Analysis Models

An analysis model accepts numeric values from a likelihood model and an impact model, and uses them to calculate a numeric risk score. How this score is calculated and used depends on whether the analysis model is qualitative or quantitative.

In either case, you can assign multiple likelihood and impact models to an analysis model. If you do, the person who performs an analysis selects one likelihood model and one impact model from those you've assigned to the analysis model.

Qualitative Analysis

A qualitative analysis model calculates its risk score by multiplying the numeric values supplied by the likelihood and impact models. However, it also maps ranges of scores to labels. As an overall risk level, it selects the label assigned to the range that includes the risk score it calculates.

Users who perform qualitative risk analysis work in words. They select labels that describe risk likelihood and impact. The likelihood and impact models translate these labels to numbers. The analysis model calculates its numeric score but returns the corresponding label. The numeric values are used only in the background.

For example:

An analyst may select Medium Low as a likelihood and Medium High as an impact for a risk. Suppose that the likelihood and impact models correlate these labels to the values 4 and 8.
The qualitative analysis model would multiply these values, producing the number 32.
That number would be included in a range (say, 21 to 40) that the qualitative analysis model correlates to a label (say, Medium Low). That label would describe the overall risk level.

To create or edit a qualitative analysis model is to:

Select the likelihood and impact models it uses. They must be the qualitative type.
Create labels that describe overall risk, such as Low, Medium, or High.
Correlate those labels to ranges of risk scores. The ranges you create must include all values from 1 to 100, with no gaps or overlaps. The higher the number, the greater the risk.
Be sure the score ranges allow for the values defined in the likelihood and impact models you've selected. Those models include values from 1 to 10, but don't have to include all of them. So you need to consider what values they do include and what products those values can generate. Here's an example:

Each of the likelihood and impact models has three labels, which correspond to the values 1, 3, and 5. So the analysis model can calculate only the following risk scores: 1, 3, 5, 9, 15, and 25. You might consider the first two of these to indicate low risk, the second two medium risk, and the last two high risk. But because the analysis model can't leave gaps between its ranges and must use all values up to 100, you might create these ranges: 1 through 3 corresponds to Low, 4 through 10 corresponds to Medium, and 11 through 100 corresponds to High. Or, you might map the range 11 through 25 to your High label and create a fourth range, 26 through 100. Knowing it will never be used, you assign it the label "NA."

Quantitative Analysis

A quantitative analysis model uses either of two formulas to calculate an overall risk score, and returns that score as an overall risk rating. The two formulas include:

A product formula:

Risk level = Likelihood × Time frame × Impact
A weighted formula:

Risk level = (Impact × Weight)^Factor1 × (Likelihood × Time frame)^Factor2

For either formula, a user sets the time frame while performing analysis on a risk. The user selects one in a set of labels, which correspond to numeric values as follows:

Day = 365.25
Week = 52.18
Month = 12
Quarter = 4
Year = 1
Decade = 0.1
Century = 0.01

To create or edit a quantitative analysis model is to:

Select the likelihood and impact models it uses. They must be the semiquantitative type.
Decide whether the model should use the product formula or weighted formula.
Supply the Weight value, and the Factor1 and Factor2 exponent values, if you select the weighted formula. The Weight value is a percentage. If you enter the value 1, for example, the formula uses the value 0.01.