12-07-2011 01:46 PM
I am in the process of exploring the number of variables needed for a data linkage.
Would like to know after stratifying the data, whether there are cells that have frequencies greater than 0 or 1 and the magnitude of these frequencies.
What is a quick procedure for calculating 1) the average cell frequency across multiple variables (e.g., Age x Gender x Year of Birth) and the 2) Min and Max Cell count?
12-07-2011 03:16 PM
You can use ODS to obtain the crosstabfreqs output table .. which you could then use as the data for doing all of your analyses. Take a look at: http://www.nesug.org/proceedings/nesug01/cc/cc4025.pdf
It shows how to obtain such tables. You could simply use proc means or proc univariate to analyze the data in such a table.