--- Claude Francoeur <[email protected]> wrote:
> We were expecting to have 2 bends in our model (y = X1 X1squared
> X1cubed). <snip> How should we interpret these results?
The easiest way to interpret a non-linear regression curve is to graph
the predicted value against the original explanatory variable. -adjust-
can be quite handy here, as it allows you to fix the other explanatory
variables at certain values (and can create confidence intervals around
the predicted values).
*------------- begin example ------------------
sysuse auto, clear
sum weight
gen z_weight = (weight-r(mean))/r(sd)
gen z_weight2 = z_weight^2
gen z_weight3 = z_weight^3
reg mpg z_* foreign price
adjust foreign = 0 price, by(weight) ci replace
twoway rarea lb ub weight, sort || ///
line xb weight, legend(off) sort
*--------------- end example --------------------
(For more on how to use examples I sent to the Statalist, see
http://home.fsw.vu.nl/m.buis/stata/exampleFAQ.html )
-----------------------------------------
Maarten L. Buis
Department of Social Research Methodology
Vrije Universiteit Amsterdam
Boelelaan 1081
1081 HV Amsterdam
The Netherlands
visiting address:
Buitenveldertselaan 3 (Metropolitan), room N515
+31 20 5986715
http://home.fsw.vu.nl/m.buis/
-----------------------------------------
*
* 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/
