1

I have been assigned a topic on "A simulation study of sample size for multilevel logistic regression."

I have searched the topic but found little reference on it.

Could you please offer some inputs, ideas or insight on the subject ?

ABC
  • 1,367
  • 3
  • 13
  • 31
  • Whether something is interesting is subjective and not a good fit for this site. – Juho Kokkala Feb 16 '15 at 13:02
  • At least please tell how do you find this topic intersting. – ABC Feb 16 '15 at 13:10
  • 2
    You have good key words for a literature search there. Beyond that, the scope for addressing directly what you find boring, tedious or irrelevant is minimal and explaining why you are wrong would be difficult; where are we supposed to start? I don't think this is a useful or answerable question. And (I'm serious here) most people on this forum wouldn't find this interesting either unless and until they were bitten by some real problem in which sample size was an issue. So (to make just one positive suggestion) finding a real case study in which the sample was small might be a start for you. – Nick Cox Feb 16 '15 at 13:30
  • 1
    @Juho Kokkala I think we should refrain from taking the question literally, as it's not about how interesting the topic is, but about what inputs, ideas or insight the community can offer on the subject. – mugen Feb 16 '15 at 13:32
  • 1
    Check this topic for an interesting subject for simulation study: http://stats.stackexchange.com/questions/126590/how-many-levels-in-multilevel-modeling-is-too-many – Tim Feb 16 '15 at 13:42
  • @mugen Thank you. I've edited my question following your comment. – ABC Feb 21 '15 at 05:15

1 Answers1

2

Here is an example: https://www.iser.essex.ac.uk/research/publications/working-papers/iser/2013-14 . The point in that article is that the sample size of the higher level units can very limited when you use it for country comparisons. This is a large area in which multi-level models are used, so if that is problematic then that is a big deal for many users of this method.

Maarten Buis
  • 19,189
  • 29
  • 59