The Psychology of Quality and More

Decision Trees (part 2: decision strategies)

Quality Tools > Tools of the Trade > 23: Decision Trees part 2

Last time, we looked at the basics of Decision Trees whilst also indicating that there is a range of possible strategies that may be used to influence the final decision. This time we take a deeper look into those possible strategies.

How it works

When decisions are made, there is seldom one right course of action and your final choice will be driven by the criteria that you use to help make the decision. There are a set of strategies that can be used with Decision Trees to help select the best alternative that reflect common decision criteria.

The possible different strategies are:

·         Maximin, which looks for the best worst-case situation and is the 'pessimist's choice'.

·         Maximax, which seeks to maximize to profit by selecting the option with the highest potential return. It can thus be viewed as the 'optimist's choice'.

·         Minimax Regret or Opportunity Loss, which takes the conservative approach of seeking to minimize the loss incurred should the best action not be taken.

·         Expected Monetary Value, or EMV, includes the concept that an item which is less likely to occur is also worth less. This is a very common strategy, based on the sound statistical and financial principle that effective value is equal to actual value multiplied by probability of occurrence.

·         Expected Opportunity Loss, which combines the Minimax Regret with the statistical principles of EMV to give a conservative statistical approach.

·         Utility puts other personal values on events and actions. For example, actions may be rejected if there is any risk whatsoever of loss.

The table below illustrates how different strategies can lead to very different valuations and hence choices of actions.

How to do it

1. After building the basic Decision Tree, decide on the strategy which best suits your situation.

2. If necessary, identify the probabilities associated with the different events. This may be done via surveys, measurement or estimation (remembering that a poor estimate will give a poor result).

3. Using the strategy, calculate the value as appropriate of each action. This can be placed on the Decision Tree diagram, as below.

4. The selected decision may be highlighted by putting a double bar across the actions that are not selected.

Next time: Why-Why Diagram

This article first appeared in Quality World, the journal of the Institute for Quality Assurance