I have survey data from households in Mexico. The data contains information about 32 states, the decile of household income that each household belongs to (10% of the nation in decile 1, 10% in decile 2, etc.) and whether or not the household has electricity (yes or no). Since some states are poorer than others, the percentage of a poorer state population with no electricity within decile one will be higher than the percentage of a richer state population with no electricity within decile one. I want to create a graph that shows this by having income decile 1-10 on the x-axis, electricity access (yes/no) as the cluster x variable, and the minimum, maximum and median percentage of population on the y-axis.
I can find these percentages by splitting the file by state (into 32 groups) and then using crosstabs with income decile as the columns and electricity access as the rows, and then using the percentage within each decile. (So the percentage of each state population within that particular decile that has or does not have electricity is shown in the table). I cannot figure out, however, how to plot just the maximum, minimum, and median values of these percentages, since boxplot is not an option when plotting directly from a crosstabs table.
Any suggestions would be greatly appreciated.