This exercise differs from the first case study in being much simpler and shorter. It was run during a two-hour tutorial with a group of ten students, from the same 1st year astrophysics class as the Zog exercise. It was designed as the capstone of a set of tutes focussing on back-of-the-envelope calculations, and was not assessed.
Once again, I devised a fictitious astronomical mystery. In this case, it was a halo red giant star with an anomalous radial velocity. The students were presented with a sketchy initial briefing (in the form of a bogus poster paper), and were allowed to divide themselves into teams. Calls for proposals were issued every twenty minutes. I'd pre-faked the pieces of data I felt it most likely the students would ask for. If they asked for something different, I made it up on the spot. From time to time, I'd issue fake press releases, reporting new discoveries about this star from other groups.
At the end, groups had to stand up and present what they had learned to the remainder of the class. The exercise was not assessed.
Once again, this exercise was extremely popular with the students. They found it compulsive, and loved the taste of real research. Surprisingly, the biggest source of error was unit conversions: students are so used to having problems presented to them in sanitised units that they found real data, with a mix of units, very hard to cope with.
This smaller scale exercise, run in class, avoided the group-work and time-management problems of the ``Universe of Zog'' exercise. On the other hand, it required considerably more class time, and covered much less of the course material.