Here is a question which is a homework question. This is problem 3.23 on page 97 of Hastie et al., Elements of Statistical Learning, 2nd. ed. (5th printing). There are several problems in my opinion:

  1. The poster does not mention that this is homework.
  2. The poster rephrases the original problem by omitting some key definitions, but the questions he asks are from the original problem.

This might be honest mistake, but I do feel somehow cheated, so I already downvoted the question. What I want to ask is whether I should edit the question to reflect that this a homework and to give the references. Also what is the policy for such type of questions?

| |

I'll answer your particular questions first:

  • Here's the current homework policy. However, this was discussed sometime ago, and we've had many new users since then. Also the policy was discussed before we had experienced "term time".
  • @robin: I don't mind the standard deviation question - although I wouldn't up vote it. To my mind, someone unfamiliar with statistics could ask this question. However, statistics homework questions can be completely contrived - see the question on birth weight.

Personally I think that we may have to start closing homework questions. If you look at the recent questions by queops, they are all blatant homework questions. One of the questions was closed, but only because it was a duplicate. If I didn't have mod powers, I would have probably voted to close quite a few of the recent questions.

If my students asked their homework questions in this site, I would be rather annoyed that they just got the answer with little/no work.

| |
  • 1
    $\begingroup$ Maybe it is worth distinguishing between repetitive attempts with homework questions, like the example you cited, and some of the older ones we got that were positive and constructive (IMO), because of the ensuing comments they generated and/or visible contribution by the OP. I agree that CV should not become a "repository" of homework questions. $\endgroup$ – chl Feb 9 '11 at 20:10
  • 2
    $\begingroup$ I think that if someone is stuck with homework and asks for help indicating what is the obstacle then such question is ok. Questions without indication that poster did some work are not ok. Questions which are homework, but are masqueraded as non-homework should be closed. Such questions bug me the most. On the other hand in this case the answer given is very good and actually helps in understanding of LAR, if proper context is given. $\endgroup$ – mpiktas Feb 10 '11 at 7:42
  • $\begingroup$ Hm, I had a thought that maybe homework questions should automatically be community wiki? Has this been discussed? $\endgroup$ – mpiktas Feb 10 '11 at 7:45

If we accept things like that about standard deviation (upvoted 6 time, I think I have few questions that are so much upvoted :))

I guess we can accept a question already asked in a book , if the book does not give the answer ... The best you can do is editing the question (or ask to) with a reference to the book and an improvement with respect to the key definitions.

| |
  • $\begingroup$ Yes, this is a typical example where the number of votes is high for a question that is rather general; there were worse situations, though. But this was a question that was asked during the beta and it seems this period generated a lot of highly voted Q&A (including mine), even for basic computing stuff or general advices. I agree with your proposition, with the limitations raised by @csgillespie about systematic HW questions from the same user. $\endgroup$ – chl Feb 9 '11 at 20:16

You must log in to answer this question.

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