kitchen table math, the sequel: Changing our Assumptions

Wednesday, February 20, 2008

Changing our Assumptions

In the earlier post , I commented that I consider people awfully good at being rational given their initial conditions--it's that their initial conditions can be so far from accurate that they have issues.

Catherine said in the comments in response
"I generally think that the nutty things people do usually make perfect sense from their perspective...Nevertheless, people do -- and then carry on doing -- a lot of things that bring them grief. Then they carry on being surprised that they are experiencing grief."
This habit of ours to repeat things that bring us grief, is a good example of bad priors. "My friend won't be late THIS time, she knows I can't afford to miss this plane flight" comes from unreasonable initial conditions: the prior probability that a friend will arrive on time is based on wishful thinking, or other friends, or what should be, but not adjusted to the evidence.

Why is it SO HARD for us to change our prior probabilities, even in the face of lots of overwhelming data? I don't know enough neuroscience to answer this question, but it's clear that by the time we're late adolescents or adults, our prior probabilities for our experiences in the world are pretty darn fixed.

We've got priors for the littlest details in our world. We've got priors for the next word in the sentence to be a verb, and priors for the next note in a song. This is why the sight-reading method and whole language methods can be so destructive: the students' brains are busy creating priors for "this word starting with "s" is going to be "sleep" ", rather than actually READING the word and finding out it's "silent".

Changing our priors is really what that "slowly pushing the wall" analogy is about. For kids, whether it's their behavior around chores or their learning to read, we're better off trying to be darn sure that the priors we're forming in their minds are true, rather than finding out later that they've become prejudiced against the symbols in front of them.


Catherine Johnson said...

Behaviorists always say that the core human approach to frustration is to keep doing what you're doing now but louder and more often.

Catherine Johnson said...

This also brings to mind one of my all-time favorite pieces of advice.

I've told this story before, so some of you guys should pass on by.

This was a graduate program in medical research at Columbia. There was a professor who was world-famous for research and for teaching; he was the mentor of half the major researchers working now, it seems.

He was teaching the brainiacs.

His main piece of advice to his students was: If what you're doing isn't working, try something else.

Catherine Johnson said...

Does "priors" capture what's going on, though?

I'm not sure.... (not sure how "priors" is defined in this setting).

Liz Ditz said...

See page 77 of the 2/25/08 New Yorker this link works for me

What Was I Thinking: The lastest reasoning about our irrational ways

Elizabeth Kolbert's review of Dan Ariely's Predictibly Irrational: The Hidden Forces That Shape Our Decisions

"In “Predictably Irrational: The Hidden Forces That Shape Our Decisions” (Harper; $25.95), Dan Ariely, a professor at M.I.T., offers a taxonomy of financial folly. His approach is empirical rather than historical or theoretical. In pursuit of his research, Ariely has served beer laced with vinegar, left plates full of dollar bills in dorm refrigerators, and asked undergraduates to fill out surveys while masturbating. He claims that his experiments, and others like them, reveal the underlying logic to our illogic. “Our irrational behaviors are neither random nor senseless—they are systematic,” he writes. “We all make the same types of mistakes over and over.” So attached are we to certain kinds of errors, he contends, that we are incapable even of recognizing them as errors. Offered FREE shipping, we take it, even when it costs us."

The review is an introduction to behavioral economics and Ariely's work.

I find it fascinating.