12
$\begingroup$

In Paired t-test or Independent t-test @Damien asks about paired vs. independent t-tests. Two commenters have asked him to say more, but I don't see how he can, given his level of evident confusion. I think it would be OK to answer this with some hints, but I don't want to violate site policy.

Thoughts?

$\endgroup$
4
  • 9
    $\begingroup$ I am one of those two, but have absolutely no objection. That members can disagree in public on where to draw a line in interpreting even site policy is at most very slightly embarrassing; it should never stop a member using their own judgment. $\endgroup$ – Nick Cox Aug 12 '13 at 15:26
  • 3
    $\begingroup$ @NickCox +1 I find public disagreement over such grey areas useful. ... as for the question, I think that after a short delay (I think a few days at least), much more fully answering such questions is not only okay but perhaps to be encouraged. The pedagogical value in them doing some of it for themselves has likely either been achieved or won't be with further delay. $\endgroup$ – Glen_b Aug 14 '13 at 21:42
  • 3
    $\begingroup$ @Glen_b Arguably it's only by seeing views on marginal cases that we learn where to draw the line, or least to get some sense of whether each of us is hawk or dove. $\endgroup$ – Nick Cox Aug 14 '13 at 21:45
  • $\begingroup$ @NickCox, your comments were very helpful (several upvotes). I'd suggest you to post them as one answer seeking to consolidate this thread. $\endgroup$ – Andre Silva Aug 23 '13 at 1:54
15
$\begingroup$

If you look at that thread Paired t-test or Independent t-test you will see that the two people who commented were @Glen_b and myself. I am confident that neither of us saw anything remotely personal or critical in this question, which is just using that example to raise a general issue. As it happened, we were also the only people to reply at the time.

Following encouragement from @Andre Silva, I try here to summarize all the comments, with some extra nuances. I incorporate freely comments made by @Glen_b.

Even if there is an explicit forum policy on some matter (e.g. self-study questions, questions that are really about programming, avoiding bad language) there can still be a judgement call about whether a particular posting has crossed the line into unacceptable.

That members can disagree in public on where to draw a line in interpreting even site policy is at most very slightly embarrassing; it should never stop a member using their own judgment. In the limit the only alternative is that a moderator decides everything and that comments of the kind made by @Glen_b and myself are disallowed. Neither fits the principles or practice of CV.

Indeed, policy is maintained, or possibly even evolves, from seeing little discussions over grey areas in public.

Arguably it's only by seeing views on marginal cases that we learn where to draw the line, or least to get some sense of whether each of us is hawk or dove. "Hawk" and "dove" here means more or less inclined compared with others with sufficient reputation to vote for closure on a particular question. Extreme hawks or doves should try to moderate their behaviour or to keep quiet!

In the particular case of self-study questions, it is arguable that after a short delay (a few days at least), much more fully answering such questions is not only okay but perhaps to be encouraged. The pedagogical value of the OP doing some of it for themselves either has likely been achieved or won't be with further delay.

$\endgroup$
2
  • $\begingroup$ @Glen_b Just to flag that the above uses some of your wording, so feel even freer than usual to edit, comment or disagree. $\endgroup$ – Nick Cox Aug 23 '13 at 9:04
  • 3
    $\begingroup$ I have no problem with your incorporating or summarizing comments in this fashion. Indeed, my understanding is that's how it's supposed to work. $\endgroup$ – Glen_b Aug 23 '13 at 17:37
11
$\begingroup$

I think it is fine to use our own judgment. One of my concerns is that the hints may be too complicated or the questioners too confused, causing them to leave and the question unanswered. It is indeed a shame.

I feel that the learner should decide if he/she needs more explicit hint, there is a nifty markdown called "spoiler" >! you can use to mask the more tempting part of the answer:

Statistics are like a bikini. What they reveal is suggestive, but what they conceal is vital. -- Aaron Levenstein

The user will have to scroll over to see the contents. That way, we made our point as an educator, and we don't have to baby the learner and come back to check on them.

$\LaTeX$ codes work in this masked block as well:

$SE = \frac{SD}{\sqrt{n}}$


Previous version of this response

Is there any function like blocking out a chunk of texts so that the questioner will know that a more explicit answer is contained, but get to decide if he/she can work on the hints (which is not blocked out) first?

Here is an illustration from the forum GameFAQs.com, which is very cautious about players giving out game-related spoilers:

enter image description here

The user would have to drag the cursor across the black box to reveal the more explicit part. Like the white text I revealed in this picture.

I feel that the learner should decide if he/she needs more than a hint. This is a Q&A platform, and I hate to see most of these fundamental questions didn't get wrapped up because the questioner left or the members didn't follow up.

$\endgroup$
6
  • 2
    $\begingroup$ Hi Penguin. There is support for a "spoiler" box using the markdown facilities. I don't recall the exact syntax offhand, but it's given in the FAQ. $\endgroup$ – cardinal Sep 26 '13 at 17:10
  • $\begingroup$ This strikes me as a new question (albeit an important one), rather than an answer to the OP's question. Would you mind turning this into a new meta.CV Q? $\endgroup$ – gung - Reinstate Monica Sep 26 '13 at 17:57
  • $\begingroup$ @cardinal Thanks! $\endgroup$ – Penguin_Knight Sep 26 '13 at 18:31
  • $\begingroup$ @gung, I revised my post to make it more relevant. Thanks. $\endgroup$ – Penguin_Knight Sep 26 '13 at 18:48
  • 4
    $\begingroup$ Nice example - for those interested $\LaTeX$ works in the hidden blocks as well. I'd still probably prefer people did not give answers right away like this for homework questions, but after awhile this is a splendid solution. $\endgroup$ – Andy W Sep 27 '13 at 12:22
  • 1
    $\begingroup$ Thank you @Andy, I added another example to show that. $\endgroup$ – Penguin_Knight Sep 27 '13 at 12:26
2
$\begingroup$

I'm not of the mind that getting someone who is clearly stuck out of the weeds is a bad thing. While I think it's fine, and indeed a good idea to ask them to clarify or see if they can push their question a bit further, I don't think there's any use in requiring some arbitrary and undefined amount of work until the poster hits "enough".

If you the answerer think they're stuck, and further work is only going to end in confused flailing and frustration (or worse, in my mind, failing to get one's question answered, never to return), then I see no reason why you shouldn't at least try a gentle nudge in the right direction.

$\endgroup$

You must log in to answer this question.

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