st: missing values & ice & categorical variables

[Andrea Bennett <mac.stata@gmail.com>]

Dear Statalister,

Is this generally the right approach to deal with survey data?

1. get rid of missing values with package such as the "ice" command  
(first deciding which variables one wants to use)

2. generate dummy variables for categorical variables (such marital  
status or level of schooling)

3. perform the regression

There is no way of generating the dummies out of the categorical  
variables and then dealing with the missing values, right? At least  
not, if I would like to work with the "ice" command (I could though  
start by recoding the missing values as dummies, first, but this seems  
to lead to distribution issues, I think)?

Many thanks for the considerations,

Andrea
*
*   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/