Search code examples
statasummary

How can I create a higher dimensional table?

I want to create a table in Stata 14, where the relative frequencies of the sexes are conditioned on whether the person is employed or not:

table example

I have tried to use the community-contributed command tabout, but I have not found a specific example that deals with this. As a workaround, I generated a variable with levels for a person that is "Male & Employed", "Female & Employed" and so on.

Is there a solution that does not require generating additional variables?

Solution

  • Consider the following example using Stata's auto toy dataset:

    sysuse auto, clear

egen price2 = cut(price), group(4)
label variable price2 "Price Category"
label define price2 0 "Lower" 1 "Middle" 2 "Upper" 3 "Luxury"
label values price2 price2

recode rep78 (3=1) (4=2) (5=2)
label define rep78 1 "Good" 2 "Poor"
label values rep78 rep78

    The community-contributed command tab3way provides the means to create such cross-tabulations:

    tab3way price2 rep78 foreign, rowpct nofreq rowtot format(%5.2f)

Table entries are row percentages
Missing categories ignored

------------------------------------------------------------
          |         Car type and Repair Record 1978         
Price     | ------ Domestic ------    ------- Foreign ------
Category  |   Good    Poor   TOTAL      Good    Poor   TOTAL
----------+-------------------------------------------------
    Lower |  53.85   46.15  100.00     25.00   75.00  100.00
   Middle |  76.92   23.08  100.00     25.00   75.00  100.00
    Upper |  45.45   54.55  100.00     14.29   85.71  100.00
   Luxury |  63.64   36.36  100.00            100.00  100.00
------------------------------------------------------------