Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

st: new package simpplot available from SSC


From   Maarten Buis <[email protected]>
To   [email protected]
Subject   st: new package simpplot available from SSC
Date   Mon, 18 Jun 2012 16:51:08 +0200

Thanks to Kit Baum a new package, simpplot, is now available from SSC.
This package is intended for use after a simulation in which you
stored the p-values of a test. Such a simulation experiment is
typically done to check the coverage of that test. The logic of a test
is that it is a good procedure for making decisions because you will
incorrectly reject a true null-hypothesis only alpha % of the times
you apply that procedure, where alpha is the level of significance of
your choice, typically 5%. For more complex tests the computation
typically depends on some asymptotics and/or some approximations, and
we might wonder whether the chance of rejecting a true null-hypothesis
is really (close to) the nominal level of significance. We could check
that by repeatedly creating a random sample in which the
null-hypothesis is true, perform the test, and store the p-values of
each test. We should reject 5% of the iterations at the 5%
significance level, 1% at the 1% significance level, 10% at the 10%
significance level, etc. -simpplot- takes the p-values returned by a
simulation and compares the observed coverage with each possible
nominal significance level by showing the deviation from each nominal
significance level. Moreover, it also shows the region in which one
would expect most simulations to remain if we were to repeat the
simulation experiment many times and we were simulating a well behaved
test. An example of such a graph can be seen here:
<http://www.maartenbuis.nl/software/simpplot.html>.

As always, this package can be installed by typing in Stata -ssc
install simpplot-.

I hope some of you will find it useful,
Maarten

--------------------------
Maarten L. Buis
Institut fuer Soziologie
Universitaet Tuebingen
Wilhelmstrasse 36
72074 Tuebingen
Germany


http://www.maartenbuis.nl
--------------------------
*
*   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 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index