Statalist


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

Re: st: RE: correlated data


From   Nikolaos Pandis <npandis@yahoo.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: RE: correlated data
Date   Fri, 4 Sep 2009 23:45:07 -0700 (PDT)

Dear David,

Than you for the email. I ordered the book you have suggested, although I am not sure when it will arrive since it was temporarily out of stock.

Also, I have found the bland-altman ados and read on the Bland-Altman analysis/plots for agreement which focused on the agreement between 2 methods.

My question is looking for agreement between the standard and 3 new methods of measuremtns on the same sample.

1. Should a paired t-test be used between the standard and each of the other 3 methods for a total of 3 tests?

2. How about using the following:

.xi:reg vol i.methods,cluster(id)

where 
vol=continuous 
methods has 4 levels (0 for standard method and 1,2,3 for the new methods). 
cluster(id) shows that vol is measured by each one of the methods on the same sample

Is there something wrong with this approach?

Any advice is welcome.

Many thanks,

Nick




--- On Tue, 9/1/09, David Airey <david.airey@Vanderbilt.Edu> wrote:

> From: David Airey <david.airey@Vanderbilt.Edu>
> Subject: st: RE: correlated data
> To: "Statalist" <statalist@hsphsun2.harvard.edu>
> Date: Tuesday, September 1, 2009, 6:46 PM
> .
> 
> I thought the point of methods comparison studies was to
> measure the same units with different methods?
> 
> There is a good book on methods comparisons called
> "Statistical evaluation of measurement errors: Design and
> analysis of reliability studies" by Graham Dunn.
> 
> Nick Cox (and others) also have some methods comparisons
> ados available at SSC that you could try out (findit
> bland-altman), but maybe the 3D aspect is throwing you?
> 
> You might also look into the "geometric morphometric"
> literature, particularly the 3D morphometrics literature.
> 
> -Dave
> 
> > Nikolaos Pandis
> > 
> > We have a set of 3-D images constructed from cat
> scans, and we are
> > measuring volumes defined by certain anatomical points
> on the 3-D
> > images.
> > 
> > The reconstruction/measuring technique is performed
> using 3 new types of
> > software and their results will be compared with the
> results of
> > validated/reference technique.
> > 
> > The same reconstructions/cat scans are used for all
> techniques.
> > 
> > The objective is to see how close (do they differ
> significantly?) the
> > volume values recorded by each technique are to the
> values recorded by
> > the reference technique.
> > 
> > I was thinking along the lines of regression with the
> volume(continuous)
> > variable as the dependent variable and technique as
> the categorical
> > dependent variable with 4 levels. The reference level
> would be the the
> > standard/validated method.
> > 
> > However, how would I account for the fact that the
> data is correlated
> > since all measurements for the 4 methods are taken
> from the same
> > reconstructions/scans?
> *
> *   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   |   What's new   |   Site index