Deriving Probability Distributions with Pairwise Relative Comparisons

This presentation supports the FAIR contention that we need to use ranges or distributions for probability and impact in FAIR for risk management. More importantly, this presentation shows how the PERT-styled distributions used in FAIR analyses can be supplemented with pairwise comparisons that can reduce ‘noise’ inherent in measuring uncertainty, thus producing more accurate distributions based on individual judgments as well as judgments from groups of individuals. The process is based on pairwise comparisons of a range of uncertain outcomes, such as the frequencies of an event. 

Please welcome Professor Ernest Forman of George Washington University to the stage!