The aim of this research was to document the content on pro-ana blogs and to uncover how the unique socio-technical features of blogs (interactivity, self-disclosure, masspersonal communication) facilitate social support among members. A final sample of 48 pro-ana blogs provided 624 individual units for coding. Results indicate that prevalent forms of social support were emotional support, esteem support, and informational support. A new category, reciprocal self-disclosure, was also revealed to be quite frequent. Blogs are spaces where social support is sought and communicated among members of the pro-ana network. Interpretation of blog communication and implications for treatment and research are discussed.

No abstract

To explore the feasibility of social media for message testing, this study connects favorable viewer responses to antismoking videos on YouTube with the videos' message characteristics (message sensation value [MSV] and appeals), producer types, and viewer influences (viewer rating and number of viewers). Through multilevel modeling, a content analysis of 7,561 viewer comments on antismoking videos is linked with a content analysis of 87 antismoking videos. Based on a cognitive response approach, viewer comments are classified and coded as message-oriented thought, video feature-relevant thought, and audience-generated thought. The three mixed logit models indicate that videos with a greater number of viewers consistently increased the odds of favorable viewer responses, while those presenting humor appeals decreased the odds of favorable message-oriented and audience-generated thoughts. Some significant interaction effects show that videos produced by laypeople may hinder favorable viewer responses, while a greater number of viewer comments can work jointly with videos presenting threat appeals to predict favorable viewer responses. Also, for a more accurate understanding of audience responses to the messages, nuance cues should be considered together with message features and viewer influences.

No abstract

No abstract