# Sensitivity analysis

Sensitivity analysis is one of the stages of decision making. Allows you to assess how would change the choice of an optimal decision, if you have changed the basic economic size or operating conditions. Sensitivity analysis is useful, because:

1. allows to highlight the main characteristics of the problem having an impact on the decision,
2. allows to assess the impact of the various factors, on management goals,
3. allows to reach optimal solutions in the case of a decision repeated in slightly modified conditions.

Sensitivity analysis can also include an evaluation of the implementation of the selected variant of decision. Using this technique, managers can be reassured that they made correct decision. And if not, what were the reasons? If the decision-making context has been correctly recognized? If there was set a proper goal? If they have considered all variants? If in the light of the ex-post evaluation managers would change their original decision?

## Sensitivity analysis methods

Sensitivity analysis is a method used to determine how changes in certain input variables will affect the output of a model or system. The following are common steps used in a sensitivity analysis:

1. Identify the input variables that are likely to have the greatest impact on the output. These are typically referred to as "critical" or "sensitive" variables.
2. Determine the range of possible values for each input variable. This can be done by reviewing historical data or by using expert judgment.
3. Select a method for varying the input variables. Common methods include perturbing one variable at a time (i.e., one-at-a-time or OAT) and using a multi-dimensional approach (i.e., Monte Carlo simulation).
4. Run the model or system for a set of input variable combinations and record the output.
5. Analyze the results to determine the relationship between the input variables and the output. This can be done graphically or mathematically.
6. Interpret the results in terms of the system or model's behavior and decision-making.
7. Repeat the analysis as needed with different assumptions and scenarios.

 Sensitivity analysis — recommended articles Decision table — Parametric analysis — Analysis of preferences — Level of complexity — Decision tree — Logistic regression model — Descriptive model — Risk estimation — Statistical hypothesis

## References

• William F. Samuelson, Stephen G. Marks; "Ekonomia menedżerska" Polskie Wydawnictwo Ekonomiczne 1998r.
• Saltelli, A., Chan, K., & Scott, E. M. (Eds.). (2000). Sensitivity analysis (Vol. 1). New York: Wiley.