5 ms·
As I'd have expected, it seems providing an explanation does indeed reduce the likelihood the user will break the rules in future, and make them more likely to
by CM30 6y ago
As I'd have expected, it seems providing an explanation does indeed reduce the likelihood the user will break the rules in future, and make them more likely to contribute in future.
Surprised there wasn't much difference between a human provided reason and a bot provided one though (and in fact, the latter performed slightly better). Wonder what the reason behind this could be?
> s.Our results show that controlling for other factors,explanations provided by automated tools or bots are associated with higher odds of moderatedusers posting in the future
Either way, I always make sure to provide an explanation why a piece of content was removed on any site I manage, with the sole exception of obvious bot submissions (since the latter are literally incapable of understanding feedback).
- js8 6y agoI wish HN would require an explanation (or a counter-argument) for downvote. Often I get downvoted and have no idea why. If the content is obviously a spam, it should be flagged and removed, not just downvoted.
- GuB-42 6y agoSlashdot has that, somehow. Up/downvotes has categories. For upvotes there are interesting, insightful, informative and funny. For downvotes there are troll, flamebait, redundant, offtopic. There are also the general categories overrated and underrated. You can also flag posts for things like spam. It is not a full review but it is better than nothing. Also, users can give out weights per category. For example if you don't want to see jokes, you can lower the value of "funny".
- CM30 6y agoAgreed. I think these types of systems work better than the standard upvote/downvote one, and I wish more sites would use them. I know Facebook reactions are a bit like this, as is a popular XenForo plugin with similar features.
- js8 6y agoYes, but I think implementing a requirement to post a reply (or at least upvote some other person's direct reply) before a downvote would be much simpler. The categories do not take away certain subjectivity of the up/down votes. It is easy to understand silent upvote as "I am in agreement and I think more people should read this". With silent downvote, the intent is clearly "less people should read this" but the reason as to why is not given. It also nicely solves the meta-moderation problem. If the reason for downvote is given, it can be validated whether the explanation is meaningful or not.
- matthewheath 6y agoThe emotional rejection experienced by the moderated user might be less when they associate that experience with an automated tool/bot rejecting their posting versus a human. It's possible that a moderated user might feel less dejected when they know it's not a human personally reviewing their content and rejecting it. Certainly I myself would feel less inclined to try posting again if a human moderator rejected my content (because I feel that I have little chance of changing their mind, and I'd feel judged) whereas I'd be inclined to try again if a bot rejected my content because it's impersonal. This is, of course, entirely irrational since automated tools can't really be persuaded in the same way a human can.
- TulliusCicero 6y agoSometimes, you don't want them to contribute anymore. A large number of the people who get banned on the subs I moderate are just belligerent assholes that we're better off without. Someone posting blatantly racist or misogynistic comments isn't someone we want around.
- naasking 6y ago> A large number of the people who get banned on the subs I moderate are just belligerent assholes that we're better off without. Nobody is "just" anything.
- TulliusCicero 6y agoI’m talking about their message board contributions, obviously. That someone might have a normal life outside of screaming at women and minorities online doesn’t have a ton of relevance on the forum where they’re doing the screaming.
- naasking 6y agoMy point is that reducing even someone's comments to a simple label is a dangerous mindset,particularly negative labels. Once you've applied that label, there is no redemption. What would this person have to do to remove a negative label you've associated with them?
- TulliusCicero 6y agoI think it's more dangerous to allow them to stick around because of a misguided sense of charity. > What would this person have to do to remove a negative label you've associated with them? First, let's stop right here. You're using weasely language to try to make it sound like noticing someone has said bigoted things, and kicking them from the community for it, is some kind of mistake, as if the bigotry was the fault of the moderator, rather than the bigot. Place blame where it belongs: if someone says something sexist, someone noticing that they appear to be sexist is fully the responsibility of the person who said the sexist things. Let's reword what you said in a way that makes sense: > What would the bigoted poster have to do to convince you they're no longer a bigot? (at least as far as internet posting goes) There, much better. Depends on what they said, of course. If they said something racist, I would expect an earnest acknowledgement of that, that racism is bad, etc. As well as evidence from posting elsewhere on Reddit that they've reformed.