I have been interested in trying to think better, so I was looking for information on cognitive biases. One of the most recommended books in this area is Thinking, Fast and Slow by Daniel Kahneman. So I got it and tried to read it.

The Basic Idea

The basic idea of the book is that cognitive biases (defined as a systematic, predictable errors) are not emotional issues, but rather arise from the way our minds work. Kahneman’s idea is that there are two types of thinking. There is fast thinking that is based on associations we make between things. It’s very fast and it’s easy to do. Then there is slow thinking that is much more deliberate, but it takes more effort. The slow thinking part of our mind is supposed to keep an eye on the fast thinking and step in when mistakes are likely, but that doesn’t always work.

The fast thinking part of our brain is really bad at math, statistics, and logic, and is the source of a lot of our cognitive biases. The associative mind aims for consistency, which gives us the halo effect. That is where you give less credence to bad information about people you had good first impressions of, and give less credence to good information about people you had bad first impressions of. It leads to anchoring, where what you say right before you ask a question has a large effect on how the question is answered (this is used a lot to bias surveys). The way fast thinking is built on associations leads to availability bias, where what you’ve heard more about seems more likely. There are several other examples given in the book.

The slow thinking part of our brain can deal with a lot of those issues, but we can’t use it all day long. We’d get exhausted from the effort it takes. Conversely, when you are exhausted it is a bad time to make decisions, because it’s harder to get the slow thinking part of our brain in gear.

Sloppiness?

Kahneman makes a good case for this fast brain/slow brain idea, and talks about a lot of studies done by him and others that support his ideas. But after my experience with Laurie Santos’ Yale Happiness Class, I am rather suspicious of psychological studies in general. And Kahneman wasn’t even as good as Santos at providing clear information on the studies in question. A lot of my problems with the Santos class were because I was seeing things in the graphs and other slides she was presenting that tripped red flags for me. Kahneman would say who did the study and the year, but he didn’t give hard information on the study results. So there were no red flags, but no sense of certainty either.

Then he started talking about his research methodology. He would come up with an idea, and then talk it through with his colleague Amos Tversky. They would come up with an answer to the question, and then do a study to confirm the answer. That seems like a really shady way to do research to me. I would think research should be about finding out what the answer is, rather than confirming the answer you think is right.

But I kept reading the book, even though it was becoming a slog. Later on in the book he gets into how this all feeds into his Prospect Theory about economics. Mind you, I think bringing more about cognitive biases and irrational behavior into economics is probably a good idea, but it wasn’t what I bought the book for.

Then I got to the phrase “I did not collect data, because the outcome was obvious.” While I agreed with him about the outcome, I wouldn’t just assume it was true because it seemed obvious to me. I would collect data. At that point the slog of reading the economics caught up with me, and I stopped reading the book.

Conclusions

I came out of this thinking that he had some good ideas about how people think, that matched with my experience. But I was still a little skeptical, for two reasons. One, I couldn’t confirm that the study’s he was quoting really supported his arguments, and least not without a lot of effort tacking down and reviewing the original studies. Two, I felt exactly as confident about these ideas as I did about The Society of Mind by Marvin Minsky. Yet it’s not quite clear to me how to reconcile those two views of the mind (not that I have put a lot of effort into doing so).

Then I started writing up this post about the book, and went to Wikipedia to double check some of the statements I was making. There is a section on Wikipedia named “Replication Crisis” which caught my eye. The first sentence says “The book has been heavily criticized for relying on shoddy and non-reproducible studies.” Well, so much for that.

This project made me reconsider my plans for the future, which included reading several more books in this area. My goal in my retirement has been to enjoy my retirement. After watching my father trapped in three rooms for years I didn’t want to repeat his mistakes. And while I don’t want some sybaritic, self-reinforcing life, I do want to enjoy my retirement. I did not enjoy reading this book. And I want to be more careful about taking my enjoyment into account when choosing projects to work on in the future.