Is it possible to quantify the quality of questions over time?

Question

Without getting in to a rant, it appears to me (and others) that the quality of questions being asked on Stack Overflow has greatly diminished. I'm curious if there has been any attempt to take a critical look at this by quantifying the questions asked on Stack Overflow over time?

No attempt to do this would be fool proof or completely accurate, but some information -- viewed with critical eyes -- can help to evaluate not only the "if" of this question, but also the "why."

Perhaps one simple attempt to do this would be to simply construct a graph where the Y-axis was total question votes or total number of questions and the X-axis were the date on which the questions were asked?

Answer 1

I feel that determining this would be very difficult just based on statistics alone. As Ivo Flipse once brought up, sometimes low quality answers and questions get upvoted a lot. We can see some things with the review tab. But with millions of questions now, getting something like this just on statistics is near impossible. The way to see this though, is through the users. Older members will be able to tell you the change, as will regulars to the site, and editors.

We know low quality posts happen - and then you bring in a key point - why? I think Jeff and Joel recently discussed this in a podcast, and came to some sort of consensus on it. People seem to actually find it easier to ask simple and low quality questions on SO, than actually just going to Google.

How can we stop it? - We really can't. This is the internet of 2011. But thanks to all the great editing tools at our fingertips, and the ability to use chat and comments, we can easily turn simple questions into something better, that can help more people.

Answer 2

I don't think votes alone would do it, but perhaps you could come up with a decent heuristic by also factoring in the average number of edits, the average number of close votes (not just closed questions), and a few other stats.