Dear Statalisters,
When using -xi- to create indicators for many categorical variables in
a regression, I cannot help noticing -xi- has some advantage that when
the original variable name is long, the name of indicator crated are
cut off somewhere then the name of the outcome variable is kind of
unexpected. I have to adjust the indicator name one by one when the
name is long. This causes inconvenience when using -xi-.
But on the other side, if there are many interactions of many
categorical variables, the original way by creating each dummy
variable for each interaction is also time-consuming.
I am curious: can some one here share how you decide to decide to use
-xi- or not to run a regression with lot of categorical variables
interactions?
Thank you for your time!
Best wishes,
Amanda
*
* 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/
