Since there are four variables to compare, you have to consider what kind of table would be the easiest to analyze the result. In this case, since variable for gender (gender_n) only has two values, you might want to create two separate tables of female and male using by gender_n. You should use lookfor to find out what is the variable used to see attitudes toward the new government (new_govt). Also, to exclude invalid answers, check the survey and you will find that negative numbers are invalid answers. Thus, use if new_govt>0 at the end.
lookfor government
132. new_govt byte %9.0g new_govt 8 :effect of new governmentsort gender_n
-> gender_n= F
Means of monthly gross pay
19 |
:populatio | 8 :effect of new government
n group | 01-bette 02-same 03-worse | Total
-----------+---------------------------------+----------
01-afric | 614.03448 527.96429 830.52 | 635.47857
02-colou | 957.2 1140 918.69231 | 975.17391
03-india | 2733.3333 1200 1750 | 2150
04-white | 2753 3679.1429 2764.5714 | 2963.1875
-----------+---------------------------------+----------
Total | 782.0101 1157 1545.2787 | 1090.1393
-> gender_n= M
Means of monthly gross pay
19 |
:populatio | 8 :effect of new government
n group | 01-bette 02-same 03-worse | Total
-----------+---------------------------------+----------
01-afric | 960.06716 581.96875 1407.12 | 955.2356
02-colou | 1067 1159 1094.8889 | 1101.913
03-india | 3320 2004.6667 2163 | 2330.1818
04-white | 7266.6667 6516.875 4748 | 5712.7556
-----------+---------------------------------+----------
Total | 1249.5 2383.5263 2654.1905 | 1816.6667
There doesn't seem to be any clear trend in income distribution among the races and attitudes about the government by gender.