2
$\begingroup$

The question When are OLS linear regression parameters inaccurate? has been rewritten 10 times at least. There should be no reason to have it [on hold]. How do I get it off [on hold]? Mind you, there seem to be 2 close votes and 4 reopen votes.

Do I

(1) Put a bounty on it?

(2) Copy it over to a new question?

(3) Bribe a moderator with flattery until they relent?

(4) Change something in the question yet again?

Ideas, please. I'm stumped. Also, even though it says that I can start a bounty in the help section to get myself out of this situation, and the question was posed 72 hours ago, there is no bounty option. So, unless someone else votes to reopen soon, all I can do is copy the question over as a separate question, wait 48 hours, then assign a bounty.

Seriously https://stats.stackexchange.com/help/no-one-answers says in part "If, despite your best efforts, you feel questions aren’t getting good answers, you can help by offering a bounty on any question more than two days old." but that option did not exist for me when the question was on hold, even when I unaccepted @Tim answer, which is another reason why bounty options can disappear, it's not there if you accept an answer.

Many thanks to the good colleagues who reopened the question. For those who feel that I did not play fair, let me point out that the question was headed for oblivion, and for penance I am now wearing the "I am your father" hat, to show everyone how evil I have become working on the dark side.

| |
$\endgroup$
  • $\begingroup$ This is explicitly addressed in our help pages: please visit stats.stackexchange.com/help/no-one-answers. BTW, "a moderator" could be any of thousands of people with enough reputation to vote to reopen. $\endgroup$ – whuber Dec 19 '16 at 0:37
  • $\begingroup$ @whuber Reopen and close votes requires a 3000 reputation of which there are 187 on this site at the moment. Look, you are a really smart guy (trying #3 above :) what are your objections to a reopen? Can you give a reason at least, please (trying #4 above :)? $\endgroup$ – Carl Dec 19 '16 at 1:01
  • 3
    $\begingroup$ Thank you for the reminder about the reputation thresholds--I should have checked them. I haven't been able to read the entire question through. When I last skimmed over it, it seemed like a polemic loosely disguised as a question. There do appear to be many interesting statistical issues brought up in that post, but I just don't have the time to wade through it and sort out where the questions actually are. That made me feel that I was not in an adequate position to evaluate whether it should be reopened, so I have not voted on it. $\endgroup$ – whuber Dec 19 '16 at 1:37
  • $\begingroup$ @whuber Interesting that, "a polemic loosely disguised as a question." What then of science? Do we reject hypotheses because we do not approve of positing them, or do we posit hypotheses to disprove them? Anyway, put yourself in my position and give some advice, but, have a care that I come by my silly opinions honestly. $\endgroup$ – Carl Dec 19 '16 at 1:55
  • 4
    $\begingroup$ The question has more than quadrupled in size from the original, and has almost completely changed in content, after an answer was provided to the original. I would suggest rolling back to something more like the original (something where Tim's answer still applies), and then post the "OLS accuracy" question separately. Wherever possible, I also recommend to "taboo" any words where there appears to be semantic confusion. $\endgroup$ – GeoMatt22 Dec 19 '16 at 6:00
  • 3
    $\begingroup$ I agree with @GeoMatt22. Honestly, if that was not mine answer I would suggest rollback myself, but since it is my answer I do not want it to look like some kind of edit wars. Changing everything in your question after it already gained some attention and was answered is quite inelegant. $\endgroup$ – Tim Dec 19 '16 at 9:00
  • $\begingroup$ @GeoMatt22 My original intent was to present the inaccuracy of least squares. I was diverted from my goal by redirection to word choice. Bias is defined differently depending on which unphysical theorem is tautologized. It was so strained a discussion that I have to apologize for the hoof in mouth. I have Tim to thank about giving a care to help me sort through the nonsense of my own making. Now, what to do about it? I will do whatever makes sense. $\endgroup$ – Carl Dec 19 '16 at 9:05
  • $\begingroup$ @Carl now the question is not on-hold anymore and doing things like applying more changes, posting it again etc. won't do any better. $\endgroup$ – Tim Dec 19 '16 at 9:30
  • $\begingroup$ @Tim I owe you one. Please read and search OLS bias. Try to answer what I was trying to ask, and I will put a bounty on it of +50 to pay you back. $\endgroup$ – Carl Dec 19 '16 at 9:35
  • $\begingroup$ @whuber It says in the help section that one can put a bounty on to open a question [on hold]. However, the bounty option did not exist when the question was placed [on hold]. Now that the question is off of hold, the bounty option has appeared. Somewhere something is incorrect, probably the help section. $\endgroup$ – Carl Dec 19 '16 at 9:43
  • 3
    $\begingroup$ I'm unable to find where the help states you can put a bounty on a closed question, Carl. The relevant page is stats.stackexchange.com/help/reopen-questions. $\endgroup$ – whuber Dec 19 '16 at 18:50
  • $\begingroup$ @whuber Hummm, you said "This is explicitly addressed in our help pages: please visit stats.stackexchange.com/help/no-one-answers. It is a link that you provided to me above, and mentions bounty reopening. That is where I got the idea, as well. So if it has been commutated, that text should be edited. $\endgroup$ – Carl Dec 21 '16 at 5:01
  • $\begingroup$ You're right, Carl, that help text implicitly assumes "any question" means "any open question." However, changing this portion of the help text is beyond the powers of any of the mods here, because it's a common part of the entire SE system. If you would like to agitate for improvement, consider posting on Meta-StackOverflow. $\endgroup$ – whuber Dec 21 '16 at 14:55
  • $\begingroup$ Reviewing the question it seems to be the longest question I have ever seen. It seems to me that it would be better to start over. Adding on to this would try anyone's patience. $\endgroup$ – Michael R. Chernick Dec 28 '16 at 2:03

You must log in to answer this question.

Browse other questions tagged .