I want to correlate many continuous variables (e.g. A1-to A20) to one ordinal variable (B)
Via the Fit model -> multivariate I get the results for all possible combinations, not only A1 to B, and A2 to B, etc. but also e.g. A1 to A2, A2 to A11, etc.
I am only interested in A1 to B, and A2 to B, A3 to B etc. You may say I could sort the data and then just use the correlation pairs I need. While this would work in this example, it will not work easily if you have A1 to A1000 or so variables and you want to know the correlation of each to B, as the number of pairs grows dramatically.