Tip:
Highlight text to annotate it
X
So there we have it: directly from Danny Kahneman, talking about system one and system two. Now
I think it's important to remember the point of this episode. We're looking at the difference
between system one and system two. He makes that distinction clear, I think, when he was
talking about some of the early work that he did with Jackson Beatty, in the 60's
actually, on pupil dilation. He just mentioned it briefly, but what he was actually measuring—we
know that pupils when you are measuring—what you can actually do is film somebody's pupil,
and you can project it onto a wall beside them, for example, and you can measure it
literally with a ruler. Measure the size of the pupil as it gets larger and smaller.
Now we know that the pupil responds to things like light. So if you have bright light, the
pupil dilates [sic]. Dark, and then it gets larger. But it also responds to cognitive effort.
So it's actually responding to system two, essentially how hard your brain is working.
It's not even how hard you feel like you're working, but it's actually how much effort,
how much heat the brain is producing almost. So what you can do—the way that they tested
it, he and Beatty, was by presenting people with a digit-span task. If I ask you to remember,
for example, 6-4-3—that was three digits. I measure your pupil. Then I add a digit:
6-4-3-2. You keep those four digits in mind, and I measure your pupil. 6-4-3-2-7. We can
keep doing this and adding digits, and as you try to remember the digits span as you
keep adding more and more digits, your pupil just keeps getting larger and larger, until
one of two things happens: one, you report the number, so you say, "Okay, 6-4-3-2-7,"
and then your pupil constricts again; or you give up.
In his book, he talks about this really nice example where they're watching—they have
somebody participating in the experiment, in this digit-span task, and they're remembering
the digits, and they're outside of the room; they're watching this large pupil in the
screen outside. When the person gave up, they just see and they say inside of it, "So
you gave up on the problem?" The person was like, "How did you know?" It's almost
like you had a window into their own mind, which is, I think, quite cool.
That's cool. Danny Kahneman also talked about this phenomenon known as anchoring.
Fritz Strack, a researcher, has a pretty cool example of this. He asked people to guess
how old Gandhi was when he died. He put them into two groups. One group he asked, "Was
Gandhi older than 140 years old when he died?" Another group he asked, "Was Gandhi older
than 9 years old when he died?" These people responded—they were guessing how old Gandhi
was when he died—and the people that were given the high anchor of 140 guessed that
Gandhi died at 67 years old. People that were given the low anchor of 9 guessed that Gandhi
died when he was 50 years old. Now he actually died when he was 78, but you can see that
the high anchor pushed people in one direction, and the low anchor pushed them into another.
Now you may think that, "Well, that's kind of reasonable if an experimenter or someone,
they might have some inside information about the correct answer to the question, and so
it's quite reasonable to anchor your decision based on a number that they say," but it
can't be working like that because there's another great example by Dan Ariely. He
asked people to bid on bottles of wine and chocolate. Before the experiment, he asked
participants to write down their social security number, just the last two digits, and then
he split them into two groups. If their social security number is higher than 50, they go
to this group; lower than 50, in other group. Then he asked them, "How much would you
bid on these things, on these bottles of wine and these chocolate?" Now he found that
people who had a high social security number, greater than 50, for example, they were willing
to pay way more for these things than people who wrote down a low social security number
before the experiment. You can see that this arbitrary writing down
of a number is influencing their decisions. I think people with a high social security
number were willing to pay 60 to 120 percent more for these things than the low social
security number, which I think is pretty cool. This idea of anchoring, I think,
is really good. It highlights what we're talking about when—the point of this episode
is that we're operating under less-than-ideal conditions, and so we have to rely on shortcuts
in order to be able to navigate the world obviously. When you're in this anchoring
position, if you're given any number, any number, obviously, even random numbers, the
same thing happens with the roll of a roulette wheel. If a number that comes up, say 10,
then you use that as your anchor, regardless of whatever random process generated it—but
under most conditions it's not random. Under most conditions, when we're operating in
the world, a number appears, that's something to start with. But in complex scenarios, if
you have no idea, for example, of the percentage of African countries in the UN, or the population
of Australia, say, you have to start somewhere, and any number that you have is better than
no number at all. So we have to deal with what we have.
But that's just anchoring. We can deal with a whole bunch of other types of heuristics
and biases. We're going to talk about one next called availability, but what we're
going to do first is revisit the faces that we've presented in the first part of the episode.
Now what we want you to do is think back to
the list of faces we've presented earlier in the episode, so don't watch the video
again. Just think back to that list that we presented. What we want you to do is estimate
whether, in that list, there were more males or whether were more females that we presented.
Think back and go into the next section and indicate whether there were more males or
more females in that list.