Two grad students and I are doing research on #mastodon and "the politics of federation," and so far for data/materials, we've collected: interviews with admins, 70 Masto threads on the topic, dozens of blog posts, and ~170 Masto CoC statements. It seems like a lot, but... what are we missing? Suggestions welcome!
@robertwgehl Random sample of tweets with matched random sample of toots from several different time periods (e.g. around major news events, etc.), label tweets and toots with metadata like "supports alt-right or Nazi philosophy", "contains racial slur," etc. and compare samples
@robertwgehl Other things you could possible code for:
* Does the post contain actual original content?
* Is the post a link to a news source with commentary?
* Is this a hot take on the topic of discussion du jour?
* Is the user posting this identified as a bot?
* Does the user claim to be an individual or a Brand™?
@robertwgehl Hrmmm. what sort of research? I'd say either interviews or case studies with users (who may or may not be aware of the politics) otherwise it seems a touch self-selecting. Perhaps a wide-canvassing survey asking questions which signal awareness of politics of masto without using the words?
@robertwgehl Without a sense of "here are the percentages of users aware of/impacted by the politics of federation" the persuasive power of the interviews and ... likely quite angry threads.... is not nearly as persuasive as a representation of the community.