"Boilerplate" mod messages are "impersonal" and leave users "confused"

Question

After receiving his first suspension, ex-moderator Robert Harvey voices this perspective on boilerplate mod messages:

I'd prefer a meta post asking for better transparency in mod messages. I used that boilerplate message all the time when I was a mod, but never understood why people got confused until I was suspended myself. The truth is that boilerplate message is every bit as impersonal as the corporate boilerplate SE has been using on Meta, and it provides no information that would help a user figure out what he's done wrong, other than the admonition "we think you were rude." – Robert Harvey 34 mins ago

Having had the pleasure of receiving a mod message about my behaviour in the past, I've got to say that I share the sentiment.

I suppose there are circumstances where boilerplate will tend to be enough. If a user is wilfully doing things that they know are unambiguously wrong, then there is no need to explain the problem with their actions to them in detail. Or perhaps they're a new user falling into a common failure mode that you've got a great boilerplate explanation of already prepared that you know people generally understand and recognise as relevant to their own behaviour. If that's the case, then, great - roll out the boilerplate!

But when dealing with cases where a broadly well-meaning user is crossing a line without realising it? It seems unlikely, then, that the boilerplate is going to tell them much they don't already know, or give them a clear idea of where the line is that they've stepped over.

When dealing with such murky cases, would it not be better practice in general for mods to err on the side of personalised messages outlining - even if only briefly - the specific behaviours that were objected to and why?

Answer 1

In an ideal scenario, two things would happen:

1. The behavior that causes the suspension would be easily collatable and dropped into the mod message automatically (or as part of a workflow)
2. The user would understand that this isn't really a debate.

The current wording of the mod templates are designed to say "You did something bad" as gently as possible, without giving the user something to "Rules Lawyer" with.

After 7 years of moderation, I can tell you that users will rules lawyer.

Even with the current mod templates, there are often (I don't have percentages), replies, and those replies are often (again, no percentages) "what did I do wrong?" or some variation of questioning the moderator's parentage.

Therefore, if the templates are meant to keep people from asking what they did wrong, they are failing at it. If they aren't doing what we expect them to do, why not change them? Why not tell users what they did wrong? They're going to ask anyway.

Answer 2

I personally (as a mod) usually don't use the templates. I think, in this case, when you have to suspend an active member, you should write a clear and personalized message. If a mod uses a template without any modifications, it shows disrespect.

I can imagine only one situation, when it's (probably) ok to send such messages - if there a lot of flags and moderators do not have time to handle them. Otherwise it just looks like laziness and disrespect.

Answer 3

When dealing with such murky cases, would it not be better practice in general for mods to err on the side of personalised messages outlining - even if only briefly - the specific behaviours that were objected to and why?

Yes, it would be. And I do.

In almost every case that I have sent one of these "please stop being rude" moderator messages, I have gone through and highlighted specific examples of the behavior that I am deeming rude and that inspired me to write the message. (Most of the time, this is a series of comments, which I will quote as a bulleted list.)

This way, the user will know exactly what behavior is problematic and they will know what they need to change if they want to continue participating on the site.

If it is not patently obvious why the content that I cite is rude, I will provide a brief explanation of why and which of our guidelines it violates.

If it's someone that I deem to be an otherwise valuable contributor (i.e., if I'm not just trying to politely show a troll the door), I have even been known to propose alternative wordings or otherwise make specific recommendations on how they can modify their behavior (e.g., skip the comments altogether; just cast downvotes and close votes).

If I'm going to take the time to bring the ban hammer down on someone, I feel like I owe them an explanation. Increased transparency and understanding is also far more likely to lead to a mutually-beneficial resolution. Aside from this, the documentation may be useful in the future for other moderators. And going through all the evidence helps me clarify my position and decide whether I want to actually include a suspension and, if so, for how long. These decisions are all based on the severity and the past history, so I'm going to go through that anyway. Why not share the outcome of my research with the party that it concerns?

So, basically, as others have said, these boilerplate messages should be nothing more than a starting point or template. The problem is, writing good moderator messages is hard and it takes time. We can't force people to take the time to write good messages. So it ends up that the templates get sent out directly, whether out of laziness, apathy, or inability. And some standard "approved" boilerplate is at least marginally better than a confusing, nonsensical message.

For what it's worth, the message composition screen shown to moderators already includes some very apropos advice:

• it's about the behavior not the person

• cite specific examples of the problem behavior with links

• always be civil and courteous

• avoid getting dragged into extended discussions

A possible objection to this practice of giving specifics may be that the user will argue back that what they said isn't actually rude—that they'll "rules-lawyer". Well, I've sent a fair number of moderator messages, and I haven't had a big problem with this in response to my "please stop being rude" messages. I suspect that's because I take the time to cite specific examples and explain why they're rude. I also take a pretty matter-of-fact tone. The messages aren't something you read and think there is any room for discussion. If a user does try to engage in a discussion, I simply won't respond more than once (note the last bullet point), and then only to reiterate what I said in the original message (with perhaps an allowance that different cultures view things differently, but requesting that the site's standards be followed, even if they're different than one's own).

Answer 4

These boiler plate templates are often a starting point for a mod message. We (at least in my sites team) include examples to show the user what content was taken as inappropriate and offer guidance on what to improve.

However, the downside to including specific information is that users can enter into a dialogue with the moderators that can sometimes end up on a fruitlessly cyclic argument that doesn’t end up being productive.

Of course, it’s down to each individual moderator how the use the templates and how much extra information or guidance to put into them.

Answer 5

One of the biggest challenges to successfully implementing a new Code of Conduct is addressing the concern that enforcement won't be predictable or fair. This concern has been a recurring theme in many of the responses to the CoC rollout, and I believe community anxiety about capricious or targeted punishments underlie a fair proportion of the negative reaction. Staff and moderators have tried to reassure community members, but this needs to be supported with action. Any action that could reduce a user's privileges should involve special attention to clarity, predictability, and fairness.

To use recent high profile cases, for both Robert and Monica's case SE representatives stated that the user was warned and didn't correct the behavior. In both cases the users disagreed. The clarity of warnings should be improved.

May I suggest the following:

• Use messages to warn users, rather than comments. Comments are transient. Messages provide an opportunity to reflect.
• Effective warnings are clear. Specifically:
  • Identify the offending behavior. Cite and describe the behavior (this is what you did, this is the aspect of that behavior that is a problem).
  • Identify the consequences for repeating the behavior
  • Be specific about what would constitute repeat behavior
  • Avoid generalizations like "be respectful" or "stop being rude"

Yes, a clear, specific warning may set you up for rules lawyering, but vague warnings are notoriously ineffective. Really this comes down to the goal of the CoC. If the goal is behavior change, make your warnings clear and specific. Vague warnings will only serve you if you believe people can't change their behavior, and you just want to purge the user base of all the bigots. In the end, I don't think that strategy will help the community.

These suggestions take some inspiration from (with all sincerity) an evidence based process for working with misbehaving children.

Answer 6

As I'm no mod or previous mod I'll just give my preference as a user

In the context of a mod message concerning behavior that ultimately lead to punishment, how I see it handled in other communities is that there is a strong preference toward making problematic behavior available and clear, preferably verbatim pasted. I wouldn't understand what a mod message is about if it's subject is being kept away from me. I also need to remember and recollect under what personal or emotional circumstances I posted something wrong.

Now I understand that some violations are a bit hard to demonstrate, such as "We've noticed you have a pattern of engaging in lengthy discussions in comments". If not possible to capture the entire context, I would still think it's appropriate to verbatim paste an example of me derailing in comments, and stating this is an example. It's still much more helpful for associating my behavior with what's bad, than reading it explained to me through rules or other person's examples.