IECM 12.0 beta User Manual > Introduction to Uncertainty Analysis >

A Non-Technical Example

To illustrate the process of defining a subjective probability distribution, let's turn to a simple example of eating lunch in a cafeteria. How long does it take from the time you enter the cafeteria to the time you pay the cashier? Assume that you enter at 12:05 p.m. on a weekday and that you purchase your entire meal at the cafeteria. The answer you give may depend on your recent experiences in the cafeteria. Think about the shortest possible time that it could take (suppose nobody else is getting lunch) or the longest possible time (everyone shows up at the same time). What is the probability that it will take 2 minutes or less? 45 minutes or less? Is the probability that it takes 10 minutes or less greater than 50 percent? etc. After asking yourself a number of questions such as these, it should be possible to draw a distribution for your judgment regarding the time require to obtain and purchase lunch at the cafeteria. Such a distribution might take the form of a fractile distribution giving the probabilities of different waiting times to purchase lunch. For example, your evaluation may conclude that there is only a 1 percent (1 in 100) chance it will take one minute or less, a 60 percent chance of 1 to 10 minutes, a 25 percent likelihood of 10 to 15 minutes, and a 14 percent chance of up to 25 minutes. These probability intervals can be drawn as a histogram and translated into a fractile distribution for a probabilistic analysis.