Changing from length interval to singleton length does not change the fact that the rest of the design is
fold (6) set (3) subset(2) MoSS (2)
So, if we do a Bonferroni table for the entries from all of these cells for each singleton length interval, there will be 72.
This is the same number as yielded OK results for Auq, so I will proceed in this way unless you want me to construct the Bonferroni tables differently.
Of course, there will be Bonferroni tables for each of the average slopes for each of the three regressions, and perhaps the average covariances (you haven't mentioned yet how you want to operate on/with those.)
