Conversation

Notices

1. Embed this notice
   𝓻𝓻𝓪 (rra@post.lurk.org)'s status on Wednesday, 06-Sep-2023 18:31:29 JST 𝓻𝓻𝓪

   • __nate__

   In June I attended a workshop on #research on and through Mastodon by @ccamara and @__nate__. Part of the discussions during that workshop focused on understanding how computational research could happen on the Fediverse. Prompted by the upcoming mastodon 4.2 release, I wrote down some thoughts on what that can mean for scraping-based academic research on the fediverse.

   My tldr: this type of research is going to happen regardless of whether people want it or not. Calls for changing research methods are insufficient. Instead, we could focus on mechanisms for cooperative actors to better navigate consent computationally. The new 4.2 toot:indexable flag possibly provides a model for this.

   https://test.roelof.info/log/toot-indexable/

   • Embed this notice
     Sexy Moon (moon@shitposter.club)'s status on Wednesday, 06-Sep-2023 18:31:27 JST Sexy Moon

     in reply to
     @rra very interesting, thank you. I believe there was discussion about making the indexable flag more flexible but was rejected but I can't remember the specifics.
     
     I don't mean to sound too negative but I attended a medical data ethics conference once and researchers there overwhelmingly were against privacy and consent lol, but this may be specific to medicine where people believe they are saving lives.
     
     Researchers moving from Twitter to Fediverse is wild, Twitter users are already weird compared to the larger population and Fediverse is super weird. Any research based on looking for your keys under the streetlamp because that's where the light is, is going to be unusably biased.
   • Embed this notice
     𝓻𝓻𝓪 (rra@post.lurk.org)'s status on Thursday, 07-Sep-2023 18:16:15 JST 𝓻𝓻𝓪

     in reply to

     • Sexy Moon

     @Moon Yeah the level of research ethics varies a lot with disciplines and countries, which is also why we shouldn't just rely on that. Making it easier to respect wishes will go a long way though.

     As for research on the fediverse I agree that it will be even more skewed demographically than Twitter if you want to make claims about 'people online' or whatever. However I think there are specific areas of research that the fediverse allows for. On interoperability and governance for example as there is a lot of corporate misinformation on that front. In European commission hearings Facebook for example claimed that e2ee interoperable chat was not technically possible, as a way to soften the interoperability requirements for their chat offerings..

     Sexy Moon likes this.
   • Embed this notice
     Sexy Moon (moon@shitposter.club)'s status on Thursday, 07-Sep-2023 18:57:59 JST Sexy Moon

     in reply to

     • Neko McCatface v2023 :verified::makemeneko:
     @roboneko @/rra@post.lurk.org There are two issues. The first is that there is just no exception to informed consent in medicine so they are asking for something new, and there are historical reasons to not want aggregate access to this kind of data. The second is medical information is some of the most sensitive information about a person. This is a controversial statement, but I think 100% data anonymization is probably impossible. A lot of these researchers are working for or sponsored by corporations so it's not really just in the hands of doctors/scientists.
     
     What I found was many medical researchers want access to all medical records and don't want you to even be able to decline consent.
   • Embed this notice
     Neko McCatface v2023 :verified::makemeneko: (roboneko@bae.st)'s status on Thursday, 07-Sep-2023 18:58:00 JST Neko McCatface v2023 :verified::makemeneko:

     in reply to

     • Sexy Moon
     @Moon @rra
     
     > researchers there overwhelmingly were against privacy and consent
     
     in my experience privacy is viewed as from the public eye, not from being a (hidden) data point in a research project. and consent? to quantify someone and crunch numbers? lmao grudgingly acknowledged at best and only because the IRB requires it. people only care because you need their blessing to get anywhere
     
     tbh I don't see an issue with that way of thinking. someone crunching personal data in private isn't at all the same thing as providing a publicly searchable index