Fooled By Expert Intuition

Published on: Sat Mar 18 2023

Series

Content

Introduction

Have you ever been in a complex situation with an expert, and out of no where, they just come up with an answer or solution ?

How is that they are able to instantaneously come up with an answer or solution ?

Is it magic ? or is it something else ?

We’re going to get to the bottom of this!

In the book “Thinking Fast and Slow”, Dr. Daniel Kahneman breaks down the two models to help better understand “expert intuition”.

The two models are:

  • Dr. Gary Klein (High validity intuition)
  • Dr. Paul Meehl (Low validity intuition, statistical approach)

These two models contain the key to understanding expert intuition.

Expert intuition

Experts tend to have this “intuition”, almost like a sixth sense in their field of expertise.

Dr. Daniel Kahneman refers to this as “knowing without knowing how they know”.

This is how these experts are able to seemingly just come up with answers or solutions out of no where.

However, when asked how they’d knew, they just reply with “trust me, I just know” or “I had a hunch”.

So, the real question is if the experts didn’t know how they know it, do they really know it ?

That’s what we’re going to talk about in this article.

By digging into the details of these two models, we can better understand what is the decision making process like in the expert’s mind.

Expert intuition

Let’s start off with Dr. Gary Klein’s model which provides the ins and outs of how this process works.

Dr. Gary Klein’s model

In Dr. Gary Klein’s research, he has focused on studying people like firefighters, and some of those findings were later applied to military training to aid their decision making skills.

He has come up with a model for understanding this process, and it’s called the Recognition-Primed Descision (RPD) model.

The basic explanation of this model is that experts uses familiar cues in a complex and dynamic situation in order to quickly make a decision.

If you look at it from the perspective of firefighters and military operations, this is quite important in high pressure and time sensitive situations.

These decisions need to be made with high accuracy within seconds.

They are able to do this because they have patterns stored in their brain through decades of real and virtual experience that helps them identify options or solutions or answers in situations.

This decision making process is very intuitive because it relies on the expert’s pattern recognition to pick out key identifers or features from a situation.

Inner workings of Recognition-Primed Decision (RPD)

According to Dr. Daniel Kahneman, he argues that Recognition-Primed Decision (RPD) actually involves both the System 1 (Intuition) and System 2 (Analysis).

  • System 1 - Automatic (function of the associative memory)

  • System 2 - Analysis of the conclusion(s) to see if it makes sense (or if it will work)

Here is a more detailed breakdown.

In a given situation, something in the environment has provided the expert with a cue, and this gives the expert access to information stored in the memory and that information provides the answer.

While, System 2 may involve “slower thinking” or thinking deliberately, he argues that experts has honed the ability to quickly use both type of thinking in these specific situations.

Recognition-Primed Decision model (RPD) visualized
Recognition-Primed Decision model (RPD) visualized

The process:

  1. Cue - A pattern is recognized in a situation (System 1)

  2. Information (Memory) - The pattern is matched with a piece of information in our associative memory (System 1)

  3. Analysis - Analysis takes place to determine if it makes sense or it will work (System 2)

  4. Decision - A decision is made (Assuming it passes step 3)

Dr. Gary Klein’s model builds the foundation for how most people think about expert intuition and intuitive decision making.

Now the question becomes - does this model always work ?

That’s where Dr. Paul Meehl’s model, and his research findings comes in.

Let’s take a look at that and understand why this model doesn’t always work.

⚠️ Note:

In steps 1 and 2, the expert does not know this process is happening.

System 1 (intuitive) happens automatically and it just feeds the person a solution.

Dr. Paul Meehl’s model

Dr. Paul Meehl argues that some experts like clinicians, stock pickers, political scientists and pundits operate in low (or zero) validity environments.

In his research, he talks about how these experts often rely on unreliable cues (in the RPD process) which leads to decisions that may not be so accurate.

The solution that he has proposed to combat this is to defer to using statistical algorithms or methods for decision making in those situations.

Meehl’s statistical method for decision making visualized
Dr. Paul Meehl’s statistical method for decision making (visualized)

Inner workings of low validity decision making

So, why is it that some experts can’t seem to find accurate answers in certain situations ? Is it due to their skills ?

It’s not really about the skill (or experience), rather its about the situation or environment that the expert is in.

In unpredictable or noisy situations or environments, a valid cue cannot be found, and the expert intuition actually produces an inaccurate result.

Dr. Daniel Kahneman breaks down this process, and like the RPD model, it uses both System 1 (Intuition) and System 2 (Analysis)

Here is where it gets interesting, let’s go through the process.

Decision making process (in noisy environments):

  • System 1 (Intuition) - System 1 will produce an answer to a difficult question via substitution

  • System 2 (Analysis) - The answer produced is an answer that was not for the question intended but it may good enough to pass the review of system 2

Again, this process in System 1 (intuition) happens automatically, so it’s not easy for these experts to “trace their steps”.

What this means is that an expert can have judgements that answer the wrong question with high confidence (subjective confidence).

While these decisions may appear valid, Dr. Daniel Kahneman emphasizes that these conclusions are “self delusional at best, and sometimes worse”.

Again, this is due to the process of question substituion and the unpredicatable environments that some experts may find themselves in.

This is what led to Dr. Paul Meehl’s conclusion of aiding the decision making process with data (ie statistical algorithms or methods) in these noisy situations or environments.

Intuitive decision via question substitution
Intuitive decision via question substitution

Benefits of using Statistical Methods (From Dr. Daniel Kahneman):

  • More likely to detect weakly valid cues (Than human judgement - Intuitive process)

  • Likely to maintain a modest level of accuracy by using the cues consistently

Testing validity

You may be thinking... couldn’t we say this for “experts” like athletes, firefighters and surgeons ?

This comes back to this concept of validity, or more importantly, whats makes a cue more valid in some situations compared to others ?

According to Dr. Daniel Kahneman these are the conditions needed to acquire expert intuition:

  1. Predictable environment - The environment has to be sufficiently regular to be predictable

  2. Prolong learning - An opportunity to learn these regularities through prolong practices (ie over time)

Expert intuition with regularity in the environment or situation (visualized)
Expert intuition with regularity in the environment or situation (visualized)

So, these two conditions can be used to evaluate whether or not an expert will have a valid cue in any situation.

When these two conditions are met, the expert’s intuition is likely accurate.

So, I believe the main difference between validity is the regularities in the environment.

While athletes, firefighters and surgeons may face different cases in their field, much of the environment tends to remain quite regular.

This allows for highly valid cues in these situations which leads to accurate decision making (from an expert practitioner).

Conclusion

So, there you have it, quite a bit of information about the two models.

These models form the basis from which we can understand expert intuition, and how to reliably evaluate it’s accuracy.

Let’s do a quick review!

The Two Models for expert intuition:

  1. Gary Klein’s Model - Dr Gary Klein’s model suggest some experts operate in regular environment, and they can confidently rely on their intuition for decision making (especially in high pressure and time sensitive situations)

    • Recommendation: This produces high confidence in trusting the expert’s decisions
  2. Paul Meehl’s Model - Dr Paul Meehl’s model suggest some experts operate in noisy environment and in these situations, it’s better to rely on statistical methods for decision making

    • Recommendation: This produces low confidence in trusting the expert’s decisions, defer to or supplement with data (ie statistical algorithms or methods)
Two models for Expert intuition - Gary Klein and Paul Meehl’s Model
Two models for Expert intuition - Gary Klein and Paul Meehl’s Model

As a reminder, validity is not based on skill level or experience, rather, it is based on regularity of the environment and prolonged learning by the person.

Now you should have a good unerstanding of the inner workings of both models and what expert intuition is about!

All of this has a lot of implications in terms of decision making, learning and boosting your general understanding of the expert decision making process.

Hopefully this article has helped you to dispel some the magic that goes on during an expert’s decision making process.

That’s all for now...

If you enjoyed that or learned something new, please share this article with a friend or co-worker 🙏❤️! (Thanks!)

Also, feel free to connect with me on social media, and let me know which part you found most interesting!

Connect with (or Follow) me on LinkedIn - https://www.linkedin.com/in/ijerrychang 🤝


Enjoy the content ?

Then consider signing up to get notified when new content arrives!

Jerry Chang 2023. All rights reserved.