Let's try to work with users before chasing them off

Question

The Workplace gets a lot of questions whose initial versions don't fit. We correctly close those. Our closure rate is around 50% (I can't find an exact number in the tools), so this is a widespread issue, particularly with new users who aren't as experienced with our expectations.

Fortunately, closure isn't supposed to be the end of the line. The whole point of closing a question is to prevent answers while the question gets fixed so it can be reopened. Many questions don't get fixed and reopened, and instead ultimately get deleted, but we allow time for fixes to be attempted, if people want to try.

All of that is defeated if people rush in to delete questions soon after they're closed. We've had recent deletions that happened in under six hours (one in about 90 minutes!). While that's fine for questions that are abusive or completely unredeemable, I think we're being too trigger-happy with questions that could be fixed or that people are actively trying to fix.

Think back to your first internship or job. You made mistakes because you didn't know the ropes. Some of them were bad, or embarrassing. But unless they were at the "lose a big contract" or "burn down the building" level, you probably didn't get immediately fired for them, right? Instead, your supervisor took corrective action, both to fix whatever you did wrong and to instill in you the importance of not doing that again.

Our new users are like interns or perhaps fresh grads. They have some basic skills and knowledge, but they're going to get it wrong sometimes. Quickly deleting their questions is like firing them. Requesting clarification, closing, editing, and allowing time to fix it is like your supervisor's corrective action.

We're a site about the workplace. Shouldn't we emulate good workplace practices here? Save the insta-deletions for the questions that will burn down the site.

Answer 1

I think its time to turn up the enforcement of the spirit of the be nice policy.

Reduce the threshold for accepting rude/abusive flags especially in comments. If it could be considered rude then accept the flag and let the users that post snarky, and unhelpful or whiny comments suffer the penalties that were intended.

Encourage people to flag comments as rude or abusive more. Especially when they are.

The carrot is the game that these people get to play where they score points for telling other people how to live their lives on the interwebs. Time to bring in the stick for when they stop making an honest effort to be nice.

If you would not say a comment is nice or constructive, then it should be considered rude of abusive. With the penalties that start being enforced you will quickly see a change in the atmosphere here overall. Worst case if it fails then you roll back the penalties and say oops my bad.

The reality of the system is that comments can be used to trigger some people with OCD. Most people around here know who those people are that get triggered. Repeatedly intentionally triggering those people is abusive. Yes I am one and I have never made that a secret.

Answer 2

We're a site about the workplace. Shouldn't we emulate good workplace practices here? Save the insta-deletions for the questions that will burn down the site.

Yes.

Answer 3

In principle, I agree that we should try to help as many people as possible, which translates to saving as many questions as we can. However, there are some XY problems we need to consider.

---

We've had recent deletions that happened in under six hours (one in about 90 minutes!)

The option to delete a question doesn't appear automatically. For simplicity, let us ignore the deletion privilege at 10k reputation, and focus on the trusted user privilege. Trusted users can only vote to delete a post that scores -3 or lower. This leads to two questions:

1. Possible loophole in SE policy: To delete a post, 3 delete votes are required, and coincidentally, a minimum of 3 downvotes are required for the delete option to appear! Theoretically, this allows 3 trusted users to "collude" to delete posts.

   SE could discourage this by not allowing the same trusted user to cast both downvote and delete vote, or by at least imposing a time delay between the two, but given that they are cool with allowing reopen vote along with delete vote, that won't happen any time soon. Besides, that also goes against the "earn your privilege" model of SE, especially since the privilege is named trusted user.

2. Not enough upvotes on "salvagable" questions: The upvote privilege (at 15 reputation) is much easier to earn than the downvote privilege (125 reputation). For every upvote, an additional downvote is required to delete the post, or in other words, every upvote makes it harder to delete the post. So, why don't people upvote "salvageable" questions more often?

Now, considering all that, let's turn the question around and ask, "if the question wasn't that bad, why did it earn so many downvotes within 6 hours?"

---

You made mistakes because you didn't know the ropes ... you probably didn't get immediately fired for them, right? ... Quickly deleting their questions is like firing them. Requesting clarification, editing, and allowing time to fix it is like your supervisor's corrective action.

That is an interesting analogy, but it is also flawed. We don't ban or suspend people for asking one bad question (or even truckloads of bad questions, for that matter), which would be the equivalent of "firing" in the workplace.

A slightly more accurate analogy could be the following: If the intern's (or fresh grad's) work is not too bad, the supervisor will certainly work with them to fix it. However, if the intern's work is horribly bad, the supervisor will probably need to scrap their work entirely.

Whether a question is "not too bad" or "horribly bad" is subjective, but as long as we work within the SE framework, the post score serves as an objective measure for us to work with.

---

Let us not forget that the primary responsibility for asking "acceptable" questions lies with the asker, and not with the community. We certainly want to help as much as we can, otherwise we wouldn't be spending our free time here for internet points and badges. However, we should also not give an impression that people can post a few lines of text with minimal effort, and we will figure out how to make it acceptable.

In terms of the workplace analogy, we don't want the interns to think that they can just submit some half-completed work without much effort, and then expect that the supervisor will "work with them" to complete it.

Answer 4

Experiment with whatever threshold there is to place questions from new users into the "First post" review queue and see whether people use it to improve people's questions.

Users of various communities are used to the idea of new users being moderated, so this won't come as a huge shock to them.

If re-jigging the threshold ends up causing more problems that it solves, it can easily be adjusted.

Answer 5

Over at x.stackexchange.com * we let questions with insuficiencies brew while we await the OP to come back with an edit, to let them reflect upon what they wrote before jumping on them, or to further refine an already great post; unless they are 'new rep' (1 or <110), then a helpful comment will probably be forthcoming.

After a few hours (some are either not so patient or more inquisitive) a useful observation (helpful criticism) is offered, sometimes a few - it's up to the OP to utilize either checking their mail, revisit their question or use their crystal ball and either attend to the comments, fix the question/answer or let it stand.

It is only after several hours it gets to the point where the question gets a hold. Usually the 'holder' further clarifies their expectations and if the necessary improvements are not made the system automatically deletes the post.

---

* X is a few StackExchange communities, not a single one in particular. Surprisingly you'd think that these places would be more pedantic for style, accuracy and well presented thoughts, but instead they are inclusive, kind and thoughtful (except for the odd outlier).

It's not that we must do that here, especially since there's a few sites where you'd think better practices would be in effect but they are not - but consider how it looks from the other point of view:

• They don't know to study the Help and FAQ.

• There's a warning box on their question or odd followup comments on their answer.

• There's more stuff they gotta do.

The Internet is a big place. They can read and do what is needed here or simply bail and retreat to more familiar territory elsewhere.

Since accounts here are anonymous the person(s) who jump on the newcomer (or non conforming, not so new person) first paint themselves, then the community, and finally the whole SE series of websites - with the odd person vowing never to return becoming increasingly common.

Really the answer is how do you want to be perceived - an obligation to read the Help and FAQ before posting, and a more laidback or polite approach by the rare few, will keep the 'Orphan Rate' lower.

Read twice (or as needed) and post once (if needed).

[Derived from the expression: "Measure twice, cut once".]

---

It's a fine motor skill, developed over time.

Answer 6

I think you could make it better for users (new or otherwise).

Suggestion:

This current "off topic" flag could be tweaked for when a question is salvageable and just needs editing:

Questions require a goal that we can address. Rather than explaining the difficulties of your situation, explain what you want to do to make it better. For more information, see this meta post.

IMO something like:

Questions require a goal that we can address. Keeping your question short and only using relevant facts makes it clear and answerable, and avoids it being closed due to difficulty deciphering the context. eg Is it really relevant to the problem, or just something you are passionate about? For more information, see this meta post.

The bit "or just something you are passionate about" suits this site given the likelihood of in-depth scenarios.

If it's salvageable but a lot of work, you could close with the above proposed close reason and they can choose to edit it or not. At least then you've tried - given them the info they need.

I removed the bit "Rather than explaining the difficulties of your situation, explain what you want to do to make it better" because on this site (specifically) surely the difficulties of the situation is what they want to make better?

Then "Unclear what you're asking" flag could be used when it's literally not understandable at all, but not used when the question is just too long or just needs editing. E.g. a lot of time required to read/edit doesn't necessarily mean "unclear".

 

Is that question that bad, can it be salvaged?

It's entirely your choice as to whether you want to edit or not, but you could leave it for someone else if it's salvageable, or if you want to flag leave a comment explaining to the OP what to do next.

To help reduce the amount of closed questions, you could approach a question like this:

• Evaluate the question
• Is it ok?
  1. Yes -> Optionally answer, vote, comment
  2. No
     • Does it look salvageable?
       1. Yes -> Edit, or move on and let someone else edit, or vote to close with proposed off topic reason
       2. No -> Optionally vote to close

For regular users you know if a question is never going to be on topic or useful no matter how much editing is done. But if it's just a badly worded or structured question, then edit it or leave it for someone else to.

You can even leave a comment to explain why it's a problem. Maybe with more time the asker will improve it if they know what's wrong. Rather than just abruptly closed and you likely never see them again :)

 

---

Discussion on things from the comments and MaskedMand's answer, as I have a different viewpoint which might add some thought to the discussion.

Whether a question is "not too bad" or "horribly bad" is subjective, but as long as we work within the SE framework, the post score serves as an objective measure for us to work with.

If the quality of the question is subjective, as that is what the voting is based on then the voting also has to be subjective. So not really a useful "measure" to work with.

Also, each person should judge a question or answer on their own merits. Using other people's votes and flags as a means to determine your own actions risks invoking the Bandwagon effect.

 

why don't people upvote "salvageable" questions more often?

If it's "salvageable" then it needs work so it's arguably not in a state to warrant an upvote. But as it's "salvageable" it can be worked on, so the question here should be:

"Why don't people edit salvageable questions more often?"

 

if the question wasn't that bad, why did it earn so many downvotes within 6 hours?

Downvotes are people's opinion, so you won't ever accurately know why people voted unless you ask them. Opinions/downvotes can be impulsive, malicious, got out of the wrong side of bed, and can be formed from the Bandwagon effect, biased based on a culture.

So even if downvotes and/or it being closed is generally an implication that a question is bad, it's not conclusive, and nor does it mean the question was not salvageable.

Also, people voting to close/flag will likely also downvote. Such votes may have an accurate bearing on the question quality, but only if the reason they are closing is valid. If there is a true culture here for closing too quickly, then those downvotes might arguably be based on that culture and could instead perhaps have edited and not downvoted?

I'm not saying any of this is definitive, just food for thought.

 

---

IMO Community moderation is about members who frequent a site and know it well helping those who do not. Editing to make improvements, guiding people to use the site correctly, making the site great by working at getting good Q&A pairs.

I know specific points were made, but there are no obstacles to the simplistic idea and attempt of people spending a bit more time identifying if a question can be salvageable and editing it, rather than just closing it down as quickly as possible.

Just IMO :)