RS58 - Intuition

Release date: April 8, 2012

Massimo's thinking capWhen your intuition tells you something, should you listen? That depends! Relying on intuition can be anything from a highly effective strategy used by experts, to an excuse not to require evidence for your beliefs. In this episode, Massimo and Julia talk about what people mean by "intuition," where our intuitions come from, and when intuition can beat careful reasoning.

Julia's pick: "Information is Beautiful - Snake Oil?"
Massimo's pick: "Zombie Economics: How Dead Ideas Still Walk among Us"

References:

http://measureofdoubt.com/2011/11/14/what-do-philosophers-think-about-intuition/

Comments

[Ed]

Hi,
I really enjoy your podcast. I was wondering if you could clarify something from this episode?

Julia discussed the failure to take into account the background rate when interpreting some statistics. So if a rare disease has a background rate of 1/1000 in the general population, and a test that has a false positive rate of 5%, and you test positive, it is still unlikely that you have the disease.

I remember learning about this in my statistics class all those years ago, but I cannot remember how to do the math! I was wondering if you could briefly explain it? Or point me in the right direction of where to find out about it? (A quick google search didn't help me, mostly because I don't know the right search terms to use). It's a really neat demonstration of how statistics aren't intuitive, I just wish I could remember how to arrive at the right answer!

Thanks!

[Ed]

Ok, I found what I was looking for above. If anyone else is looking for it, there is a good description at: http://www.cs.ubc.ca/~murphyk/Bayes/bayesrule.html

[Gil]

Hi Ed, you can find a faitly thourough explanation with examples on the wikipedia page here: http://en.wikipedia.org/wiki/Base_rate_fallacy

[Ed]

Thanks Gil!

[Jeff]

re: 10,000 hours and expertise

If you haven't heard about The Dan Plan you might find it interesting. A guy with no golf experience quit his job so that he could spend 10000 hours practicing golf to try and become a professional golfer. He's about 2500 hours in.

http://www.thedanplan.com/

[Max]

When you do enough statistics problems, it becomes intuitive. Take the base rate 1/1000, multiply by the true positive rate over the false positive rate, say 95/5, and you get the approximate probability after testing positive. To get the exact answer, use odds instead of probabilities, e.g. 1/999 instead of 1/1000.

[Gi]

There are some studies related to gut-instinct intuition vs rationale when donating money to causes. I only remember hearing about a couple, but one study compared donation rates to help protect panda habitat. Two approaches were used: presenting possible doners with simple graphs and factual information that explained the problem, or mainly showing cute pictures of pandas. The second approach got more people to donate, but they only gave small amounts of money. When people's critical thinking was engaged, more money was raised on average (people seemed to be donating in proportion to the problem). That was just one study but this podcast reminded me of it.