Exploring the Online Behavior of Users of Online Depression-Focused Communities: Comparing Communities with Different Management Types

Abstract

Introduction

Online depression-focused communities (ODCs) are popular avenues that help people cope with depression. However, to the best of our knowledge, research on online behavior and differences among users from managed and unmanaged ODCs has not been explored.

Methods

We collected data from the most popular managed depression-focused community (MDC) and unmanaged depression-focused community (UDC) in China. Text classifiers were built using deep-learning methods to identify social support (ie, informational and emotional support) and companionship expressed in the posts of these communities. Based on the content of their posts, community users were clustered into supporters and ordinary members. Econometrics was used to analyze the factors that influence supporters’ contributions and ordinary members’ participation in MDCs and UDCs.

Results

Community response has a positive impact on supporters’ social support and time span in the UDC, but this impact is not significant in the MDC. Supporters expressing positive emotions provide more social support, and they are more willing to serve in the MDC. Supporters expressing negative emotions tend to have longer engagement with the UDC. In addition, community response has a positive effect on ordinary members’ participation in both communities, and this effect is greater in the UDC. Ordinary members expressing positive emotions are more active in the MDC, and ordinary members expressing negative emotions are more active in the UDC.

Conclusion

This study improves the understanding of users’ online behaviors in ODCs, provides decision-making support for designers and managers of ODCs, and provides information that can be used to help improve aid for people with depression provided by community and mental health professionals.

Keywords:

• managed depression community
• unmanaged depression community
• supporters’ contribution
• members’ participation
• text-mining
• econometrics

Disclosure

The authors report no conflicts of interest in this work.

Additional information

Funding

This study was funded by the National Natural Science Foundation of China (Grant Nos. 71774041 and 71531013). The authors wish to express their sincere appreciation to the editors and reviewers of this paper.