Snapshot Serengeti Talk

Maybe a helpful suggestion?

  • Annabz by Annabz

    I'm a political scientist and really enjoy being a citizen scientist in research areas that are not at all my field of expertise. In my department, we have a large project where researchers are trying to measure democracy, by letting people in several countries classify and observe democratic indicators and then report back to the research team. One clever thing they include in their reports, is how certain they are of their conclusions/observations. Maybe that could be included also here? It would make it possible for the observer of the picture to give an indication of whether they are really sure of their identification or not.

    Posted

  • DZM by DZM admin in response to Annabz's comment.

    Hey @Annabz -- that's an interesting mechanic; thanks for letting us know!

    As I understand it, Snapshot, and many other Zooniverse projects, actually manage to extract very useful data without needing that sort of extra variable... the reason being that "sureness" can be inferred from the degree to which different people's classifications agree with each other!

    There's a detailed breakdown here that might be of interest to you: http://blog.snapshotserengeti.org/2012/12/14/we-need-an-i-dont-know-button/

    However, that's not to say that future projects might not use such a variable, so it's interesting to keep in mind! There's something a little bit like it on Penguin Watch, for instance, where people signify if they've marked every penguin, or just the requested 30. 😃

    Posted

  • Annabz by Annabz

    Hi, it sounds like a great system to consider the degree of agreement, and I suppose that a large variation in identification would indicate that people are a bit uncertain In their classification. But if for example the accuracy of identification is equally good in two pics, but the sureness differs, what would that say? There would also be cases where the accuracy differs but the sureness is the same. Perhaps no useful information in this case, but a bit interesting to consider I think.
    I read about the I don't know button, also I wondered why there wasn't such an alternative but understood when I read the reply and found the discussion interesting, I'm a bit of a nerd when it comes to scientific methodology 😃. I'll try the penguin watch! This is great fun, with all the projects:-).

    Posted