I want to generate a Kruskal-Wallis test p-value using columns in my existing dataset: Y and X. But, I need to do this for every separate combination of two other columns in my dataset: {X2 & X3.} In some software these two extra columns, X2 and X3, would be placed simultaneously into a "by" variable. But I don't see how to include the other variables, X2 and X3 in the Kruskal-Wallis test in Spotfire. Would like to see this implemented.