The simplest way is to restructure your data from wide to long (VARSTOCASES), split the file by ID, invoke OMS-OMS END, run a regression command, open the OMS save file and, there are the B's by id. It reads like you'd like to compute/plot correlations between the B's and other variables and to do that you'll need to match files the OMS save file (you may want to do some file cleaning out at this point, which you'll understand when you look at the variable list) back to the original dataset.

Gene Maguin

There are examples from Singer and Willett for R too:

http://www.ats.ucla.edu/stat/r/examples/alda/ch2.htm

  Here is a link to all examples for the text, showing other chapters and programs:

http://www.ats.ucla.edu/stat/examples/alda/default.htm

See the examples for Singer and Willett's Applied Longitudinal Data Analysis at UCLA's Institute for Digital Research and Education site.

http://www.ats.ucla.edu/stat/spss/examples/alda/chapter2/aldaspssch2.htm

  Examples for other chapters may be useful when you start to model.

thanks for your shairng