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

st: [Fwd: Comparing dummy categories – logistic]


From   "J. Jabs" <jaj8@cornell.edu>
To   statalist@hsphsun2.harvard.edu
Subject   st: [Fwd: Comparing dummy categories – logistic]
Date   Tue, 20 Sep 2005 16:22:19 -0400 (EDT)

I have 2 questions at this point:
1. How do I find the effect between dummy categories from logistic
regression? (I will also have this issue when I use ‘regress’ with a
subset of the data (those with tp=1 then the outcome becomes min. in food
prep as a continuous variable).  I was looking at the ‘adjust’ command but
I’m not sure it will do what I want in logistic (or that I understand it)

There are numerous predictors with interactions so I’ll just include a few
to illustrate the point:
xi3: logistic tp age i.agekid partner i.hrwork agekid*partner
i.prhrwork*partner

tp=outcome of time in food preparation (0=none; 1=any)
age is continuous
i.agekid has 4 levels (age youngest kid: 0 =kid 0-4yr; 1= kid 5-12yr; 2=
kid 13-17yr; 4=no kids)
i.prhrwork has 5 levels (hours partner work: 0=full time; 1=part-time;
2=vary; 3=not work; 4=no partner)
partner is categorical (0=no partner; 1=partner)

I understand how to compare the levels to the reference category (ie: for
i.agekid for categories from 0 to 1 or 0 to 2 ), but how do I compare
between groups (ie: for i.agekid for categories from 1 to 2; 2-3; 1-3)?

2.  Is there a way to account for those without partners? For example in
the interaction i.prhrwork I can compare differences in partner work time
as compared to full time, but it does not seem that the comparison of 0 to
4 (partner works full time to no partner) is logical?  I have this issue
for numerous variables that have interactions besides the one described
(no kids vs. age youngest; vs. # kids).
The statistics people here do not have an answer so does anyone have
suggestions?  I'm sure that other people have dealt with this?
I’ve searched online and stata archives, though haven’t located anything
of help. Any assistance would be appreciated.

Regards, Jennifer
jaj8@cornell.edu
(I addressed this to wrong email so sorry if it gets cross posted)




*
*   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