Second order cumulative probability plot | Vose Software

Second order cumulative probability plot

See also: Cumulative probability plots, Second order cumulative probability plot, Presenting results introduction, Graphical descriptions of model outputs

A second order cdf is the best presentation of an output probability distribution when you run a second order Monte Carlo simulation. The second order cdf is composed of many lines, each of which represents a distribution of possible variability or probability generated by picking a single value from each uncertainty distribution in the model.

This is a second-order plot of a discrete random variable. The step nature of the plot makes it difficult to read.

This is another second-order plot of a discrete variable, where the probabilities are marked with small points and joined by straight lines. The connection between the probability estimates is now clear, and the uncertainty and randomness components can now be compared: at its widest the uncertainty contributes a spread of about 2 units (red dashed line), whilst the randomness ranges over some 8 units (blue dashed line), so the inability to predict this variable is more driven by its randomness than by our uncertainty in the model parameters.

This is a second-order plot of a continuous variable where our inability to predict its value is equally driven by uncertainty (red dashed line) about the model parameters as by the randomness of the system (blue dashed line). This is a useful plot to provide decision-makers because it tells them potentially how much more sure one would be of the predicted value if more information could be collected, and thus the uncertainty reduced.

Read on: Overlaying cdf plots