Who should attend
Participants should have a grounding in statistical methods (models, parameters, estimates, covariance matrices, confidence intervals, concepts of repeated sampling, and some appreciation of the underlying theory (least squares or maximum likelihood)).
Equally important is a knowledge of the subject matter and models used in the subdiscipline. Modern data analysis is far more than statistical theory and numerical methods; this fact is a major emphasis of Akaike's work, the book by Burnham and Anderson (2002), and the textbook by Anderson (2008).
Often people in a two- or three-person team attend these sessions and this is sometimes ideal. Usually, one or two people have deep experience in the subject matter, while the other person has good quantitative and computer skills. This team situation often benefits from the workshops offered.
Everyone learns a lot from even a one-day workshop. Those with a modest background in the statistical sciences learn new approaches and begin to ask different questions. Those people with a more substantial quantitative background will also learn a lot and begin the rethink their training in methods based on null hypothesis testing. Most attendees are intrigued by subjects such as model selection uncertainty, model selection bias, the nature of quantitative evidence, and the simplicity of the main concepts. The sessions are informal and encourage interaction as the material is presented.
Information-theoretic methods are based on deep statistical theory; however, the application of these methods is relatively simple. The methods are easy to understand and compute; they seem compelling.