Decision Making Decision method

Decision Tree

A decision tree represents choices, uncertain events, and consequences as branches and makes justified options and assumptions visible.

A decision tree represents choices, uncertain events, and consequences as branches. It clarifies options and assumptions; numeric probabilities belong only where they are justified.

Core question

Which actions and uncertainties lead to which outcomes, and what assumptions support the choice?

Plan the work Method guide

Start comparison

Give feedbackIs this method useful to you?

Participants: 1-6
Time: 30–90 minutes for a small tree; data collection and valuation are separate.

Which actions and uncertainties lead to which outcomes, and what assumptions support the choice?

Decision Tree method illustration showing its working structure

Question
Frame decision and outcomes
State decision, alternatives, consequences and criteria; distinguish a decision diagram from an ML model.
Nodes
Distinguish decision, chance and terminal nodes
Use decision nodes for controlled choices, chance nodes for uncertain events and terminal nodes for outcomes.
Evidence
Add evidence and probabilities only when justified
Add conditional probabilities at chance nodes only with evidence; branches at each node must sum to 1. Cite source and time basis for values.
Choice
Compare paths and document choice
Review consequences, sensitivity and qualitative criteria; record rationale, owner and revision signal.

Update the work record when new evidence, changed constraints or the planned review changes an assumption.

A good fit when

For multistage decisions with clear alternatives and uncertainty. Statistical decision-tree learning for classification or regression is a different technique.

Ideal for

Option evaluation
Policy design
Product or process logic

Not good for

Broad strategy questions without criteria
Creative ideation

Outputs

Decision question and criteria
The record contains the decision, alternatives, consequences, and criteria and distinguishes a decision diagram from an ML model.
Tree with clear node semantics
The setup uses decision nodes for controlled choices, chance nodes for uncertain events, and terminal nodes for outcomes.
Traceable branch assumptions
The branch record contains conditional probabilities at chance nodes only where evidence supports them; values cite source and time basis, and branches at each node sum to 1.
Reasoned choice and review date
The review record contains consequences, sensitivity, qualitative criteria, rationale, owner, and revision signal.

Decision Tree working record

Fictional example

Method-specific prompts and a fictional worked example; template prompts remain empty.

Frame decision and outcomes

Name the purpose and decision owner.

**Decision node:** Add a desk at a new rental branch? Options: add / do not add.

Distinguish decision, chance and terminal nodes

Label every branch and node type.

**Chance node after each option:** Peak demand high or low; these events are outside direct control.

Add evidence and probabilities only when justified

Leave unsupported probabilities unknown.

**Terminal nodes:** Describe wait time, operating cost and customer experience qualitatively. Do not add numbers without evidence.

Compare paths and document choice

Test whether plausible assumptions change the choice.

**Next step:** Collect peak-period data, then assess conditional probabilities per path and document their source.

OpenLearn explains decision-analysis branches and chance probabilities; IBM provides separate ML tree terminology. This guide distinguishes decision, chance and terminal nodes and uses no unsupported probabilities.

Decision trees and dealing with uncertainty | Open University OpenLearn

First step

Write the decision, options, external events, consequences and decision criteria.

Open the method guide

Good to do first

A hypothetical failure makes concrete risks, signals, and fitting mitigations visible.

Decision Making

Pre-Mortem

Supplies anticipated failures as possible external events in the tree.

Where to go next?

Decision Making

Trade-off Analysis

Also compares the chosen paths against accepted trade-offs.

Paper illustration of an assumption matrix with prioritized test cards

Decision Making

Assumption Surfacing

Makes the assumption behind the identified data gap explicit and plans a test step.

Deep dive

Purpose

For decisions with follow-on paths and dependencies, a linear comparison is often not enough. A Decision Tree shows how options branch under conditions and which consequences hang on each branch.

How it works

The decision question is spanned out along conditions, events, and possible states. This reveals which paths are expensive, unlikely, or strategically risky and where gaining information has the greatest value. This translates uncertainty into a testable path model.

In detail

A Decision Tree shows its strength when a decision depends on conditions, events, and follow-on options. Uncertainty becomes visible as a path structure and shows where additional information would really be valuable. The tree only stays useful if relevant branches are modeled and consequences are assessed honestly.

Facilitation

Define the starting question, the most important branches, and the decisive uncertainties before the tree grows too wide. It becomes helpful where the value of information shows which branch should be clarified first. Close with a preferred path, the triggers for reassessment, and the biggest known remaining uncertainties.

The method guide walks execution through 4 phases, timing, 4 pitfalls, and clear stop criteria.

Open method guide

Your perspective

Does this method help with your work?

Your feedback helps shape the next improvements to methodatlas.

Sources and topics

Topics

Decision Tree Options

Sources

Decision trees and dealing with uncertainty | Open University OpenLearnopen.eduOpenLearn explains decision-analysis branches and chance probabilities; IBM provides separate ML tree terminology. This guide distinguishes decision, chance and terminal nodes and uses no unsupported probabilities.What is a Decision Tree? | IBMibm.comOpenLearn explains decision-analysis branches and chance probabilities; IBM provides separate ML tree terminology. This guide distinguishes decision, chance and terminal nodes and uses no unsupported probabilities.

Artifact templates

Decision Tree: WorksheetPlan a session or initiative to match the actual scope of work.

markdown

decision-tree-working-template.md

Plan a session or initiative to match the actual scope of work.

Decision Tree: Worksheet

Question

To be confirmed

Desired outcome

To be confirmed

Scope

Prepare the full method Time: 30–90 minutes for a small tree; data collection and valuation are separate.

Complete this worksheet on paper or in your own document during the work.

Preparation for this scope

Method setup

Materials: Whiteboard, decision evidence, probability/value sources, criteria and decision owner.
Roles: Decision owner · subject experts on uncertainty · data specialist for quantitative evaluation.
Advance information: Choose materials, data access and participants to fit the research or decision question.
Overall time needed: 30–90 minutes for a small tree; data collection and valuation are separate.
Setup: Open the method-specific structure. Fictional examples are for orientation only; keep the blank worksheet separate from actual findings.

Prepare for the work steps

Question: Frame decision and outcomes

Write the decision, options, external events, consequences and decision criteria.
Whiteboard, decision evidence, probability/value sources, criteria and decision owner.