bootstrapping p-values

[Nicolas Thomas Ewbank]

Does someone have a nice and understandable way to bootstrap p-values for the t-test? We've looked online but none of it seems to make any sense. We were hoping one of you could help us out

[Carlo Lazzaro]

Nicolas;
did you take a look at Example 3 in -bootstrap- entry, Stata 13.1 .pdf manual?
As an aside, if I understand well your post, you would actually bootstrap the -t- of -ttest-, not the p-value.

[Nicolas Thomas Ewbank]

Thanks for the reply Carlo. I can not however find the example. might you have a link?

[Carlo Lazzaro]

Nicolas,

I have no link, but the running title and the page you're after are: bootstrap — Bootstrap sampling and estimation 201.
By the way, you can easily have access to all this stuff by simply typing -help bootstrap- from within Stata and scroll down the menu of the -Also see- button top right the screen.

[Nick Cox]

http://www.stata.com/manuals13/rbootstrap.pdf but (1) you're assuming to be the latest version of Stata unless you tell us (1) the manual comes with Stata.