Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: very small sample comarsion


From   Ronan Conroy <rconroy@rcsi.ie>
To   "statalist@hsphsun2.harvard.edu" <statalist@hsphsun2.harvard.edu>
Subject   Re: st: very small sample comarsion
Date   Sat, 26 Feb 2011 11:55:17 +0000

On 26 Feb 2011, at 09:33, ajjee wrote:

> Dear All
> 
> I have many samples of very small samplesize (n=5 and some times n=3) and I
> want to compare the the sample means of two samples. But I don't think I'll
> be able to apply t-test for these sample of very small size. In this
> situation, which methods I should use?. Thanks in advance.


The problem here is that your statistical power is almost nil, and only massive difference between the samples will be significant – and if this is the case, commonsense would have told you that they were different without the need to collect data.

Small samples generally contain an absence of evidence. The trouble is that many people cannot tell the difference between an absence of evidence and evidence of absence. 

One note of optimism: if you have many small samples, can you tell us if these samples are completely independent, or whether they have some relationship. It may be possible to gain power by combining. 




Ronán Conroy
rconroy@rcsi.ie
Associate Professor
Division of Population Health Sciences
Royal College of Surgeons in Ireland
Beaux Lane House
Dublin 2


*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/


© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index