Choosing an Appropriate Distribution

Current @RISK users, both novices and experts, business and financial analysts, economists, statisticians, scientific researchers using @RISK or any other Monte Carlo simulation platform may greatly benefit by taking this course.  Students who also want to be introduced to @RISK as a general simulation methodology will benefit from this course.

It is intended to answer the common question modelers have whenever they are building a model: How to choose appropriate distributions for the variables, or “moving parts” of a Monte Carlo simulation model they are attempting to build. The principle of GIGO (“garbage in, garbage out”) applies here dramatically well. Build a model with appropriate distributions that clearly reflect the statistical nature of your variables and you will end up with a robust model to withstand reality testing. Build a model with lousily chosen distributions and your model will be as weak and questionable as any of your input variables.

This course starts by introducing a decision tree as a structure to help decide on multiple distributions. The world of statistical distribution functions is endless. @RISK uses some 97 different distribution functions to choose from. And this is not the end of it, since you can create, as we will show, your own distributions.