So here is the raw data. Not unexpectedly the very first to the 2nd note is much more likely to be a climb than a decrease, and just the opposite for the final two notes (although as I mentioned in the first post, that was a bit difficult to judge for many of the songs I was using in that they often faded out rather than ended). I should point out that I only evaluated note sequences that went up or down (I didn't count the many instances in which 2 consecutive notes stayed the same). I was surprised by the low frequency of major 3rd intervals. Anyway, it was a fun project.