A Linguistic Signaling Model of Social Support Exchange in Online Health Communities

Health care consumers and patients are increasingly using online health communities (OHCs) to exchange social support and enhance their well-being. The success of OHCs in promoting health, however, depends not just on posting activity by participants, but, crucially, on whether or not responses are subsequently received. While previous studies have considered various mechanisms by which the likelihood of social support provisioning can be increased (e.g., the establishment of social capital), the impacts of linguistic signals have yet to be considered. Therefore, we consider whether or not linguistic signals in posts -- including sentiment valence, linguistic style matching, readability, post length, and spelling -- impact the amount of support received. Adopting an overarching theoretical framework of signaling theory, this study proposes a model that explains the signaling roles of linguistic features within OHC posts in promoting social support provision from OHC participants. The research model is empirically tested on a large dataset collected from an OHC platform covering multiple health conditions. Results show that affective linguistic signals, including negative sentiment and linguistic style matching, are effective in invoking both informational and emotional support from the community. We also find that informative linguistic signals including readability, post length, and spelling are positively associated with informational support receipt, while readability and spelling are also positively associated with emotional support receipt. Overall, this research not only enriches our current understandings of the linguistic signaling in OHCs, but also provides practical insights into improving social support exchange in OHCs.