Plausibility and Probability

I must have read hundreds of books over the years, many of which have been great. Of those I’ve read, I can point to two books which fundamentally and almost instantaneously transformed my views on life, love and happiness. They are (in no particular order)

  1. Life after Death by Deepak Chopra
  2. The Selfish Gene by Richard Dawkins

Now, there’s a third one. “Thinking, Fast and Slow” by Daniel Kahneman. Prof. Kahneman teaches at Princeton these days. He’s a Nobel winner in Economics, and has done path breaking work over the last four decades in understanding the psychology behind how our minds process information. He’s truly a treasure.

I’d like to share some nuggets from the book over a few posts. If you like this sort of thing, you should get the book and check it out for yourself.

Background: Prof. Kahneman breaks the working of the mind into System 1, which processes data, reflexively forms patterns and draws conclusions, and System 2, which applies logical rules to examine the soundness of drawn conclusions. One of his assertions, borne out from his studies, is that System 1 is hyper-active, and System 2 is extraordinarily lazy.

Plausibility and Probability

One of the things we humans often do is to mix plausibility with probability. To understand this, Prof. Kahneman and his colleagues designed what is now famously called ‘the Linda problem’. Consider the description of a fictitious Linda below.

Linda is thirty one years old, single, outspoken and very bright. She majored in philosophy. As a student, she was deeply concerned with issues of discrimination and social justice, and also participated in anti-nuclear demonstrations.

After reading Linda’s profile, respondents in studies were asked a simple question: Which of the following is a more probable alternative?

  1. Linda is a bank teller.
  2. Linda is a bank teller and is active in the feminist movement.

The correct answer is the first. Think Venn diagrams. The circle with ‘feminist bank tellers’ is wholly included in the circle of all bank tellers. Therefore, the probability of Linda being just a bank teller is higher than the probability of her being a bank teller AND a feminist. The more specific you are about an event, the lower are the chances of its occurrence.

Here’s the fascinating part. In the survey, 89% of undergraduate students at top ranked American universities picked option 2. When they administered the test to doctoral students at Stanford Business School, all of whom had taken advanced courses in decision sciences and probability, they got a similar result! 85% chose option 2. A majority of really smart people violated the most fundamental rule of probability. They chose the more plausible but less probable event over the other.

What (the system 1 of) our mind does is to jump to the most plausible or coherent conclusion. It does not consider the likelihood of the conclusion. The mind substitutes likelihood with representative-ness of the event. This is how we make errors in judgment that can have far reaching impact on society at large.

This is not to say that plausibility is unimportant and should be ignored. It has its benefits. Plausibility, for example, is the basis for profiling at US airports. Resistance to these methods may be a laudable moral position, but also a rather simplistic one. If your name is Shah Rukh Khan and there’s a known terror suspect out there of a similar age with the same name, chances are pretty high that you’re going to be detained at the airport. Such predictive techniques are based on pattern matching, where the preference is to form a quick conclusion and investigate at leisure. As long as such methods are not manipulated but implemented fairly, they will lead to benefits for the society even if it means costs for the impacted few.

People who are taught new statistical facts about human behavior are often impressed to the point where they will tell their friends about what they heard. This does not mean that their world view has changed. Learning is about applying lessons to our own experiences, and not about repetition of facts. The test of learning lies in whether our understanding of experiences (and how we live) has changed .

In the words of the good professor himself, “changing one’s mind about human nature is hard work. Changing one’s mind for the worse about oneself is even harder.”

6 Replies to “Plausibility and Probability”

  1. I chose like a Stanford doctoral student too. The insights are unusual but as with most mind studies, not unpredictable. The tendency towards accepting plausibility seems to be evolutionary, a survival tool, and perhaps a key factor in the creative process. Much of the literature on handedness, for example, seems to point in a similar direction too.

  2. Yep. Applying it is not easy as some of these responses/reactions are near involuntary. I guess the post was a long-winded way of saying what we already know – jumping to conclusions is not the way to go.

  3. Yep. Applying it is not easy as some of these responses/reactions are near involuntary. I guess the post was a long-winded way of saying what we already know – jumping to conclusions is not the way to go.

  4. ha ha, I didn’t miss the bit about Stanford doctoral students 🙂 Yes, the tendency is described as evolutionary – quick assessments on fight/flight have been made/necessary over millions of years.

Comments are closed.