An Introduction to Decision Theory

I recently had a conversation with a student that yielded an interesting case that could be used as an introduction for people to decision theory. With that in mind we made a video and this article to do just that.

What is decision theory?

Decision theory is basically a tool people can use to make better decisions. It usually is not used in short-term time sensitive decisions (although it can be).

World leaders, Chief Executive Officers, and leaders that make big decisions in general will use it regularly. For the layperson there are a few advantages to learning decision theory:

1. Making big decisions with long-term consequences like ‘should I quit my job’, ‘should I buy that house’, or ‘should we start a family at (insert time)’. 

2. Understanding the way one’s bosses – bosses – boss thinks. This can lead to potential career improvements (promotions, etc.) when one can better anticipate what is useful for people higher up the organizational hierarchy.

3. Being better informed on public policy. By viewing political decisions through a decision theory ‘lens’ it can be easier to see precisely how their decisions are better or worse. It can also be easier to prove precisely why a decision may be wrong or right, ethical or unethical, etc.

A basic overview of how it works:

The idea is to map out all the points into the future one can predict in what is effectively a modified flow chart.

Decision theory has three ‘nodes’ that have associated shapes:

• Decision nodes (Squares)

• Chance nodes (Circles)

• End nodes (Triangles)

Any time something is within the person’s control and they are deciding, one makes a Square.

Any time something is outside of someone control and the ‘decision’ is effectively being made by an outside force, one makes a circle.

When a path has effectively ended in a result or has gone as far as one can predict it is capped off with an end node. As an example, the decision tree that was made for the case we will be using is this:

Context for the case:

The student asked a friend to take a university exam for them, was caught, confessed to their professor that they did so, and was basically asking for advice on this question:

“My professor has not gotten back to me yet and I do not know if consequences are coming, should I follow up and ask them what the progress of this is?”

Further context was provided:

The student had previously lost 10+ family members in an active warzone a few weeks prior to this exam (naming the specific one is not relevant). This resulted in them requiring significant mental health intervention to address the trauma.

In their vulnerable state they self-admittedly made a bad decision in a panic.

Now depending on one’s personal bias, that context could be viewed a variety of ways. Two of these are for example:

One could be highly sympathetic to the student and feel that this decision was understandable.

Another could assume that the student is ‘making it up’ or is utilizing the passing of their family to ‘get away with’ doing a bad thing.

In either case, the personal judgement is actually irrelevant. This is actually a significant point when it comes to utilizing decision theory and also in advising clients in risk management.

From the individual’s perspective this context is the truth. It is generally not the professional’s responsibility to adjudicate the matter, but to operate off of the relevant information to answer the question.

With this in mind, here is the logic in how we can create the decision tree (the flowchart).

Making the decision tree:

The main question asks what the optimal decision is based on outcomes. The individual, client, or in this case the student decides based on the endpoints what the optimal outcomes are.

The first point in this tree is actually not a decision. Instead it is the question: has a disciplinary process already been initiated?

If so, contacting the teacher again is largely irrelevant and the following question becomes what action will the student take when contacted for disciplinary action.

This disciplinary action generally involves university staff follow-up with the student and ask for an explanation, come to a determination, and then inform the student of the consequences.

This is represented in the upper path. The student identified three options in this case: say nothing, tell a lie, tell the truth (family passing trauma context).

On the lower path, the assumption is that the teacher/professor did not initiate the disciplinary process (they forgot, did not consider it a priority, etc.).

If the student does nothing, then the problem essentially ‘goes away’. Conversely, if they ask the teacher about the progress of this, they essentially loop back to the start of the decision tree with the initial chance node.

After the decision tree has been made, look at each outcome (not listed above), and work backwards figuring out the optimal decisions to reach those points.

This will often result in a set of decisions where one can ‘solve’ at every decision point what they will do to regardless of what chance-nodes force upon them creating optimal ‘paths’. 

While it is not always possible to perfectly solve a decision tree, for this student, they actually had an optimal solved result which is below:

The student decided that the best outcome was for the problem to go away on its own, and that in the event they had to deal with disciplinary action, that they would provide the truth of the circumstances that led them to that decision.

They were confident that these were the preferred outcomes.

Going back to the original question: “My professor has not gotten back to me yet and I do not know if consequences are coming, should I follow up and ask them what the progress of this is?”

The answer would be: no.

To some, this outcome may seem obvious, and they may think that using decision theory for this is going overboard.

In the case of this student, prior to mapping out their situation and visualizing it they were highly distressed. After doing so, they realized that they had already made up their mind on what to do, but just needed confidence.

The main valuable takeaway was as follows: You already know what you want to do in this situation, stop worrying about it and get on with your day.

Learn more and watch the video here.

Vezina is the CEO of Prepared Canada Corp. and is the author of Continuity 101. He can be reached at info@prepared.ca.