Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down at the end of May, and its replacement, statalist.org is already up and running.


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

RE: st: RE: Reconstructing the distribution from interval data


From   "Kieran McCaul" <Kieran.McCaul@uwa.edu.au>
To   <statalist@hsphsun2.harvard.edu>
Subject   RE: st: RE: Reconstructing the distribution from interval data
Date   Fri, 10 Sep 2010 04:12:31 +0800

...


My first thought when reading Ronan's original post was "If people want to know how many procedures someone has performed in the last year, why don't they just ask them?  Why do they group the responses in the questionnaire?"

I suppose that by grouping the responses in the questionnaire, the authors are tacitly acknowledging that there will be a degree of error in recall of the exact number.  If they had asked for the exact number, then I suspect the distribution would have displayed some evidence of this; a tendency to report 5, 10, 15, etc.  That would still be better than grouping the responses in the questionnaire though.

With regard to the zeros.  I suppose there are going to be a small proportion of surgeons who have done a procedure in the past, but are not quite sure whether it was in the last year or not.  So there is likely to be some misclassification error, but probably small, probably non-differential.


-----Original Message-----
From: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of Ronan Conroy
Sent: Thursday, 9 September 2010 5:49 PM
To: statalist@hsphsun2.harvard.edu
Subject: Re: st: RE: Reconstructing the distribution from interval data

On 8 MFómh 2010, at 19:34, Austin Nichols wrote:

> The ML approach seems preferable; for income data, see e.g.
> http://fmwww.bc.edu/repec/bocode/g/gbgfit.html
> and if you assume a family of distributions for #procedures
> (Poisson, maybe?), you can estimate those parameters the same way.


Good point! A colleague has also pointed out that one frequency is  
known exactly: the number of physicians who did no procedures. This  
allows the calculation of a Poisson parameter. Calculating the  
expected frequencies within each of the categories allows some measure  
of goodness of fit.


Ronán Conroy
Associate Professor
Division of Population Health Sciences
=================================

rconroy@rcsi.ie
Royal College of Surgeons in Ireland
Epidemiology Department,
Beaux Lane House, Dublin 2, Ireland
+353 (0)1 402 2431
+353 (0)87 799 97 95
+353 (0)1 402 2764 (Fax - remember them?)
http://rcsi.academia.edu/RonanConroy

P    Before printing, think about the environment





*
*   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/

*
*   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 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index