Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
From | Richard Williams <richardwilliams.ndu@gmail.com> |
To | statalist@hsphsun2.harvard.edu, statalist@hsphsun2.harvard.edu |
Subject | Re: st: Sample size for maximum likelihood estimates |
Date | Thu, 11 Aug 2011 12:05:54 -0500 |
At 10:57 AM 8/11/2011, Maarten Buis wrote:
On Thu, Aug 11, 2011 at 5:20 PM, dk wrote: > Take example If, my dependent variable is respondent go for purchasing > x product or not. I have response 60 % purchase and 40 % do not > purchase. Then is it possible to use 20 explanatory variables. That still depends on the distribution of your explanatory variables (low variance means less power which means more observations are needed), the correlations between these variables (higher correlations means less power which means more observations are needed) and whether or not these include interaction terms (interaction terms mean less power which means more observations are needed). I am sorry, there really is no generic answer to your question. -- Maarten
Some of the suggestions that Scott Long has tossed out are summarized on pp. 3-4 of
http://www.nd.edu/~rwilliam/stats3/L02.pdfAs Maarten says, there doesn't seem to be any nice simple formula for deciding what the right sample size is. But it does seem like you need a larger sample for a logit or probit analyses than you would for an OLS regression.
------------------------------------------- Richard Williams, Notre Dame Dept of Sociology OFFICE: (574)631-6668, (574)631-6463 HOME: (574)289-5227 EMAIL: Richard.A.Williams.5@ND.Edu WWW: http://www.nd.edu/~rwilliam * * 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/