[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
Re: st: Meta-analysis and corr2data
At 01:48 AM 3/9/2004 -0800, n p wrote:
I think you could. You would run corr2data multiple times to create each
of the fake data sets, adding an indicator to each data set to indicate
which group it was (e.g. gen group = 1, gen group = 2, etc.) Then, combine
all the files together.
statistics etc.). The question is if it is right to
use the corr2data command to create three (or more)
independent samples for A, B and C arms with given
mean, SE and N to obtain via regression two betas and
two SEs (B vs. A and C vs. A). I know that I could use
only two arms for each meta-analysis ( B vs. A, C vs.
A and so on) but I think that this would induce some
multiple comparison error so I decided that I have to
calculate all possible betas of each study ( B vs. A,
C vs. A and so on) simultaneously.
If in doubt as to whether what you are doing with corr2data is
legit: Generate alternative data sets, using a different seed. Results
should be exactly identical, down to several decimal places. If not, what
you are doing is probably illegitimate. For example, computing logs of
variables would not be legitimate.
Basically, the idea behind corr2data is that you can use it when the means,
correlations and standard deviations are sufficient to do all the required
calculations. If anything more than that is needed, corr2data can't handle
it. In some programs, like SPSS, you could enter means, correlations and
SDs directly. In Stata, you achieve the same thing via the creation of
fake data with the desired means, etc. But, you have to remember that the
data are fake! And what you can do is limited.
Richard Williams, Notre Dame Dept of Sociology
OFFICE: (574)631-6668, (574)631-6463
WWW (personal): http://www.nd.edu/~rwilliam
WWW (department): http://www.nd.edu/~soc
* For searches and help try: