st: Interactions of 2 explanatory variables within a categorical variable?

[Maggie Skiles <maggie.skiles@gmail.com>]

Dear all,

I am performing a logistic regression with variables sex (binary), age
(binary), income (maybe linear?) and region (categorical - 3 dummies,
4 categories).

I am wanting to look at how these variables play a role on my outcome.
But, I am also trying to see how these variables - sex, age, and
income - play a role WITHIN each region. Just based on the EDA, these
do change largely within the regions.

I have two questions:

1) I realize that I can look at interactions of region 1 and gender,
region 2 and gender, etc. Is there any way to look at the interaction
of 2 variables WITHIN a region. For example, does the effect of income
change by age WITHIN a category?  (I have chosen to not create
subgroups for regions, for various reasons)

2) After doing a lowess plot, it appears that my continuous variable
(income) is curvelinear and requires a linear spline with a knot at
45. However, when I look at the effect of income within each region,
the curves vary greatly (one looks like the knot could be at about 15,
one is nearly linear, one is more of a inverse w shape), so it seems
the knot at 45 is not really applicable. I could create categories for
income instead. But, does anybody have thoughts about how to go about
this? I would really like to assess how income plays a role within regions.

Thanks in advance,
Maggie
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/