----- Original Message -----
From: <[email protected]>
To: <[email protected]>
Sent: Monday, May 12, 2003 12:29 PM
Subject: st: bootstrapping of skewness-adjusted t-statistics
> Dear Stata-Listers,
>
> I'm analyzing initial returns and long-run performance data for a sample
of
> some 200 initial public offerings. The data is skewed to the right side,
all
> tests reject the assumption of normality. Instead of using a simple
one-sided
> t-tests to find out whether the returns differ significantly from zero I
> want to run a bootstrapped skewness-adjusted t-test. I was able to
calculate the
> skewness-adjusted t-statistic (suggested by Johnson 1978) in excel, but I
> have no idea how to bootstrap the test. The literature on that issue can
be
> summarized as follows:
> - take 1000 subsample of size n/4 from the original sample
> - calculate skewness-adjusted t-statistics for each subsample
> - compute the standard deviation of these 1000 t-statistics
> - standardize the t-statistic of the empirical sample by dividing through
> the standard deviation of the bootstrapped 1000 t-statistics
> - compare the resulting value to the corresponding critial value of the
> standard normal distribution.
>
> Thanks in advance for any suggestions on the johnson t-test in stata and
the
> bootstrap command (whether it is bstrap, bs, bstat or bsample)
>
> Jan Kuklinski
> Witten/Herdecke University
>
It would not be safe to accept the values that Excel calculates. For a
recent review of Excel see:
B. D. McCullough and Berry Wilson "On the Accuracy of Statistical Procedures
in Microsoft Excel 2000 and Excel XP"
Computational Statistics and Data Analysis 40(4), 713-721, 2002
Abstract
"The problems that rendered Excel 97 unfit for use as a statistical package
have not been fixed
in either Excel 2000 or Excel 2002 (also called Excel XP). Microsoft
attempted to fix errors
in the standard normal random number generator and the inverse normal
function, and in the
former case actually made the problem worse."
However, there is a Johnson test in Stata. Try -findit johnson- to locate
and download.
Hope this helps,
Scott
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
