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]

Re: st: Box-Tidwell Test


From   Nick Cox <[email protected]>
To   "[email protected]" <[email protected]>
Subject   Re: st: Box-Tidwell Test
Date   Tue, 26 Apr 2011 13:34:36 +0100

Sorry, but this is another version of the same question and the answer is the same too. Lowess can't help you with dummies. What you can write on linearity is that the issue does not bite or even arise with your choice of predictors.

(I don't know what there is to 'try' in my previous, which was a one- sentence demonstration of the fallacy you have swallowed.)

Nick

On 26 Apr 2011, at 12:57, [email protected] wrote:

Hi


Ok I will try this. But I have another questions regarding testing the linearity assumption. Another way to check is the Locally Weigthed Scatterplot Smoothed (lowess) and the literature says that graphs with u- shaped and not continues shapes contravene the assumption.

I tried the test with:
lowess wabsunion themen, jitter(2) bwidth(.2)

and got as a result, a straight line, with some slope. It is for sure not s-shaped form as it is supposed to be, yet it is not u- shaped and continuous. So is it ok to write that the linearity assumption is ok? I know the case is special as I have 2 dummy variables with values (0-1), so its seems logical not to have a s- shaped graph, but I have to write something about linearity.

I am very sorry, i know my questions must seem totally stupid but this is the first time I use STATA and regression
*
*   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