Research talk:Understanding Content Review Practices of Wikipedia Editors

Spelling

The main thing I do on wikipedia nowadays is fix typos, but that does often lead me to detect vandalism, sometimes vandalism that has stuck in wikipedia for many months. Hopefully we are correct in assuming that vandals can't spell. If not we are in trouble. WereSpielChequers (talk) 14:56, 28 October 2018 (UTC)

ha - this kind of thing is useful insight for our design team, thanks. --Andicat (talk) 16:22, 29 October 2018 (UTC)

representative sample?

Thanks for doing this research, looking forward to outcomes.

I'm curious though whether you considered whether your recruitment method will result in a representative sample. What is the population you're trying to describe? For example, I can imagine that editors from the student population behave differently than editors from the general population would (e.g. use different tools, workflows). Also, if you expect these results to be generalizable for the English Wikipedia, you may want to consider recruiting editors from different countries, with varying degrees of comprehension of English (e.g. native speakers, near-native speakers, en-3, en-2). To use the example of WereSpielChequers, someone who is not a native speaker and doesn't feel 100% comfortable with the language, may be less likely to do spell checking as a chore. I can also well imagine that American editors consider different content problematic than, say, Indian editors - especially when topics have a geographic significance. This may impact how they approach the decision process. Effeietsanders (talk) 06:23, 29 October 2018 (UTC)

Thanks! Good points. We are not aiming to collect a representative sample or to generalize for this initial observation. Instead, we aim to observe some set of real-world practices as a sanity check before we build tools to do larger scale studies. --Andicat (talk) 16:21, 29 October 2018 (UTC)

Same IP

One of the biggest predictors of an edit being vandalism is that other edits by the same person have been identified as vandalism. So some tools such as Rollback will show you other edits by the same account or IP once you've reverted an edit for vandalism. But this is only useful where those other edits have not already been reverted - rollback could really do with an upgrade here. I'd be very interested if Rollback would prompt me to look at live edits that may be by the same vandal. There is also an opportunity if we have dev time to show other edits in the same IP range as the vandalism you've just reverted. Currently we can IP range block, but spotting vandals who are resetting their IP address is difficult if they change the page they edit. WereSpielChequers (talk) 11:28, 30 October 2018 (UTC)