According to the about section of Tumblr.com at the time of study creation, Tumblr users can, “post texts, photos, quotes, links, music, and videos from your browser, phone, desktop, email or wherever you happen to be.” Users also have a “dashboard” where posts from other blogs appear; users can comment, reblog, and add hashtags. Tumblr has an “ask” feature, where users can send messages, both as themselves and anonymously. There is also a direct messaging feature.
Tumblr was chosen because the setup of the site mimics more traditional “blog” sites, while combining the features of newer social media sites (e.g., videos, hashtags), giving users a variety of ways to engage. We also believed this site allowed us to be authentic about our identities, without encouraging users to see us as a recovery resource.
An account was created for our lab after IRB approval. With the exception of the survey post, we did not create any content, only reblogging informational posts or motivational posts already on Tumblr. We did not interact with any users; the messaging feature was disabled. A secure Qualtrics link was shared in a post, using recovery-related hashtags such as #EDrecovery, #eatingdisorderrecovery, and #EDtreatment. International recovery resources were listed at the top of each survey page, as was a reminder that participants could skip questions. A combination of 18 Likert and open response questions were included in the survey.
The initial survey page asked if users were above the age of 13; if participants indicated they were not, they were unable to continue the survey. If they indicated they were 13 or older, they were taken to a consent page, where the potential risks and benefits of the study were explained using language appropriate for youth assent. The IRB approved a waiver for the acquisition of parental consent, as it was not possible to contact user’s parents through Tumblr. No personal or identifying information was collected, and we did not ask questions about participants’ EDs beyond diagnosis.
For this paper, five of the questions were examined using theoretical thematic analysis at a semantic level (Braun & Clarke, 2006). The six phases of thematic analysis as described by Braun and Clarke (2006) were followed. The first author generated the initial codes, searched for and reviewed themes, and defined and named themes. The second author and a research assistant checked codes and reviewed themes. The third author participated in defining and naming themes. All authors participated in producing the report.
The first question asked, “Does the recovery community on Tumblr help you with your eating disorder recovery?” with response options including no, a little, sometimes, often, and most of the time. Two open-ended questions then followed, asking, “What about the recovery community helps your eating disorder recovery?” and “What about the recovery community do you think makes your eating disorder recovery harder?” Finally, the open-ended question, “After viewing eating disorder recovery posts on Tumblr, do you usually feel better or worse about your own recovery?” was asked, with a follow-up open-ended question of “Why or why not?” Participants’ responses have not been edited, to stay true to their voices.