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: RE: RE: RE: When to use ML or bootrapping


From   Nick Cox <[email protected]>
To   "'[email protected]'" <[email protected]>
Subject   st: RE: RE: RE: When to use ML or bootrapping
Date   Tue, 30 Nov 2010 15:27:58 +0000

Not clueless, but these questions are well covered in the literature.... 

Nick 
[email protected] 

Amy Dunbar

Thank you for the references, Nick.  I knew they weren't alternatives, but my students asked me about bootstrapping in a class, and I realized I didn't know when one would use that technique.  I added the ML question because that has been something I wondered about for a while.

For example, if ML can be used for linear models, why would you use ML instead of OLS?  And yes, I realize just how clueless that must sound.

On Behalf Of Nick Cox

These aren't really alternatives in any sense. Indeed it is common to combine maximum likelihood (ML) estimation and bootstrapping. 

Often, ML provides estimates and bootstrapping gives (more) honest indications of their uncertainty than you might otherwise get. 

However, you can bootstrap anything you can program, more or less, including procedures in which ML cannot be specified or is not possible for whatever reason. 

Conversely, ML might well be possible in circumstances in which bootstrapping is problematic (e.g. many time series problems). 

The book you refer to doesn't aim to cover likelihood from scratch. You need an intermediate book that covers both in context. I like 

Davison, A.C. 2003. Statistical models. Cambridge U.P. 

Rice, J.A. 2007. Mathematical statistics and data analysis. Duxbury or whatever it's called this year. 

But the choice is very personal, and you might well prefer something with more business or economic examples. 

Nick 
[email protected] 

Amy Dunbar

Although these questions are not directly Stata related, I hope someone will suggest sources to help me.  I want to know how one knows if you should consider ML or bootstrapping.

I read the preface to Maximum Likelihood Estimation with Stata, Fourth Edition, http://www.stata.com/bookstore/pdf/ml4-preface.pdf
I don't see a chapter on when ML is appropriate, unless that is covered in the practical implications discussion.

As for bootstrapping, Chapter 13 in Cameron and Trivedi helped me understand how to use bootstrapping, but not when I should consider using bootstrapping.

I apologize for my newbie questions, but I would really appreciate some intuition.


*
*   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