Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

RE: st: An allowance for clustering should only increase SE's?

From   "CDSC - Nichols, Tom" <>
To   "''" <>
Subject   RE: st: An allowance for clustering should only increase SE's?
Date   Mon, 18 Apr 2005 13:56:26 +0100


Thank-you very much.  
I will take your advice and only allow for clustering if this increases the

I can see that it is possible for units within clusters to be more different
than units sampled randomly from the population as a whole.  But I think the
opposite is more likely to be true (at least for the clusters I have).

I found references to the intra-cluster correlation (or intra-class)
correlation saying that estimates can be negative.
But rather confusingly, they usually also say that the total variance can be
divided into a between cluster component and a within cluster component, and
that the intra-cluster correlation is   var(b)/[var(b)+var(w))]
Some references even give this as the definition of the intra-cluster
Accordingly most calculations are based on an ANOVA table and only lead to
positive values.


Joanne Garrett wrote:

> Negative values
> imply that people within a cluster (or PSU) are more different from each
> other than they would have been had you sampled randomly (i.e., not by
> cluster). Intuitively this makes no sense.
As the great diagnostician Charcot was fond of saying, 'ça ne l'empèche 
pas d'exister'. There is one occasion on which this may indeed be true: 
where the PSU is a group of people who form a team or social unit 
characerised by a division of labour. A family, for example, or a 
community mental health team, will combine people with different 
specialisations and points of view. Two people randomly sampled from 
within such a PSU will be more sharply differentiated than two people 
sampled from the population.

Certainly, I recally that my mother and I were capable of producing a 
kappa of close to minus one in matters of musical taste.


Ronan M Conroy ( 
Senior Lecturer in Biostatistics 
Royal College of Surgeons 
Dublin 2, Ireland 
+353 1 402 2431 (fax 2764) 
Just say no to drug reps

This email and any files transmitted with it are confidential and
intended solely for the use of the individual or entity to whom
they are addressed.
If you have received this email in error please notify the
originator of the message. This footer also confirms that this
email message has been scanned for the presence of computer viruses.

Any views expressed in this message are those of the individual
sender, except where the sender specifies and with authority,
states them to be the views of The Royal College Of Surgeons in Ireland.


*   For searches and help try:

The information contained in the EMail and any attachments is confidential
and intended solely and for the attention and use of the named
addressee(s). It may not be disclosed to any other person without the
express authority of the HPA, or the intended recipient, or both. If you
are not the intended recipient, you must not disclose, copy, distribute or
retain this message or any part of it. This footnote also confirms that
this EMail has been swept for computer viruses, but please re-sweep any
attachments before opening or saving. HTTP://

*   For searches and help try:

© Copyright 1996–2017 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index