66R-11: Selecting Probability Distribution Functions for use in Cost and Schedule Risk Simulation Models
AACE International, August 24, 2012

Price: $0 (Member) / $100 (Non-Member)
1-CLICK DOWNLOAD FOR MEMBERS

This recommended practice (RP) of AACE International provides guidance for selecting probability distribution functions (PDFs) for use in probabilistic decision and risk management (DRM) simulation models within cost engineering and total cost management (TCM). These DRM models are used to analyze asset and project cost, schedule, profitability and similar measures in consideration of risks. While AACE's definition of risk is fully defined elsewhere, this RP uses the term risk to address both threats and opportunities.

This RP will describe the basic characteristics of commonly used PDFs and define their advantages and disadvantages for use in typical probabilistic modeling. It is intended to provide practical advice for non-statisticians who are applying risk analysis tools such as Monte Carlo modeling. In this area of practice, the data inputs are generally subjective; they are not based on sampling of data about a population or statistical analysis of empirical data. In that regard, this RP does not cover the topic of curve fitting or statistical analysis of empirical data such as linear regression. The goals of this RP are to help users find PDFs: a) with an inferred “goodness of fit” to the opinions and perceptions of participants in risk analysis exercises; and b) that are reasonably simple to apply, understand and communicate. It does not cover methods to elicit these opinions and perceptions.

[Look Inside]