15. Distributions: creating, editing and using

This section focuses on setting up distributions in your model. Distributions can be used in any of the following model types:

• Probabilistic Sensitivity Analysis (PSA) - parameter uncertainty

• • PSA runs the model multiple times, generating a set of expected values based on different sets of sampled inputs. Variance within the set of expected values reflects the impact of parameter uncertainty. In this context, distributions are used for parameter uncertainty.

  • Find more about PSA on Single Outcome Models in Probabilistic Sensitivity Analysis.

• Microsimulation models

  • Microsimulation sends multiple trial runs through the model via random walks. The analysis will generate outcomes for every trial run. The overall expected value of each strategy is calculated from the aggregated results from the full set of trial runs. In this context, distributions can be used to vary model inputs for each trial run.

  • Review the Microsimulation section for more details.

• Probabilistic Sensitivity Analysis with Microsimulation

  • Probabilistic Sensitivity Analysis and Microsimulation on Legal Models.

Derivations and detailed explanations of the distribution formulae provided here may be found on many math/statistics web sites, and in most texts on probability theory. See, for example, Christensen, Ronald; Data Distributions: A Statistical Handbook (2nd Ed.); Lincoln, Massachusetts: Entropy Limited, 1989.