Raphael can tinker with means and sds, but that still won't guarantee
that values fall within the range specified. That points up that the
original question is not well-posed without a specification of
distribution. On face of it, it is unlikely to be normal.
Nick
[email protected]
[email protected]
Raphael, try the code below. you can change the standard deviations,
etc.... to get what you want.
*/----------- START-----------------------
clear
set obs 10
drawnorm mean1 mean2 mean3, means(0.09 0.11 0.18 ) sds(0.04 0.1 0.1)
gene id = _n
reshape long mean, i(id) j(group)
drop id
sort group
oneway mean group, scheffe tabulate
*/------------END-------------------------
This code will generate 3 groups with equal size (n=10 - in the example
above). You can create a loop to get a dataset with a specific
characteristics or copy-paste the cole below in trial-error attempts.
Raphael
I would like to generate three groups of random numbers ranging
between 0.01 and 0.20, where one of the group is statistical
significant from the other groups. Can any one help?
*
* 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/
