Previous Entry Share Next Entry
The FAQ Button
device
jducoeur wrote in querki_project
[I'm back from my honeymoon -- between that and the wedding, that's why you haven't seen much of me in the past month. I'm gearing back up to full speed now.]

Here's an esoteric idea -- really, it belongs more on The Art of Conversation than here, but this is where I'm talking about what we might do in Querki, and there's an outside chance this will prove worthwhile.

I was just in a private discussion about the unofficial SCA group on Facebook, talking with some folks who were taking the moderators to task for being excessively harsh and ham-handed in their moderation. Basically, anything that breaks the group's rules just gets silently deleted -- no discussion, no appeal, nothing. The straw that broke the camel's back was a couple of weeks back -- posts about the Florilegium (an ancient and carefully-tended anthology of posts about the SCA) got banned because the site is ad-supported. That's gotten a lot of folks hot under the collar, and I can understand why: in a number of cases, the rules are being applied too strictly, in a way that lacks nuance. But I can understand the bind that the moderators are in: moderating a group that is that big, that loud and that diverse is just plain hard under the best of circumstances. And given Facebook's craptastic tools for moderating discussions, there is no other really time-efficient way to handle edge cases.

Which led me to think about how this should ideally work, of course, and that led me to the idea of "the FAQ button". The notion goes like this.

As a community, you define conversational rules. In the case of the SCA group, that includes things like:

-- No local event announcements.
-- No commercial posts, or links to commercial sites.
-- No off-topic posts. (Except for the once-a-month Off Topic Weekend, which is an intentional free-for-all.)

In the model I'm thinking of, you can define these *formally* as rules of the community. They are listed as a FAQ, and have individual representations in the system.

As a moderator, when I see a post that I think violates the rules, I can simply delete it outright, but that isn't the model that is encouraged. Instead, the preferred mechanism is to click something (a drop-down, say) that says, "This post violates rule #2 -- No commercial posts". The post is *screened* (not deleted), in the LJ sense of being hidden from everyone but the moderators, and a notification is sent to the poster. The poster, in turn, gets an "appeal this decision" button -- if they click that, they can fill in a brief description of why they believe that this should be an exception. The appeal is sent to the entire moderation staff, who have a window of opportunity to discuss the appeal, and vote on the matter. After the time window has passed, the message either gets unscreened or deleted.

That's a bit complex, and overkill for small communities, but seems like an approach that's likely to work well for larger ones (say, over a thousand participants). I don't know if Querki is ever going to get many groups of that size (that's not the primary target problem), but it's possible. And some of the ideas here (like the easy mechanism for appeal) are probably worthwhile even for more modest moderated communities.

We'll see -- for now, it's an idle speculation, and an idea for builders of conversation tech to consider. But moderation will eventually be important enough in Querki that I may play with it. Thoughts?

  • 1
It's a bit like some of the techniques used on Stack Overflow.

Many of the questions there aren't suitable for one reason or another, and there are a variety of responses--warnings, someone else editing them, a mod putting them "On Hold" or locking them, or closing them (aka deletion). When a question is questionable, you can flag it for reasons, selected from a list--and which ask for further detail in many cases (e.g. if it's a dup, what is it a dup of?). You can also see a count of the reasons other people have flagged it for as you flag it; useful social reinforcement. And a minyan (well, 5) of flaggers can result in immediate action to close.

What's interesting is that since SO uses an earned-trust-based model (you earn points for beneficial participation), there are lots of folks who have the right to flag a question for moderator intervention. Plus, there are full-time mods who review this stuff; and there is metamoderation, where people can go in and agree or disagree with moderator actions.

There's also a separate place (meta.SO) to complain about moderators' actions, which closes the loop a little bit--or at least lets people complain about being moderated in a place that isn't the actual conversation place. This works pretty well.

Interesting -- I'm only a lightweight SO user so far, so I hadn't come across this aspect of it. Not surprising, really: SO is one of the better community technologies I know. My only real complaint is that it is *so* ornate that it can be confusing as hell until you understand its depths, so it has a surprisingly steep learning curve.

But it's a good reminder that I should dig into SO more deeply, and see what ideas are worth adapting (probably in simplified form, generally) for Querki. My requirements are significantly different -- Querki is going to need moderation in a broader way, will tend to have much smaller and better-defined communities, and requires somewhat better ease of use -- but I'd bet that I could learn a lot there...

  • 1
?

Log in

No account? Create an account