The course I'm taking this week is based on this: http://www.sei.cmu.edu/tsp/psp.html
So far we've been studying methods of logging our time while programming - keeping track of how long it takes to fix each bug, etc.
Lots of repetition. Lots of reminders not to be lazy, reminders why we need to do this. Most teachers seem repetitive and slow to me, though. There's some propaganda around. Many of the people here are subvertly skeptics (and not here by choice, exactly). Probably the teachers are expecting skepticism, as well. Unfortunately, this means the lot of us get the sermon.
There's some questionable science in the book for this course. The ideas are good, but the data given to back them up is weak in several places. That's ironic, since the course is all about gathering data carefully and completely. Many of the ideas, however, could be called obvious and useful. The problem areas are mostly acknowledged and left as an exercise for the reader. (example: it doesn't really matter that the test cases in the book have insufficient data since you need to gather your own anyway.)
Things they have right:
Collecting data is important.
This process is very personal and private. You cannot effectively compare people by comparing the data. The process must be self-induced.
One of the presenters as an entertaining French accent.