Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

st: interpreting standardized coefficients in multiple regressions with product terms


From   "Pek-Hooi Soh" <psa28@sfu.ca>
To   <statalist@hsphsun2.harvard.edu>
Subject   st: interpreting standardized coefficients in multiple regressions with product terms
Date   Tue, 3 Mar 2009 12:18:27 -0800

Hi Stata users,  I used xtnbreg and standardized some of the key variables (listcoef does not work for nbreg) and estimated the regression.

Regarding the interpretation of standardized coefficients, I found some examples in Long and Reese (page 377-8) but they don't show regressions with interaction terms. 

If I have a standardized regression with standardized coefficients -0.2 for X1 and 0.56 for X1 and X2, will a change of 1 st dev in X1 give the following effect on Y?
-0.2 + 0.56*(X2)

The change of 1 st dev in X2 is this.
1.67 + 0.56*(X1)

Do I assume a value of X1 and X2 in each case, by taking the min and max and calculating the resulting effect? 

How do I compare the effects of X1 and X2 on Y if there effects range from a negative value to a positive value? 

In an article it says if a standardized coefficient is closer to 1, it has a stronger effect on Y. Does this apply to a standardized regression with product terms?

Will appreciate your explanations and help.

Cheers

Pek



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