Re: st: using Freeman-Tukey arcsine transformation with metan command

["Roger B. Newson" <r.newson@imperial.ac.uk>]

PS if Jessica is meta-analysing proportions, instead of differences 
between proportions, then the obvious alternative might be to use the 
logit transformation, estimating odds instead of the usual odds ratios. 
Jessica does not state the Stata version being used. However, in the 
case of Stata 12, a good source on odds (as distinct from odds ratios) 
is in the current Stata Journal (Buis, 2012). In previous versions of 
Stata, the comments on geometric means given in Newson (2003) also apply 
to odds.

Best wishes

Roger

References

Buis ML. Stata tip 107: The baseline is now reported. The Stata Journal 
2012; 12{1}: 165–166. Purchase from
http://www.stata-journal.com/article.html?article=st0251

Newson R. Stata tip 1: The eform() option of regress. The Stata Journal 
2003; 3(4): 445. Download from
http://www.stata-journal.com/article.html?article=st0054

Roger B Newson BSc MSc DPhil
Lecturer in Medical Statistics
Respiratory Epidemiology and Public Health Group
National Heart and Lung Institute
Imperial College London
Royal Brompton Campus
Room 33, Emmanuel Kaye Building
1B Manresa Road
London SW3 6LR
UNITED KINGDOM
Tel: +44 (0)20 7352 8121 ext 3381
Fax: +44 (0)20 7351 8322
Email: r.newson@imperial.ac.uk
Web page: http://www.imperial.ac.uk/nhli/r.newson/
Departmental Web page:
http://www1.imperial.ac.uk/medicine/about/divisions/nhli/respiration/popgenetics/reph/

Opinions expressed are those of the author, not of the institution.

On 02/04/2012 18:24, Roger B. Newson wrote:
> I don't know whether Jessica is really meta-analyzing proportions, or
> differences between proportions (which would seem a more appropriate
> parameter for which to use the arcsine transformation, at least to me).
> However, the arcsine was proposed tentatively by Daniels and Kendall
> (1947) as a possible Normalizing and variance-stabilizing transformation
> for Kendall's tau. The -somersd- package (Newson, 2006) is downloadable
> from SSC, and estimates Somers' D and Kendall's tau-a, offering the
> arcsine transformation, and the alternative hyperbolic arctangent or
> Fisher's z transformation, for the purposes of defining confidence
> intervals for both of these parameters. And, of course, a difference
> between proportions is a special case of Somers' D. It should be
> entirely possible to use either of these transformations to meta-analyze
> differences between proportions, using the SSC packages -metan- and/or
> -parmhet- and/or -regpar- and or -parmest- (which has a -metaparm-
> module to do meta-analyses).
>
> I hope this helps. Let me know if you have any further queries.
>
> Best wishes
>
> Roger
>
>
> References
>
> Daniels, H. E., and M. G. Kendall. 1947. The significance of rank
> correlation where parental correlation exists. Biometrika 34: 197–208.
>
> Newson R. Confidence intervals for rank statistics: Somers' D and
> extensions. The Stata Journal 2006; 6(3): 309-334. Download from
> http://www.stata-journal.com/article.html?article=snp15_6
>
> Roger B Newson BSc MSc DPhil
> Lecturer in Medical Statistics
> Respiratory Epidemiology and Public Health Group
> National Heart and Lung Institute
> Imperial College London
> Royal Brompton Campus
> Room 33, Emmanuel Kaye Building
> 1B Manresa Road
> London SW3 6LR
> UNITED KINGDOM
> Tel: +44 (0)20 7352 8121 ext 3381
> Fax: +44 (0)20 7351 8322
> Email: r.newson@imperial.ac.uk
> Web page: http://www.imperial.ac.uk/nhli/r.newson/
> Departmental Web page:
> http://www1.imperial.ac.uk/medicine/about/divisions/nhli/respiration/popgenetics/reph/
>
>
> Opinions expressed are those of the author, not of the institution.
>
> On 02/04/2012 17:38, Jessica Keithlin wrote:
> > Hello Stata Community,
> >
> >
> > This is my first post on Statalist so I hope it turns out OK =)
> >
> >
> > I was wondering if anyone has any experience using the Freeman-Tukey
> > arcsine transformation/back-transformation for a meta-analysis? I will
> > be performing a meta-analysis on proportions (as opposed to the
> > typical OR/RR etc) and much of the literature recommends this
> > transformation as a way of dealing with the weighting/variance issue.
> > After the transformation I plan on running a random effects analysis
> > using the metan command (a lot of heterogeneity between my studies).
> > Has anyone used this approach before? Does anyone have recommendations
> > for the coding of the back transformation after the analysis is
> > performed? Any cautions or thoughts on the transformation or
> > alternative approaches I might not have heard of?
> >
> >
> > Any input would be greatly appreciated.
> >
> >
> > Thank you!
> >
> >
> > Jessica Keithlin
> > Centre for Public Health and Zoonoses
> > University of Guelph
> >
> >
> >
> >
> >
> >
> > *
> > * 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/
*
*   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/