8
$\begingroup$

I didn't get any answers or comments to my question. On reflection, I wondered whether I made a mistake by talking about MCMCglmm, which as far as I can tell is not a well known R package, and probably isn't cucial to my question. Should I perhaps re-write it to simply say:

  1. I have used multiple imputation to obtain a number of completed datasets
  2. I have used Bayesian methods on each of the completed datasets to obtain posterior distributions for a parameter (a random effect).

How can I combine / pool the results for this parameter ?

Or would this be too vague ? Please advise how I can improve my question.

$\endgroup$
8
$\begingroup$

I would perhaps just put your quoted text at the beginning of the question, and leave your other information afterwards. Also if you edit it gets bumped back to the top of the feed, so it may attract more attention.

If your concerned about MCMCglmm scaring people off, I would suggest changing your title, perhaps something along the lines of How can I pool posterior means and credible intervals after multiple imputation?

You shouldn't be too worried about the content of the question being responsible for not getting answers (IMO it is well written and concise). More complicated questions are harder to answer though, and given the complexity of the question you shouldn't hold your breath waiting for an answer. You asked it two days ago, have you found any information to help guide you? You can always answer your own question.

$\endgroup$
  • $\begingroup$ Thank you Andy. I guess I am impatient - but I tend to find my other questions that got good answers attracted attention quite quickly. I have made a little progress - so I added the "Update" yesterday, but that's all. I will make the suggested changes. $\endgroup$ – Joe King Aug 5 '12 at 12:24
  • 4
    $\begingroup$ +1, this is the right advice: also change the title, & keep the original content in the question but below the edit (you could use a horizontal rule to mark it off as well). He's further right that there are fewer CV contributors who can answer more complicated questions & they might not notice before it leaves the main page. If you still don't get enough attention, a bounty could be offered. $\endgroup$ – gung - Reinstate Monica Aug 5 '12 at 15:42
  • $\begingroup$ @gung I have tried to implement your advice, including offering a bounty, but I still haven't got an acceptable answer, although comments from 2 users have been helpful. Is there anything else I can do ? $\endgroup$ – Joe King Aug 10 '12 at 21:04
  • $\begingroup$ I'm afraid I don't know. Your question is beyond my expertise (you don't have to go too far to get past it). You could try updating this question w/ what's happened thus far to bump it up on the meta, & someone else might have more advice. If a good portion of your question is really about R, you could also ask to have it migrated to SO, just to try your luck there. I'm sorry for your frustration. $\endgroup$ – gung - Reinstate Monica Aug 10 '12 at 22:21

You must log in to answer this question.

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