. D.

(2022). A predictive model of the knowledge-sharing intentions of social Q&A community members: A regression tree approach. International Journal of Human–Computer Interaction, 38(4), 324–338.

Cao

Zhu

Qiu

(2023). Consequences of information feed integration on user engagement and contribution: A natural experiment in an online knowledge-sharing community. Information Systems Research. Institute for Operations Research and the Management Sciences (INFORMS), 35(3), 1114–1136.

Chen

Hung

(2010). To give or to receive? Factors influencing members’ knowledge sharing and community promotion in professional virtual communities. Information & Management, 47(4), 226–236.

Chen

X. H.

Y. T.

(2021). Network reputation, action similarity and knowledge dissemination effect of knowledge disseminators in question-and-answer communities. Journal of Management, 18(5), 741–750.

Chen

X. H.

Zhou

Y. C.

(2020). A study on knowledge contribution behavior of respondents in Zhihu Q&A community influenced by peer effect. Journal of Intelligence, 39(4), 450–458.

Chen

J. W.

(2021). Research on answer quality prediction for community Q&A in chemical industry. Journal of Northeast Normal University (Natural Science Edition), 53(3), 81–88.

10.

Y. Z.

Jia

L. D.

(2017). Configuration perspective and qualitative comparative analysis (QCA): A new path in management research. Management World, (6), 155–167.

11.

Fiss

P. C.

(2011). Building better causal theories: A fuzzy set approach to typologies in organization research. Academy of Management Journal, 54(2), 393–420.

12.

Gao

S. L.

Sun

X. Y.

Chen

J. B.

Guo

Y. L.

(2020). Study on knowledge sharing behavior of users in social Q&A community. Journal of Service Science and Management, 13(4), 688–701.

13.

Gao

J. Z.

(2018). The impact path of entrepreneurship policy on urban entrepreneurship: A qualitative comparative analysis based on fuzzy sets. Journal of Technology and Economics, 37(4), 68–75.

14.

Guan

Wang

Jin

Song

(2018). Knowledge contribution behavior in online Q&A communities: An empirical investigation. In Computers in Human Behavior, 81, 137–147.

15.

Guo

S. L.

(2021). Research on the usefulness ranking method of answers in social Q&A communities by integrating Word2vec and WGRA—Crip Q&A as an example. Library Intelligence Work, 65(23), 126–135.

16.

Han

Zhuang

F. Z.

Shi

Z. Z.

Xiang

(2014). Energy model for rumor propagation on social networks. Physica A: Statistical Mechanics and Its Applications, 394, 99–109.

17.

Hao

W. Q.

Meng

Duan

Z. Z.

(2023). The theoretical logic and configuration path of urban digital career change from the perspective of dynamic capability——Based on the qualitative comparative analysis of fuzzy set in key cities in China. E-Government, 7, 73–86.

18.

Hong

Mao

A. T.

(2020). Study on the adoption mechanism of knowledge contribution from open innovation community users. Journal of Modern Information, 40(5), 33–40.

19.

Imlawi

Gregg

(2020). Understanding the satisfaction and continuance intention of knowledge contribution by health professionals in online health communities. Informatics for Health & Social Care, 45(2), 151–167.

20.

Isabel

O. J. D. A.

Minhyung

(2021). Gamification effects on users’ motivation to contribute knowledge in a Portuguese Q&A community. Aslib Journal of Information Management, 73(4), 578–599.

21.

H. Y.

G. J.

Liang

Y. K.

(2022). Is action more powerful than voice?—Research on user interaction and continuous contribution behavior of user creativity update in open innovation community. Management Review, 34(4), 80–89.

22.

Jia

Lin

Qin

Pan

Zhou

(2022). Impact of monetary and non-monetary social functions on users’ knowledge-sharing intentions in online social Q&A communities. Internet Research, 33(4), 1422–1445.

23.

Jin

(2020). Examining the factors influencing user satisfaction and loyalty on paid knowledge platforms. Aslib Journal of Information Management, 73(2), 254–270.

24.

Jing

Tao

Fan

X. C.

(2015). Patient value co-creation in online health communities: Social identity effects on customer knowledge contributions and membership continuance intentions in online health communities. Journal of Service Management, 26(1), 72–96.

25.

Johannisson

(1987). Beyond process and structure: Social exchange networks. International Studies of Management & Organization, 17(1), 3–23.

26.

Khansa

Liginlal

Kim

S. S.

(2015). Understanding members’ active participation in online question-and-answer communities: A theory and empirical analysis. Journal of Management Information Systems, 32(2), 162–203.

27.

Zhang

S. Y.

(2014). A review of research on the information needs of network users at home and abroad. Library and Information Service, 58(5), 111–123.

28.

J. J.

Guo

S. L.

(2022). Factors influencing users’ knowledge adoption behavior in social Q&A communities in a cognitive perspective: An exploratory analysis based on rooting theory. Intelligence Science, 40(2), 91–98.

29.

Liao

J. Y.

(2020). Review of influencing factors of knowledge contribution behavior of users in social Q&A community. Industry and Technology Forum, 19(3), 95–96.

30.

Liu

(2022). Research on the influencing factors of knowledge contribution behavior of academic social network users [Dissertation, Liaoning Normal University].

31.

Liu

J. Q.

(2022). Knowledge-blockchain of distributed innovation from the perspective of blockchain technology. Information Science, 40(3), 45–54+62.

32.

Liu

József

Vassilis

H. X.

(2017). Applying configurational analysis to IS behavioural research: A methodological alternative for modelling combinatorial complexities. Information Systems Journal, 27(1), 59–89.

33.

Zhang

X. X.

Y. W.

Zhang

(2022). Research on multiple motivations and effects of users’ knowledge contribution in professional virtual community. Library Development, 2, 131–140+158.

34.

Zhang

X. X.

Zhang

L. M.

Guo

S. L.

(2020). Research on the configuration of factors influencing users’ willingness to pay for knowledge in online question answering Community from the perspective of rationality and deviation. Library and Information Service, 64(19), 89–98.

35.

Ragin

C. C.

(2000). Fuzzy-set social science, University of Chicago Press.

36.

Ragin

C. C.

Strand

S. I.

(2005). Using qualitative comparative analysis to study causal order-comment on Caren and Panofsky. Sociological Methods and Research, 36(4), 431–441.

37.

Ragin

C. C.

Strand

S. I.

Rubinson

(2008). User’s guide to fuzzy-set/qualitative comparative analysis. University of Arizona, 3(9), 87–101.

38.

Shen

S. Y.

Liu

J. Y.

(2020). Research on the optimization method of answer quality evaluation system of question and answer community. Data Analysis and Knowledge Discovery, 5(2), 83–93.

39.

Shi

C. C.

Fan

W. G.

Qiu

L. F.

(2021). How do learning effects influence knowledge contribution in online Q&A community? A social cognitive perspective. Decision Support Systems, 149, 113610.

40.

Sun

Y. Q.

(2019). Research on information behavior based on configuration perspective: Research topic, theoretical construction and analysis method. Journal of Information Resources Management, 9(3), 4–17.

41.

Tang

W. Z.

H. S.

Zhang

(2020). Domain problem-solving expert identification in community question answering. Expert Systems, 37(5), 12582.

42.

Tang

X. B.

Liu

J. N.

(2021). Research on automatic indexing of questions in Q&A community based on BERT and TF-IDF—A case study of Q&A community in Jintou network. Information Science, 39(3), 3–10.

43.

Wang

(2022). Research on influencing factors and mechanism of user knowledge contribution behavior in open innovation community [PhD thesis, Jiangsu University of Science and Technology].

44.

Wang

Liu

Xiao

S. S.

(2022). Which feedback matters? The role of expressions and valence in continuous high-quality knowledge contribution in the online Q&A community. Decision Support Systems, 156, 113750.

45.

Wask

W. M.

Faraj

Teigland

(2004). Collective action and knowledge contribution in electronic networks of practice. Journal of the Association for Information Systems, 5, 493–513.

46.

Yang

J. Q.

G. H.

(2020). Research on producer distribution of online Q&A community information: An empirical analysis based on metafilter music. Journal of information science, 39(10), 1060–1068.

47.

Yang

X. Y.

Huang

R. W.

Yuan

Q. J.

(2020). Research on user knowledge contribution decision-making process of professional virtual community based on ethnographic decision tree modeling theory. Information Science, 38(9), 141–148.

48.

Zhang

W. W.

Zhu

Jiang

(2021). Research on the impact of social learning on the knowledge contribution behavior of different types of users in professional virtual communities. Information and Documentation Services, 42(5), 94–103.

49.

Zhang

X. R.

Zhang

J. Y.

(2020). Organizational knowledge, institutional environments, and the syndication of venture capital firms. Journal of Industrial Engineering/Engineering Management, 34(2), 1–10.

50.

Zhang

X. X.

Z. M.

Guo

S. L.

(2018). Information theory & practice. Information science, 41(11), 38–44+50.

51.

Zhang

Luo

Wang

Niu

(2019). Understanding the formation mechanism of high-quality knowledge in social question and answer communities: A knowledge co-creation perspective. International Journal of Information Management, 48, 72–84.

52.

Zhang

Z. F.

Q. D.

(2010). A review of research on community question answering systems. Computer Science, 37(11), 19–23+54.

53.

Zhou

Wang

(2020). A research on knowledge contribution behaviors of open source software community users. Science Research Management, 41(2), 202–209.

54.

Zhou

Wang

Y. Y.

Deng

S. L.

(2020). Research on online health community users’ participation based on social capital theory. Journal of Information Resources Management, 10(2), 59–67+129.

Research on Factors Influencing Knowledge Contribution in Online Q&A Communities: Based on an Empirical Study of QCA on Zhihu

Abstract

Keywords

Introduction

Literature Review

Research on Knowledge Contribution Behavior in Online Q&A Communities

Research Review and Innovation Points

Theoretical Foundation and Research Design

Social Exchange Theory

Knowledge Contribution Behavior

Research Objectives

Research Methods and Steps

Research Methods

Research Steps

Data Collection and Data Calibration

Variable Measurement

Data Calibration

Analysis and Discussion

Necessity Analysis

Conditional Portfolio Analysis

Antecedent Construct Analysis of Knowledge Contribution Behavior

Antecedent Constructs of High Degree Knowledge Contributing Behavior

Antecedent Constructs of Low-Degree Knowledge Contribution Behavior

Conclusion and Insight

Research Findings

Practice Inspiration

Research Limitations and Prospects

Footnotes

Declaration of Conflicting Interests

Funding

Ethical Approval

ORCID iD

Data Availability Statement

References