Cross table for "select all that applied" using PROC REPORT

[xeonchenzi]

Hi,

I have a problem to create cross table for survey data contained "select all that applied" questions. The data is like below (Headache and Insomnia are two options in one "select all that applied" questions, so they may overlapped):

ID	Treatment	Headache	Insomnia
1	1	0	0
2	0	1	0
3	1	1	0
4	1	1	0
5	1	1	0
6	0	0	0
7	1	1	1
8	1	1	0
9	0	1	0
10	1	1	0
11	1	1	1
12	1	1	0
13	0	1	0
14	0	0	0
15	1	0	0
16	1	1	0
17	1	0	1
18	1	1	0
19	1	1	0
20	1	1	0

The usual cross table is var by Headache or var by Insomnia separately with (0,1)x(0,1).

But I wonder how I can get a cross table crossing all options = 1 together, like the following table:

Treatment	Headache=1 [N (%)]	Insomnia=1 [N (%)]
1	xx (a%)	xx (b%)
0	xx (c%)	xx (d%)
Total	n_1 (100%)	n_2 (100%)

In this case, only a+c=100% make sense.

For the completeness, I also wonder how to get a "transposed" table like:

	Treatment
	1  [N (%)]	0 [N (%)]	Total
Headache=1	xx (a%)	xx (b%)	n_1 (100%)
Insomnia=1	xx (c%)	xx (d%)	n_2 (100%)

In this case, of course only a+b=100% make sense.

Hope someone could help. Thanks so much!

[Reeza]

Have you tried PROC MEANS?

proc means data=have noprint;
class treatment; 
ways 0 1;
output out = want sum= N mean = PCT; 
run;

With 0/1 variables if you take the sum that's the equivalent of counting the number of 1s. If you take the mean, that's the same as calculating the percent. 

The CLASS statement tells SAS to calculate it by treatment. 

WAYS tells it to do the total and one way summaries. 

WANT data set will have everything you need to create your table and then you'll need to figure out how to display it.