Suppose I wanted to predict the likelihood of and degree of delays and cost over-runs associated with a nuclear plant currently under construction. How would people recommend I do so?
Reference class forecasting: get a list of previously constructed nuclear power plants, look up how much they were delayed and over budget, then use the empirical probability distribution of delays and cost over-runs. (Bent Flyvbjerg, cited by Tripitaka, turns out to be very keen on RCF.)
Suppose I wanted to predict the likelihood of and degree of delays and cost over-runs associated with a nuclear plant currently under construction. How would people recommend I do so?
Study existing literature. http://en.wikipedia.org/wiki/Bent_Flyvbjerg this guy got a lot of good press in germany, apparently he has written extensively on big infrastructure projects and cost overruns. I find Megaprojects and Risk: An Anatomy of Ambition
Reference class forecasting: get a list of previously constructed nuclear power plants, look up how much they were delayed and over budget, then use the empirical probability distribution of delays and cost over-runs. (Bent Flyvbjerg, cited by Tripitaka, turns out to be very keen on RCF.)