HKSM

Key Points

• Graphical model linking decisions, uncertainties, and outcomes to show how they influence each other.
• Used in planning and risk analysis to structure complex choices and dependencies.
• Nodes represent decision, chance, and value/utility elements; arrows show influence or information flow.
• Helps identify key drivers, required data, and where to apply sensitivity analysis.
• Often used to build decision trees, expected value calculations, or Monte Carlo simulations.
• Directed and acyclic; arrows indicate direction of influence, not its strength.

What the Diagram Shows

• The decision(s) under consideration and their timing.
• Uncertain variables that affect outcomes and their relationships.
• Objectives or value measures impacted (for example, cost, schedule, benefit).
• Information precedence: what is known before a decision is made.
• Dependencies among risks, controls, and results.

How to Construct

• Define the decision problem, alternatives, and success criteria.
• List key uncertainties, decisions, and outcome measures relevant to the choice.
• Classify each item as a decision node, chance node, or value/utility node.
• Draw nodes and connect them with arrows to indicate direct influence or information needed.
• Ensure the diagram is acyclic and time-ordered so information flows logically.
• Review with stakeholders to validate assumptions; refine as needed.
• Optionally add probabilities, utilities, and use the diagram to derive a decision tree or simulation model.

Inputs Needed

• Problem statement and decision alternatives.
• Project objectives, evaluation criteria, and value/utility preferences.
• Risk register entries, assumptions, constraints, and dependencies.
• Historical data or expert judgment for likelihoods and impacts.
• Stakeholder perspectives and risk attitudes.
• Timeline or process context to determine information availability.

Outputs Produced

• A visual map of how decisions, uncertainties, and outcomes relate.
• A prioritized list of influential variables and assumptions to test.
• Data requirements for quantification and sensitivity analysis.
• A structured basis for decision trees, expected value analysis, or Monte Carlo simulation.
• Updates to the risk analysis plan and risk register.

Interpretation Tips

• Read arrows from influencer to influenced variable; parents inform the child node.
• Arrows into a decision node indicate information available before making that decision.
• Absence of an arrow suggests conditional independence given the parent nodes.
• Do not infer magnitude or sign of effect from the diagram; quantify separately.
• Keep the diagram focused on the decision; omit unrelated process detail.
• Use sensitivity analysis to test which parent nodes most affect the value node.

Example

A team must choose between single sourcing and dual sourcing for a critical component.

• Decision node: sourcing strategy (single vs dual).
• Chance nodes: supplier reliability, price volatility, lead time variability.
• Value node: total project value (cost, schedule adherence, risk exposure).
• Influences: sourcing strategy affects exposure to supplier failure and admin overhead; reliability affects schedule; price volatility affects cost.
• Data to collect: failure probabilities, lead time distributions, cost premiums for dual sourcing.

Pitfalls

• Overloading the diagram with too many variables, reducing clarity.
• Introducing feedback loops; classic influence diagrams are acyclic.
• Confusing correlation with causation when adding arrows.
• Ignoring information timing for decisions and observations.
• Using inconsistent probability definitions or mixing scenarios improperly.
• Failing to validate the structure with subject matter experts and stakeholders.

PMP Example Question

During quantitative risk analysis, the team wants to show how the decision to outsource testing could change the probability of schedule delay and the project's net benefit, and what information is needed before making that choice. Which tool should the project manager use?

1. Fishbone (Ishikawa) diagram.
2. Influence diagram.
3. Control chart.
4. Stakeholder register.

Correct Answer: B — Influence diagram.

Explanation: Influence diagrams model relationships among decisions, uncertainties, and objectives and highlight information precedence, making them suitable for structuring decision and risk analysis.