madeleinelissette

I read this because it is a popular work by two very big names (Keeney and Raiffa) in my own technical field, decision analysis. Since I often have to explain technical ideas for a general audience and recommend courses of action to non-specialists, I wanted to see how K&R do that. So I thought it a great little book, and recommend it highly, except for one curious lapse. It recommends eliciting utility functions without using reference lotteries. Since utility functions capture attitude toward risk, my understanding and belief is that they cannot properly be elicited without explcitly considering chance events, i.e. lotteries. If this is not an inexplicable goof, then I'd have to say that it is a conscious simplification, motivated by the tedious difficulty of elicitation with reference lotteries. In fact, I usually recommend assuming risk neutrality and skipping the lotteries, at least for the first pass through a problem, unless risk is the primary source of difficulty in the problem (as it often is in petrochemicals and pharmaceuticals). Other than that, I thought this was a fine book.

2022-11-14 16:18