2
$\begingroup$

This very interesting question has an answer that relies on the dubious assumption that the frequency response of time-varying systems exists in continuous time.

I believe the answer is blatantly wrong and should not appear on the site and should be deleted. However, since I've been involved in the discussion in the comments there, I don't think it's up to me to do that.

I've given a specific example, but my general question is:

  • Should plainly incorrect answers be deleted?

and as a follow up:

  • What happens if they are the accepted answer?
$\endgroup$
5
  • 2
    $\begingroup$ Just for the sake of argument: when should an answer be deleted vs downvoted? An answer with many downvotes could be a kind of warning to readers that, even if the answer may seem sensible at first sight, the community believes it is actually wrong. $\endgroup$ – MBaz Oct 11 '16 at 0:34
  • 1
    $\begingroup$ @MBaz I've deleted answers in the past when they are spam, self-promotion, or duplicate answers to multiple questions. You make a good point, though: we should let the voting process run it's course... I'm just not sure there are enough voters on his question to get a good read. It's a little theoretical. $\endgroup$ – Peter K.♦ Oct 11 '16 at 0:45
  • $\begingroup$ Yeah, I agree that some answers need to be deleted (I've flagged a few myself). I was thinking about answers such as the one you linked. Personally, I've never downvoted a question because of the reputation hit (I know that's silly). Maybe I should start with this one.... $\endgroup$ – MBaz Oct 11 '16 at 1:24
  • $\begingroup$ From the comments of that answer, it was clear that it was disputed, which is a warning sign to the reader. $\endgroup$ – Olli Niemitalo Oct 11 '16 at 11:53
  • $\begingroup$ @OlliNiemitalo “The four stages of acceptance: 1. This is worthless nonsense. 2. This is an interesting, but perverse, point of view. 3. This is true, but quite unimportant. 4. I always said so." $\endgroup$ – msm Oct 12 '16 at 15:18
-1
$\begingroup$

For the interested readers: this is the "blatantly wrong" reference that is cited and defines the frequency response of the time varying channel (Eq 2.20). Author's page here.

Furthermore, please look at this answer and the comments therein, then you will find out Wikipedia is also wrong. Note that it is admitted by the author in a comment that his answer is wrong.

$\endgroup$
5
  • $\begingroup$ -1 Nowhere in the comments do I say my answer is wrong. It is correct. I'll update my answer to show where the term cross-correlation comes from and why the sample cross-correlation, which Wikipedia uses, is only an estimate of the "true" cross-correlation. $\endgroup$ – Peter K.♦ Oct 13 '16 at 12:28
  • $\begingroup$ "the =0 part isn't" [correct]. My comment refers to the =0 part only. I have no confusion with cross and autocorrelation. Also you need to locate the "wrong" in wikipedia. $\endgroup$ – msm Oct 13 '16 at 14:53
  • $\begingroup$ I believe there is a wide-spread misunderstanding of where cross-correlation came from, any why the two definitions differ. Please see the update to my answer on that question and see if that helps. $\endgroup$ – Peter K.♦ Oct 13 '16 at 15:57
  • $\begingroup$ Again: My problem is only with $=0$ part of your "definition" in (1), (3), and (4). Why is changing of subject? Is cross-correlation always zero by definition? $\endgroup$ – msm Oct 13 '16 at 23:00
  • $\begingroup$ perhaps because you didn't make yourself very clear. Thank-you for the clarification. $\endgroup$ – Peter K.♦ Oct 14 '16 at 2:17

You must log in to answer this question.

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