Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

[no subject]


From   Cynthia Omokaro <cynthia.omokaro@yale.edu>
To   statalist@hsphsun2.harvard.edu
Date   Thu, 14 Aug 2008 12:13:53 -0400

Hello,
I am trying to run a meta analysis using Stata but having difficulty with the
plots generated by meta or metan. I am using the meta code and wanted the
effect size in log scale but unable to do so. Can you please send me the code.
Also, i tried using the binomial option on metan but the graph is not in log
scale even though I checked off log scale.  I noticed that it is using the
approximate method for the confidence interval instead of the absolute. I am
assuming that one of the reasons while the scale is not in log scaled is
because some of the CI is negative.  Further inquiry shows that the approximate
method gives a negative CI, but the absolute method (i.e. Cii sample size count)
does not give a negative CI.  However, metan or meta are both set-up to use the
approximation method.  the  proportion values are (0.0005-0.008). How do I work
around this?

Thanks
Cynthia Omokaro
Research Coordinator,
Yale University

*
*   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