Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: Box-Cox estimation


From   Eva Tung <eva2tung@yahoo.com>
To   statalist@hsphsun2.harvard.edu
Subject   st: Box-Cox estimation
Date   Thu, 6 May 2004 07:54:36 -0700 (PDT)

Hello all,
 
I have the following questions about Box-Cox
regression.
 
1. In the test of significance of the Box-Cox
transform parameters (theta and lambda), the p-value
for lambda in my model is greater than 0.05. I read
the reference manual and it interprets an
insignificant lambda as right-hand-side transformation
not significantly adding to the regression. However,
it seems that the null hypothesis is lambda being zero
which implies a natural logarithmic form of the
variable. Does an insignificant lambda in fact mean
that natural logarithmic form of the variable cannot
be rejected?
 
2. When only one independent variable is transformed,
the output states that lambda is not identified in the
restricted model for LR test but test statistic is
still provided. Is the statistic valid for hypothesis
testing?
 
3. There is no robust option for Box-Cox
estimation. How can I obtain robust standard errors
for the estimation?
 
Thanks in advance.
 
Eva 





	
		
__________________________________
Do you Yahoo!?
Win a $20,000 Career Makeover at Yahoo! HotJobs  
http://hotjobs.sweepstakes.yahoo.com/careermakeover 
*
*   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/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index