Is there a way to do a crosstab of two categorical
variables, say, using -tabulate- and get confidence
intervals for the proportions? I know -svytab- does
this for weighted data, but it looks like I have to do
it in two steps for unweighted data from a simple
random sample. Thanks, B. Pizacani
One possible solution might be to use official Stata's -binreg- (or -glm-
with -link(iden) family(bernoulli)-), together with the -parmby- program in
the -parmest- package, downloadable from SSC. -parmby- creates an output
data set with 1 observation per model parameter and data on confidence
intervals and P-values. For more information on how to use -parmby-
together with other packages to produce tables of confidence intervals, you
might like to see my Stata Journal article (Newson, 2003), which you can
download as a pre-publication draft from my website (see my signature
below), using either a browser or the Stata -net- command.
