Bookmark and Share

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

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

Re: st: RE: Measures of association for a small sample

From   Ronan Conroy <>
To   "" <>
Subject   Re: st: RE: Measures of association for a small sample
Date   Thu, 12 Jan 2012 16:40:21 +0000

On 2012 Ean 10, at 23:48, Steve Samuels wrote:

> Your 13 regions do constitute a"population".  You can certainly compute measures of association as descriptive statistics. I don't see that statistical tests are justified unless you have a notion of a "super-population" from which the 13 regions were sampled: otherwise, what are the repetitions implied in the probability statements of type 1 error and type 2 error?. But opinions differ on this.

Such patterns as you find among the 13 regions are still worth measuring, which is one function of a measure of association.

Ronán Conroy
Associate Professor
Division of Population Health Sciences
Royal College of Surgeons in Ireland
Beaux Lane House
Dublin 2

*   For searches and help try:

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