Return to Answer

##There's no lower threshold for research needed for questions.

And why would there be? Just look around. Many of our best, most viewed and best answered questions come from simple, clear and completely unresearched questions.

I've dallied around this in previous iterations of this answer but it's really that simple. There's probably a positive correlation between question quality and amount of research but I don't see that it's causative. Just as you can have great questions with no research, you can have a thoroughly researched question that's still pretty crap.

Explaining the problem is way more important than demanding some sort of proof of having already searched for an answer.

So forget "research" and focus on the core question quality.

There's no lower threshold for research needed for questions.

And why would there be? Just look around. Many of our best, most viewed and best answered questions come from simple, clear and completely unresearched questions.

I've dallied around this in previous iterations of this answer but it's really that simple. There's probably a positive correlation between question quality and amount of research but I don't see that it's causative. Just as you can have great questions with no research, you can have a thoroughly researched question that's still pretty crap.

Explaining the problem is way more important than demanding some sort of proof of having already searched for an answer.

So forget "research" and focus on the core question quality.

At the lower end, where it matters and there is no overt mention of research, there is no helpful or reproducible way to determine the amount of effort that went into researching a question before it was asked:

• Defining a problem accurately is hard if you don't know the system well. There are a dozen competing names for any given UI metaphor over the various desktops and that can make searching for the correct phrases a pretty tough thing. I still get it wrong.

• Sharing useful information about past research is great (and asked for) but as above, plenty of people search and find nothing relevant. I wouldn't find a long list of failed search terms useful for answering the question.

So ultimately, and in direct answer to the question, there are no mandated expectations for research, only that the questions are on-topic and answerable.

You probably have your own expectations and you're free to vote how you like.

There used to be a research close reason on Stack Overflow. The expectations of reasonable research were so subjective that it caused a flood of close votes against the more simple questions. They had to pull it. Here's Shog9's take from back then:

See, the intent here was to handle the sorts of "here's my spec, please write code for me" questions that were already being closed - not expand closure to damn thousands of existing questions with good, useful answers.

Shog then posted a much wider question late last year about how we treat effort. There is some overlap with research so again, worth mentioning. Effort is great but that's not what we actually want to measure. We care much more about the quality of the question.

So by extension, this "metric" you keep asking for just isn't warranted (even if it were possible). It doesn't help us work out if something is a good, on-topic, answerable question; it's just a distraction.

##There's no lower threshold for research needed for questions.

And why would there be? Just look around. Many of our best, most viewed and best answered questions come from simple, clear and completely unresearched questions.

I've dallied around this in previous iterations of this answer but it's really that simple. There's probably a positive correlation between question quality and amount of research but I don't see that it's causative. Just as you can have great questions with no research, you can have a thoroughly researched question that's still pretty crap.

Explaining the problem is way more important than demanding some sort of proof of having already searched for an answer.

So forget "research" and focus on the core question quality.

Just to bring this in line with your revised question, this answer does talk about close votes because they are an integral part of how we handle questions that haven't met question-quality standards.

At the lower end (where it matters), there is no helpful or reproducible way to determine the amount of effort that went into researching a question before it was asked:

• Defining a problem accurately is hard if you don't know the system well. There are a dozen competing names for any given UI metaphor over the various desktops and that can make searching for the correct phrases a pretty tough thing. I still get it wrong.

• Sharing useful information about past research is great (and asked for) but as above, plenty of people search and find nothing relevant. I wouldn't find a long list of failed search terms useful for answering the question.

So ultimately, there are no mandated expectations on users, only that their questions are on-topic and answerable.

You probably have your own expectations for quality and you're free to vote.

There used to be a research close reason on Stack Overflow. The expectations of reasonable research were so subjective that it caused a flood of close votes against the more simple questions. They had to pull it. Here's Shog9's take from back then:

See, the intent here was to handle the sorts of "here's my spec, please write code for me" questions that were already being closed - not expand closure to damn thousands of existing questions with good, useful answers.

Shog then posted a much wider question late last year about how we treat effort. There is some overlap with research so again, worth mentioning. Ignore arbitrary levels of research and focus on the quality of the actual question.

So what you're asking us to define here ("a metric") is neither possible or desirable. Again, just use the normal criteria you would for judging if it's a good question or not, try not to get sucked into whatever the reasons behind asking it, and move onto the next.

At the lower end, where it matters and there is no overt mention of research, there is no helpful or reproducible way to determine the amount of effort that went into researching a question before it was asked:

• Defining a problem accurately is hard if you don't know the system well. There are a dozen competing names for any given UI metaphor over the various desktops and that can make searching for the correct phrases a pretty tough thing. I still get it wrong.

• Sharing useful information about past research is great (and asked for) but as above, plenty of people search and find nothing relevant. I wouldn't find a long list of failed search terms useful for answering the question.

So ultimately, and in direct answer to the question, there are no mandated expectations for research, only that the questions are on-topic and answerable.

You probably have your own expectations and you're free to vote how you like.

There used to be a research close reason on Stack Overflow. The expectations of reasonable research were so subjective that it caused a flood of close votes against the more simple questions. They had to pull it. Here's Shog9's take from back then:

See, the intent here was to handle the sorts of "here's my spec, please write code for me" questions that were already being closed - not expand closure to damn thousands of existing questions with good, useful answers.

Shog then posted a much wider question late last year about how we treat effort. There is some overlap with research so again, worth mentioning. Effort is great but that's not what we actually want to measure. We care much more about the quality of the question.

So by extension, this "metric" you keep asking for just isn't warranted (even if it were possible). It doesn't help us work out if something is a good, on-topic, answerable question; it's just a distraction.