4
$\begingroup$

Please correct me if I am wrong.

When I review the past questions, I have feeling that there seem a period of time where there were obviously less registered users but more participation? The most noticeable quantity difference is the number of vote. It seems hard now for any questions and answers to get more than 10 votes? but I didn't see any obvious quality drop in new questions and answers? Is this my illusion?

If not, then why? What's the difference between now and the past? Sorry, as a quant, I didn't support my argument with any numbers though I don't think I have access to relevant numbers.

$\endgroup$
  • 1
    $\begingroup$ Do you think it is just the passage of time? Maybe upvotes are a Poisson process? $\endgroup$ – Richard Herron Oct 6 '11 at 2:05
3
$\begingroup$

The gist of richardh's comment is spot on. I continue to receive up-votes for things I posted even months after I originally posted them. If anything, I think this is a good sign. It means there are many users who find great value in some of the older questions and answers, and just because they are no longer on the front page doesn't mean they aren't being looked at.

If you want to get a sense for this process (without actually downloading the full data dump, as Jeff suggested), look at the user pages for some of the high-rep users and check out the "reputation" tab. On my page, for example, you'll notice I got an up-vote yesterday for an answer I posted August 8. In this case, it is because Brian B posted a new answer yesterday, and apparently some folks decided to up-vote both his new answer and my original answer, perhaps because they liked both and felt they didn't want to disturb the existing ordering (I often do this myself). Shortly thereafter, a new user posted yet another answer. Now it looks like a very active question to someone new visiting the site, whereas for the first few weeks it looked like only Lliane and I were interested in it.

Another question here, which was recently revived by a post to hacker news, has a similar story. Note that as of this moment this question has over 7K views, whereas 22h ago it had under 1K views. This happened a few weeks ago with another programming-related question. Our community tends to handle these spikes relatively well, but in that case the question eventually had to be protected.

PS: Participation really is up quite a bit. To see that, check out quantcast. Even ignoring the hacker news spikes, traffic is up quite a bit from the early days. The page-views graph is less spiky (since those who just visit the one advertised question and never come again receive less weight) and shows that after an initial flurry of activity in February, the site started to lose traffic from March to May, but then picked up again in August and will hopefully continue growing strongly. Also, much of the initial activity was due to a limited group of people posting many questions and answers, whereas now the site has a much broader base of users.

$\endgroup$
  • 1
    $\begingroup$ I was surprised myself by the recent Hacker News influx. Half a year later I feel mostly comfortable with R, so maybe I will jot down my transition impressions when I get some spare time. Pertaining to the question - I think we may have a less enthusiastic attitude now (compared to beta), but participation is more balanced and increasing, which is a good sign. May/June was the hardest time, let's hope it's over. $\endgroup$ – Karol J. Piczak Oct 6 '11 at 19:32
2
$\begingroup$

Well, if you'd like to investigate, the data is creative commons and you can download it from the Creative Commons Data Dump -- you can also download just the Quant files using a good Torrent client, no need to pull down everything.

It is true that we do see certain sites (Android, Webapps) that tend to get a lot of "drive-by" users that dwarf the number of active, avid users even 1+ year in. I would be surprised, however, if this was the case for quant as it's a highly technical subfield, whereas the scope of "all Android devices" and "all web applications" is far, far larger.

$\endgroup$

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .