Hi everyone,
I'm very new to Stata and learning to use it myself.
I am trying to use stata to run frequency table since it give 95% CI as well. However, I seem to have problem with how Stata deals with missing value. I'm referring to the dummy dataset as below
ID Sex Age_grp
1 1 #NULL!
2 1 1
3 1 2
4 1 2
5 #NULL! 3
SPSS will give me the frequency table as below
Count Column N %
Sex Male 4 100.0%
Female 0 0.0%
Age_grp Under 18 1 25.0%
18-30 2 50.0%
31-45 1 25.0%
46-60 0 0.0%
60+ 0 0.0%
Stata
proportion Sex Age_grp, percent

Percent estimation Number of obs = 3
--------------------------------------------------------------
| Logit
| Percent Std. Err. [95% Conf. Interval]
-------------+------------------------------------------------
Sex |
Male | 100.00 0.00 . .
|
Age_grp |
Under 18 | 33.33 27.22 0.26 98.98

18-30 | 66.67 27.22 1.02 99.74

In the table above, it seems like Stata will only include record with none missing value for all the variables (ID 2, 3, 4). In a table with 7-8 variables, e.g. basic respondent characteristics, the Number of obs will be much smaller (maybe only 1/2 of the records) when only including records with none missing values.

If I run "proportion Sex, percent" or "proportion Age_grp, percent" separately, Stata will report 4 obs. However, it will take too much work to merge the Stata output in MS DOS format together to create a table like SPSS can produce.

Could someone please let me know if anything can be done to make Stata report on the 4 obs as in the table above?
Also, is the latest Stata version create output table in format-able Word table, or it is still MS DOS table?

Thanks a lot

Jiro