Skeletal cases

Want to develop students' investigatory skills? Then instead of presenting complete case-studies, provide only the initial "presenting information", and only add to it in response to their questions. There are a number of variations on the theme:

• Background: Provide just the superficial information with which a student might be confronted on first acquaintance with the situation. Feed in additional background only if asked about it: this is useful for fairly basic professional skills of gathering data on cases. After a set period of time or number of questions, see how efficiently the students have got at the important material.

• Investigation: A more sophisticated variation, perhaps starting from a higher information base, is to offer to answer questions, but only when the student explains why she is asking. This identifies the issue of testing hypotheses, and the question of what evidence might be gathered in order to test an hypothesis. It applies clearly to medical diagnosis, but it could also apply to fault-finding in a computer network, or a car engine.

• Potential accounts: Almost the reverse strategy, particularly useful on in-service courses with practitioners who are locked into recipes of the order of "When such-and-such happens, I always...": outline the presenting situation and asked them to multiply as many potential explanations as possible, before eliminating any of them. This is a good exercise for the divergent thinkers in the group, and for any occasion when you wish to encourage what John Keats called "negative capability"—the art of not jumping to conclusions, but tolerating uncertainty and confusion where necessary.

• Raw data: a variant on the above is to start with raw statistical or documentary data.

It strikes me that these all find their way in various ways into what Shulman calls "signature pedagogies"—the distinctive ways of teaching which are characteristic of many professions—such as the ward round in medicine. This linked article bemoans the lack of a signature pedagogy for teaching. I would have thought there were many, including microteaching.

Of course you don't always have to generate the problem yourself. If you are teaching people with some experience under their belts, they are very likely to come with their own real-world cases, and those can generate some of the most fruitful teaching sessions you will ever have. (Assuming that the "cases" have not merely been contrived to test you out). However, don't just engage with the case yourself, and don't simply "throw it open" to everyone. The idea is that the process can work on two levels (at least);

• the substantive level of working on the case under discussion, and also
• the process level of discovering the best way(s) of approaching that discussion.

If you miss out on reflection, with the class, on the second level you lose much of the benefit.