Decision Tree Analysis in Project Management

Decision trees are commonly used in business, engineering, and computer science to analyze decisions and their potential outcomes as a process of elimination. 

A decision tree visually displays decisions and their potential outcomes, consequences, and costs, making it easier to take action with confidence. Not only can this method help individuals make better choices in their daily lives, but businesses can also use it to optimize their decisions about inventory, service models, and other aspects of the supply chain. 

What is a Decision Tree?

A decision tree, also known as a decision diagram or decision table, is a chart that represents an algorithmically-organized list of decisions and possible consequences. It is widely used in project management, computer modeling, planning, algorithmic design, and program logic. A typical decision tree consists of: 

  • Decisions at its interior nodes; 
  • Branches that lead to various scenarios associated with each decision; 
  • Probabilities or payoff tables at each of these scenarios’ nodes. 
Decision trees are generally used in the planning performance domain

These trees can be used for both qualitative and quantitative analysis.  When reading a decision tree, follow these steps: 

  • Read every question starting from left to right until you reach a terminal node (the question ends). 
  • If there is no yes branch ending in that terminal node, stop here and go back up to where you came down on your way down. 
  • If there is a yes branch ending in that terminal node, follow it and make whatever choice leads to that branch's conclusion before going back up again. 
  • Repeat step 3 until you reach a leaf (end) node of your decision tree.

Expected monetary value (EMV) within the decision tree

Within decision analysis, EMV (Expected Monetary Value) allows for a side-by-side comparison of different alternatives. So, for example, your company can ask what's better for our company will you be investing in new equipment or will you be hiring another employee? To answer that question, you need to calculate the expected monetary value (EMV) for each option using your decision tree: The first step is figuring out what you stand to gain from either option. For instance, let’s say that you have a projected profit margin of $5 with your current employee; if instead you hire one more employee, your expected profit margin will be $7. Next, take into account any potential losses that might occur from either option. In our example, we’ll assume there are no losses associated with either choice. Finally, multiply both potential gains and losses by their respective probabilities to arrive at each option's expected monetary value (EMV). Now you know which choice has higher EMV! Keep in mind that some options may not have a positive EMV—if so, those should be eliminated right away. Remember: all other things being equal, choose whichever option has higher EMV.

Decision Tree Analysis Steps

  • Isolate decision alternatives. Decisions must be made in isolation so that the analysis can account for only the decision being analyzed,
  • Depict the decisions in a tree format. This means that each different course of action is shown as a node, or box, on the tree. Each node has three elements: The initial event or situation that triggers your choice,
  • Take into account the odds of different scenarios and give them respective payoffs to figure out the average monetary value for each possible scenario,
  • Evaluate scenarios using criteria,
  • Select the best option,
  • Document results and next steps

In which Sectors Decision Tree Analysis is used

As mentioned, decision trees are often used in fields such as business and economics. A few examples of real-world applications include: 

Healthcare: Used in medical settings, decision trees can help predict how patients might respond to different treatments, with the goal of identifying which courses of action would most benefit them. Computer Science: As a programming tool, decision trees can help programmers construct databases and software programs that have clear paths for execution. 

Engineering/Construction: Trees can be used to plan out all aspects of construction or engineering projects—everything from budget planning to timeline creation. 

Law: For lawyers and judges who need to determine appropriate courses of action for given scenarios in order for cases to be resolved correctly, decision trees can be extremely helpful tools.

See also: Alternatives Analysis

Trend Analysis

Regression Analysis

Root Cause Analysis

Reserve Analysis

Multicriteria Decision Analysis