Distribution Selection with No Data Using VBA and Excel
Trevor A. Craney, Nathan White · Quality Engineering · 2004
Often in statistical analyses or Monte Carlo simulations, analysts are required to assume distributions for variables with no data. Without a method or tool to help select from potentially competing distributions, analysts will often default to common or well-known distributions, thus limiting them from accurately modeling their beliefs. An algorithm and flowchart are provided for both a simple format and a complex format to aid in the selection process from an initial list containing numerous distributions. A software program supporting this methodology has been developed and is illustrated. This program is especially useful when selecting parameters for a chosen distribution.