12
$\begingroup$

This issue arose with a question about gathering web page viewing statistics which was closed by three high-rep users and a moderator. There were legitimate reasons for voting to close (it was not well formulated), but it makes me wonder:

  • Does this community consider questions about collecting data to be germane?

  • If such questions aren't welcome here, then where should they appear?

My own opinion is that designing data-collection activities as well as the practical and logistical details of collecting data are a traditional (and important) part of the statistical profession, so I am inclined to use a very light hand in moderating such questions, but I would like to know the sense of the community.

$\endgroup$
  • $\begingroup$ I'll just say that I agree with Rob. Data collecting questions should be allowed, but this particular question was closed for other reasons. Maybe it should have been moved to the other site. $\endgroup$ – mpiktas Mar 3 '11 at 15:31
  • $\begingroup$ @mpiktas Thanks, but please bear with me while I repeat myself: this inquiry is not about that specific question. It is about the community's interest in data collection methods. $\endgroup$ – whuber Mar 3 '11 at 15:43
  • $\begingroup$ @whuber, sorry for misunderstanding. I am all for the data collection methods. It is integral, although less glorious part of statistics. If we look at the questions and answers in this site, I think we can safely say that the site is geared more towards practical real life solutions, hence not allowing questions about data collection methods would be strange. $\endgroup$ – mpiktas Mar 3 '11 at 16:43
  • $\begingroup$ @mpiktas Thank you. I agree, but because I don't recollect seeing many such questions (indeed, none at all), I was wondering. I had to create the data-collection tag for this question! $\endgroup$ – whuber Mar 3 '11 at 16:46
  • 1
    $\begingroup$ @whuber I agree with your last comment. It seems we have seen several questions dealing with data management (with R or Excel, for example), but not data collection. It's an important aspect of statistical work, though. $\endgroup$ – chl Mar 4 '11 at 9:49
  • 2
    $\begingroup$ @chl The more I think about this, the more important this question seems to be. The universal complaint among professional statisticians is "the client should have consulted with us before doing the experiment." In other words, we need to encourage people to seek out this site before they have a problem in statistical analysis: we want them to find their way here with questions on experimental and observational design and questions about how to collect data. Maybe we should consider updating the FAQ and perhaps seed the site with relevant questions? $\endgroup$ – whuber Mar 4 '11 at 16:01
  • 2
    $\begingroup$ @whuber Yes, this reminds me of Fisher's famous quote, "To call in the statistician after the experiment is done may be no more than asking him to perform a post-mortem examination: he may be able to say what the experiment died of". I think updating the FAQ would be a good idea, indeed. We already have a lot of people involved in experimental science (and also students with varying degrees), so that I'll expect more questions to come in the near future if we promote them. $\endgroup$ – chl Mar 4 '11 at 16:15
  • $\begingroup$ @chl The FAQ is now updated. $\endgroup$ – whuber Mar 4 '11 at 17:01
12
$\begingroup$

I agree that data-collection is a core part of statistics. But the specific question seemed more of a technical question about how pageviews on websites are registered which would probably be more appropriate on webmasters.stackexchange.com.

$\endgroup$
  • $\begingroup$ Uhh Super User has nothing to do with pageviews or websites whatsoever ;-) $\endgroup$ – Ivo Flipse Mar 19 '11 at 10:59
  • $\begingroup$ @Ivo. Fair enough. webmasters.stackexchange.com is the place to go. Answer now corrected. $\endgroup$ – Rob Hyndman Mar 20 '11 at 5:54

You must log in to answer this question.

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