It has been some time (almost two years) but I am back teaching one of my favorite electives - Advanced Analytic Techniques.
This course is unlike any of the other courses I teach. Rather than focus on a specific body of knowledge, this course allows students to explore their own interests while learning to use, and more importantly, evaluate various analytic techniques used by intelligence professionals.
While each student is hyper-focused on a single technique and topic, each week we take a quick look at a technique that no one in the class is examining; something the class is interested in that we would otherwise not be able to get to.
The week starts with each student going out and finding relevant articles from peer reviewed journals and elsewhere which they then summarize and post to the Advanced Analytic Techniques blog. Each student then reads the summary and votes on whether or not they thought the article was "interesting" or not. They are also required to post a couple of comments in order to get the dialogue going and to give the original poster some feedback.
From these articles, we are trying to get a sense of the technique -- How to describe it, what are the technique's strengths and weaknesses, how to actually use it in practice, etc. We are also trying to begin to evaluate the technique. We are not trying to evaluate the technique in general, though. Rather, we are trying to evaluate the technique with respect to its utility in intelligence analysis.
Specifically, we are looking to see if the technique actually improves forecasting accuracy, if it is relatively simple (or, at least, if complex, does that complexity pay off with remarkably better results), can it be used across intelligence disciplines (i.e. is it flexible), if it works well with the kinds of unstructured data typical to intelligence analysis and, finally, if the technique facilitates the communication of the results to a decisionmaker.
Once we get into class, one of the teams conducts an exercise utilizing the method. The exercise is designed primarily to give us a feel for how the technique works in practice. Due to time constraints, we typically try to keep this exercise focused on the core elements of the technique.
Finally, we put together the posts that summarize what we have learned about the technique over the week. Since I have a fairly large class (large, at least for a graduate seminar...), I have two teams that work mostly independently on their posts. Comparing these two views of the same topic, based on the same journal articles and the same exercise, but with often dramatically different interpretations, is often a learning experience to itself.
So far this term we have looked at
- Multi-criteria Intelligence Matrices
- Decision Trees