Sensitivity Dashboard
Sensitivity Dashboard
Section titled “Sensitivity Dashboard”This tool helps you identify which parameters matter most for your risk profile, enabling focused optimization efforts.
Why Sensitivity Analysis?
Section titled “Why Sensitivity Analysis?”Risk models have many parameters. Not all are equally important:
- Some parameters have wide uncertainty but low impact
- Others have narrow ranges but dominate the outcome
- Knowing which is which helps prioritize investigation and mitigation
| Parameter | Elasticity | Risk Range | Interpretation |
|---|---|---|---|
| Mitigation Effectiveness | -2.33 | $112 - $8 | High impact - prioritize |
| Outage Damage | +0.38 | $33 - $90 | Moderate impact |
| Outage Probability | +0.37 | $35 - $102 | Moderate impact |
| Error Probability | +0.31 | $36 - $78 | Moderate impact |
| Security Incident Damage | +0.31 | $36 - $93 | Moderate impact |
| Error Damage | +0.31 | $36 - $108 | Moderate impact |
| Security Incident Prob | +0.31 | $35 - $108 | Moderate impact |
- Mitigation Effectiveness: Increasing this from 0.30 to 0.95 could save ~$104/month
- Security Incident Prob: Reducing this from 0.05 to 0.00 could save ~$74/month
- Error Damage: Reducing this from 5000 to 200 could save ~$72/month
Understanding the Outputs
Section titled “Understanding the Outputs”Tornado Diagram
Section titled “Tornado Diagram”The tornado diagram shows one-at-a-time sensitivity:
- Each bar represents one parameter varying from min to max
- Center line is the base case risk
- Longer bars = more influential parameters
- Parameters sorted by influence (most important at top)
Elasticity
Section titled “Elasticity”Elasticity measures relative sensitivity:
Elasticity = (% change in risk) / (% change in parameter)| Elasticity | Interpretation |
|---|---|
| > 1.0 | Risk more than proportional to parameter |
| 0.5 - 1.0 | Moderate sensitivity |
| 0.2 - 0.5 | Low sensitivity |
| < 0.2 | Negligible impact |
Top Drivers
Section titled “Top Drivers”The top 3 parameters that most influence your total risk. Focus investigation and mitigation here first.
How to Use This Tool
Section titled “How to Use This Tool”1. Customize Parameters
Section titled “1. Customize Parameters”Adjust the base values and ranges to match your specific situation. The sensitivity analysis will recalculate automatically.
2. Identify High-Impact Areas
Section titled “2. Identify High-Impact Areas”Look for parameters with:
- Long bars in the tornado diagram
- High absolute elasticity values
- Wide possible ranges
3. Prioritize Actions
Section titled “3. Prioritize Actions”For high-sensitivity parameters with uncertainty:
- Invest in better estimates (data collection, expert elicitation)
- Consider robust decision-making approaches
For high-sensitivity parameters you can control:
- Prioritize mitigation investments here
- Calculate ROI of reducing the parameter
For low-sensitivity parameters:
- Don’t over-invest in precise estimates
- Simple approximations are often sufficient
Limitations
Section titled “Limitations”One-at-a-Time Analysis
Section titled “One-at-a-Time Analysis”This dashboard uses one-at-a-time (OAT) sensitivity analysis:
- Varies each parameter while holding others constant
- Doesn’t capture interaction effects between parameters
- May underestimate total uncertainty from correlated parameters
Linear Approximation
Section titled “Linear Approximation”Elasticities assume approximately linear relationships:
- Actual risk functions may be highly nonlinear
- Extreme parameter values may have disproportionate effects
Range Dependence
Section titled “Range Dependence”Results depend heavily on assumed parameter ranges:
- Wide ranges will show high importance
- Narrow ranges will appear less important
- Choose ranges that reflect actual uncertainty
Advanced: Monte Carlo Sensitivity
Section titled “Advanced: Monte Carlo Sensitivity”For more accurate sensitivity analysis, use the Risk Calculator with parameter variations:
- Run baseline simulation
- Vary one parameter’s distribution
- Compare output distributions
- Repeat for each parameter
This captures nonlinear effects and parameter correlations better than OAT analysis.