Although I only read parts of this study, it basically was saying that when attempting to measure students undergraduate experience you will find greater variance among majors within an institution than between equivalent majors across institutions.  See: