Decision Tree Analysis
A visual and quantitative approach for mapping a series of decisions and chance events, assigning probabilities, and calculating expected results to compare options when outcomes are uncertain.
Key Points
- Represents decision nodes (squares) and chance nodes (circles) with branches showing possible paths.
- Uses probabilities and outcomes to compute expected value (e.g., EMV) for each alternative.
- Supports selecting among options when results are uncertain, such as choosing risk responses or vendors.
- Assumes mutually exclusive paths and can include costs, benefits, and utility to reflect stakeholder preferences.
Example
A PM must choose how to deliver a module. Build in-house costs $300k with a 40% chance of a $120k delay penalty (EMV = 300k + 0.4 x 120k = $348k). Buy a COTS license for $260k and integrate: Vendor A (60%) integration cost $90k, Vendor B (40%) integration cost $200k, so expected integration = 0.6 x 90k + 0.4 x 200k = $134k; total expected cost = $260k + $134k = $394k. The decision tree favors building in-house ($348k) over buying ($394k).
PMP Example Question
A project manager needs to compare several mutually exclusive risk responses that involve sequential decisions and uncertain outcomes. She wants to assign probabilities and calculate expected monetary value to justify her choice. Which tool should she use?
- Monte Carlo simulation
- Decision tree analysis
- Pareto chart
- Sensitivity analysis
Correct Answer: B — Decision tree analysis
Explanation: Decision tree analysis maps decision and chance nodes and computes expected values for each path. Monte Carlo simulates many scenarios but does not depict branching choices; Pareto charts rank causes; sensitivity analysis shows which variables drive results.