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 on April 23, 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: metan with repeated-measures designs?


From   "Ploutz-Snyder, Robert (JSC-SK)[USRA]" <robert.ploutz-snyder-1@nasa.gov>
To   "statalist@hsphsun2.harvard.edu" <statalist@hsphsun2.harvard.edu>
Subject   Re: st: RE: metan with repeated-measures designs?
Date   Sat, 24 Jul 2010 13:04:10 -0500

You are correct in that all I have is means from manuscripts, so I  
can't work w  raw data.

Interestind idea to estimate from a single study.  Hmmm



On Jul 23, 2010, at 2:43 PM, "Tiago V. Pereira" <tiago.pereira@mbe.bio.br 
 > wrote:

> what if you create a single effect size (or something) like :
>
> Delta_i = (pos[right leg]-pre[right leg])-(pos[left leg]-pre[left  
> leg])
>
> for the ith time (i = 1,2,3...n).
>
> The variance for Delta_i is very straightforward for normal variables.
>
> Then, use a multivariate meta-analysis approach.
>
> Ben is right,  -mvmeta1- may be what you are looking for. However,  
> as you
> are working with summary data (probably from reported data, am I  
> right?),
> you don't have an estimate for the covariances for the effect sizes.
>
> A strategy would be to search in the literature for these estimates  
> (or
> calculate them using a single study with raw data) - then try several
> other values in sensitivity analyses.
>
> I believe  there is no Stata code to perform meta-analysis of
> repeated-measures exactly as you are looking for.
>
> Tiago
>
>
>
>
>
>
>
>
>
> -----------------------------
> Statalisters,
> Please forgive my re-post, but I'm hoping there's a gem of info out  
> there??
>
> I'm looking for help using -matan- (or other) command for meta  
> analysis of
> studies that all used repeated measures designs.  In the most simple  
> form,
> I'd like to be able to estimate an across-studies effect for Pre/Post
> differences when all I can extract from the literature are means and  
> SD
> pre and post for several studies (i.e. no mean difference score with
> SD-diff were reported in any studies).
>
> -metan- is great, but it's written for independent groups  
> designs...  is
> there a way to accommodate repeated-measures designs with -metan- or  
> other
> user-written commands?
>
> Rob
>
> *
> *   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