Correlation between user name (userNNNNNN) and question quality?

Question

Is it just my imagination or is the ratio of good questions lower for users whos name is "userNNNNNN"? Or is there just a glut of smelly questions lately and there happens to be a large number of userNNNNNN users now?

Is this testable?

Similar to Correlation between reputation and use of actual name on SO, but possibly more testable. The marked answer there uses statoverflow.com which seems to no longer exist.

Updated

This naive query

select count(Score) as [Count], avg(Score) as [Average]
from Posts where OwnerDisplayName like 'user%'
union
select count(Score) as [Count], avg(Score) as [Average]
from Posts where OwnerDisplayName not like 'user%'

Produces this result

Count  Average 
------ ------- 
350    1       
626458 3

I would have thought it to be less than 1, but it looks like avg() rounded or truncated. And 626,458 posts (non 'user%') seems low. What did I do wrong?

Answer 1

Your query's wrong. Here's one that's more reliable:

Unregistered user statistics

You're looking for users where the DisplayName is NULL, not LIKE 'user%'. If you query for just users using your filter you only get something like 200 results - and there are way more unregistered users than that.

According to this query, the average unregistered user has:

• 2.1 questions
• 0.6 votes per question
• 1.1 answers
• 0.3 votes per answer
• 9.4 reputation.

Pretty dismal stats, but hardly surprising considering that these people couldn't even be bothered to register (which takes 10 seconds if you have an account with Google, Yahoo, etc.)

Compare with the same stats for registered users - and this includes the long, long tail of Stack Overflow:

• 3.7 questions
• 0.4 votes per question
• 9.6 answers
• 1.3 votes per answer
• 181.9 reputation

So yeah, unregistered users don't participate much, they just hit and run. Everybody who expected anything else, raise your hand. Thank you.

As for question quality specifically, the numbers seem to be about the same according to vote tallies, anyway. But then, that's probably more indicative of the fact that people hardly ever vote up/down questions on Stack Overflow based on their actual merits; instead we're just seeing all the random/sympathy question voting people do in order to get their Electorate badges.

It's pretty hard to measure question quality, so that's about as much data as you're going to be able to dig up.

Answer 2

Oh, I don't know. I've seen some really crappy questions from registered users.

There probably is a correlation. These are folks who just want an answer to their question and want it now. They don't want to bother with all of that registering, looking around and getting a feel for the "culture" of the site, searching to see if their question was already answered, and all that other touchy-feely stuff. Because they just want their answer. Now, dammit. And if they don't get it they're off to some other place they Googled. Heck, they're probably already gone.

So why bother to write a decent question?

Answer 3

userNNNNNN didn't bother give himself a name because he isn't planning to use the site regularly. He wants one answer to one question. So he posts it here.

"Casual user." Questions may not be that deep because the person doesn't have a lot of questions -- just the one. Probably a younger person.

Not every person on the Internet is a forum troll, you know.