Characteristics associated with social support in online smoking cessation communities

Objective. Online social support communities for smoking cessation are popular among users, but their therapeutic benefit has not been widely established and moderating their content can be resource-intensive. In an effort to examine the extent to which social support for behavior change exists in online communities, the author performed an analysis of comments posted in the Reddit smoking cessation sub-community. This analysis also sought to determine if automated coding can identify supportive messages and potentially harmful comments. Method. The content analysis included 772 comments collected from the top 20 posts in the online community in the preceding year. Comments were coded manually using a modified version of the Social Support Behavior Code, and a second coder analyzed a random sub-sample to determine reliability. Comments were also automatically coded using LIWC2007. Results. Analyses demonstrated that social support was present in comments, primarily in the form of informational and esteem support (21.7% and 32.5% of comments, respectively). However, 8.5% of comments contained negative behaviors. The automated analysis also provided evidence that social support categories are associated with specific linguistic features. Specifically, esteem support was associated with a greater percentage of positive affective words, while negative behaviors were associated with a greater percentage of negative affective words and social words. Conclusions. The results provide evidence that beneficial supportive communication may arise in online communities, but also that there are risks associated with participation. This research also demonstrates the feasibility of using new analysis techniques to meet the challenges of monitoring large online communities.