While in Stata7SE the command works properly, in Stata8SE it just skips
some variables, i.e. not for all variables sums are calculated.
For the example cited above only means of age, highlengstay, and
lowlengstay as well as sums of d0age_11, mort30, highvol, highmort,
lowcount, and lowmort are calculated.
Has anybody an idea why it doesn't work correctly or has anybody faced a
similar problem already (although I have looked for any contributions in
this matter, I couldn't find any).
