 |
Download a pdf copy of this help file here |
|
Download a pdf copy of this help file here |
See also: Modeling expert opinion introduction
List of subjective distributions
- Beta
distribution
- Bradford
distribution
- Ascending
Cumulative distribution
- Descending
Cumulative distribution
- Discrete
distribution
- Discrete
Uniform distribution
- GTU
distribution
- JohnsonB
distribution
- Kumaraswamy
distribution
- Kumaraswamy4
distribution
- Modified
PERT distribution
- PERT
distribution
- Reciprocal
distribution
- Relative
distribution
- StepUniform
distribution
- Triangle
distribution
- Uniform
distribution
The subjective distributions included in ModelRisk are particularly suited for estimating expert opinion. When we want to model a quantity we are uncertain about, a Subject Matter Expert (SME) is consulted to provide an estimate for it.
Rather than using a single point estimate, or minimum, most likely, maximum values, we model the uncertainty about a quantity with a distribution.
The PERT and Triangle distributions are most commonly used for expert estimation. These take absolute minimum, most likely and absolute maximum possible values as parameters.
These distributions have known drawbacks so it is better use the more flexible Modified PERT or GTU distributions.
See Distributions used in modeling expert opinion for a more thorough explanation of subjective distributions.
To combine the opinions of multiple experts, use the Combined Distribution window.
For an explanation about choosing, modifying and inserting a distribution with ModelRisk , please refer to the select Distribution window.