1
00:00:00,090 --> 00:00:02,490
The following content is
provided under a Creative

2
00:00:02,490 --> 00:00:04,030
Commons license.

3
00:00:04,030 --> 00:00:06,330
Your support will help
MIT OpenCourseWare

4
00:00:06,330 --> 00:00:10,720
continue to offer high quality
educational resources for free.

5
00:00:10,720 --> 00:00:13,320
To make a donation or
view additional materials

6
00:00:13,320 --> 00:00:17,280
from hundreds of MIT courses,
visit MIT OpenCourseWare

7
00:00:17,280 --> 00:00:20,130
at ocw.mit.edu.

8
00:00:20,130 --> 00:00:23,070
ABBY NOYCE: First, a
little bit about how

9
00:00:23,070 --> 00:00:25,650
we make decision making.

10
00:00:25,650 --> 00:00:29,310
So, one model of how
we make decisions

11
00:00:29,310 --> 00:00:32,369
is what's called the
expected utility model.

12
00:00:32,369 --> 00:00:34,230
And the expected
utility model basically

13
00:00:34,230 --> 00:00:36,960
says that you look at--

14
00:00:36,960 --> 00:00:39,420
if you have two
possibilities, you

15
00:00:39,420 --> 00:00:43,620
look at what you think the
likely outcomes of each

16
00:00:43,620 --> 00:00:47,010
of these possibilities are, how
probable those outcomes are,

17
00:00:47,010 --> 00:00:50,195
and how you value
those outcomes.

18
00:00:50,195 --> 00:00:53,430
So, if you think that
if I make choice A,

19
00:00:53,430 --> 00:00:57,210
then I'm very likely to have a
good thing happen but there's

20
00:00:57,210 --> 00:01:00,410
a small possibility of having
something really bad happen,

21
00:01:00,410 --> 00:01:02,850
or choice B is kind
of neutral, then

22
00:01:02,850 --> 00:01:05,820
depending on just how you
value these different outcomes,

23
00:01:05,820 --> 00:01:07,538
you're going to make
different choices.

24
00:01:07,538 --> 00:01:09,680
AUDIENCE: Like expected value?

25
00:01:09,680 --> 00:01:12,820
ABBY NOYCE: It's all about--
yeah-- expected value.

26
00:01:12,820 --> 00:01:14,850
So, you can think
of this as like--

27
00:01:17,730 --> 00:01:19,980
if you're trying to make
decisions like what to do--

28
00:01:19,980 --> 00:01:21,540
It's Friday night.

29
00:01:21,540 --> 00:01:23,601
You've got an exam on Monday.

30
00:01:23,601 --> 00:01:24,975
What are the things
you could do?

31
00:01:24,975 --> 00:01:27,495
So, you've got a decision.

32
00:01:30,030 --> 00:01:33,590
So, plan A is stay
home and study, right?

33
00:01:33,590 --> 00:01:36,630
You could stay home.

34
00:01:36,630 --> 00:01:41,220
You could be a diligent student.

35
00:01:41,220 --> 00:01:44,970
And let's suppose that if
you stay home and study,

36
00:01:44,970 --> 00:01:47,610
then it's almost certain that
you'll do well on the exam.

37
00:01:47,610 --> 00:01:50,042
AUDIENCE: There's a 1% chance
you'll waste your whole time

38
00:01:50,042 --> 00:01:51,630
playing computer games.

39
00:01:51,630 --> 00:01:53,921
ABBY NOYCE: Right, let's keep
this a little bit simpler

40
00:01:53,921 --> 00:01:54,810
for a minute.

41
00:01:54,810 --> 00:01:57,250
Your other option is
to do something else.

42
00:01:57,250 --> 00:01:58,712
What might you do
on a Friday night

43
00:01:58,712 --> 00:02:00,420
if you don't want to
stay home and study?

44
00:02:00,420 --> 00:02:01,477
AUDIENCE: Go to a party.

45
00:02:01,477 --> 00:02:02,560
ABBY NOYCE: Go to a party.

46
00:02:09,660 --> 00:02:11,930
Go to a party.

47
00:02:11,930 --> 00:02:14,611
What are some possible
outcomes of going to a party?

48
00:02:14,611 --> 00:02:17,110
Good things that might happen,
bad things that might happen.

49
00:02:17,110 --> 00:02:17,970
AUDIENCE: You meet new people.

50
00:02:17,970 --> 00:02:18,720
AUDIENCE: You might have fun.

51
00:02:18,720 --> 00:02:20,011
ABBY NOYCE: You might have fun.

52
00:02:26,790 --> 00:02:30,570
Or it might be a party where you
don't know anybody and nobody

53
00:02:30,570 --> 00:02:32,200
seems to want to talk to you.

54
00:02:32,200 --> 00:02:34,930
It might be a party
where you have no fun.

55
00:02:34,930 --> 00:02:37,397
So, what do you think--

56
00:02:37,397 --> 00:02:39,480
give me a relative likelihood
of these two things.

57
00:02:39,480 --> 00:02:41,010
Are these equally
likely outcomes?

58
00:02:41,010 --> 00:02:42,858
Is one more likely
than the other?

59
00:02:42,858 --> 00:02:44,610
AUDIENCE: Well, it
depends whose party.

60
00:02:44,610 --> 00:02:45,570
ABBY NOYCE: Depends whose party.

61
00:02:45,570 --> 00:02:46,340
Think of a party.

62
00:02:46,340 --> 00:02:49,830
Pick one at random, some
kid you know from school

63
00:02:49,830 --> 00:02:51,220
but you don't know really well.

64
00:02:51,220 --> 00:02:53,908
But they're having a party
and you heard about it.

65
00:02:53,908 --> 00:02:55,390
AUDIENCE: Were you invited?

66
00:02:55,390 --> 00:02:56,920
ABBY NOYCE: Were you invited?

67
00:02:56,920 --> 00:02:57,848
I don't know.

68
00:02:57,848 --> 00:03:01,440
AUDIENCE: We just go
to the party anyway.

69
00:03:01,440 --> 00:03:02,940
ABBY NOYCE: I've done that.

70
00:03:02,940 --> 00:03:03,720
So, all right.

71
00:03:03,720 --> 00:03:06,990
Let's say that
there's a 75% chance

72
00:03:06,990 --> 00:03:09,570
that you'll have
fun and a 25% chance

73
00:03:09,570 --> 00:03:10,740
that you'll be miserable.

74
00:03:10,740 --> 00:03:12,930
And we'll say that this is
a certainty, that there's

75
00:03:12,930 --> 00:03:15,030
a 100% chance that you'll
do well on your exam

76
00:03:15,030 --> 00:03:18,210
if you stay home and study.

77
00:03:18,210 --> 00:03:21,570
How good or bad
are these outcomes

78
00:03:21,570 --> 00:03:25,950
on a scale from like minus 10 to
10, where 0 is kind of neutral

79
00:03:25,950 --> 00:03:28,500
and minus 10 is the worst thing
that could possibly happen

80
00:03:28,500 --> 00:03:30,600
to you and 10 is amazing?

81
00:03:30,600 --> 00:03:34,320
Where is having fun at a
party on this value scale?

82
00:03:34,320 --> 00:03:35,430
Is it a 2?

83
00:03:35,430 --> 00:03:37,110
Is it an 8?

84
00:03:37,110 --> 00:03:38,590
Is it a minus 3?

85
00:03:38,590 --> 00:03:39,090
AUDIENCE: 6

86
00:03:39,090 --> 00:03:39,690
ABBY NOYCE: Probably not.

87
00:03:39,690 --> 00:03:40,320
A 6?

88
00:03:40,320 --> 00:03:41,600
Sure.

89
00:03:41,600 --> 00:03:44,360
AUDIENCE: 7.5.

90
00:03:44,360 --> 00:03:46,970
ABBY NOYCE: How bad is it to
go to a party and be miserable?

91
00:03:46,970 --> 00:03:48,230
AUDIENCE: 0.

92
00:03:48,230 --> 00:03:49,740
ABBY NOYCE: It's just neutral?

93
00:03:49,740 --> 00:03:50,400
It's not bad?

94
00:03:50,400 --> 00:03:52,470
AUDIENCE: That's a negative.

95
00:03:52,470 --> 00:03:53,560
AUDIENCE: Minus 5.

96
00:03:53,560 --> 00:03:54,070
ABBY NOYCE: Give me a negative.

97
00:03:54,070 --> 00:03:55,200
Someone give me a
negative number.

98
00:03:55,200 --> 00:03:55,700
Minus--

99
00:03:55,700 --> 00:03:56,750
AUDIENCE: Negative 7.

100
00:03:56,750 --> 00:03:59,450
ABBY NOYCE: Negative 7?

101
00:03:59,450 --> 00:04:00,920
All right.

102
00:04:00,920 --> 00:04:02,869
How much of a good--

103
00:04:02,869 --> 00:04:04,910
how important is it to
you to do well on an exam?

104
00:04:04,910 --> 00:04:06,451
AUDIENCE: 20.

105
00:04:06,451 --> 00:04:07,950
ABBY NOYCE: Scale
of minus 10 to 10.

106
00:04:07,950 --> 00:04:08,780
AUDIENCE: 10.

107
00:04:08,780 --> 00:04:10,720
ABBY NOYCE: It's a 10?

108
00:04:10,720 --> 00:04:13,834
This is super important?

109
00:04:13,834 --> 00:04:16,250
So, we probably want something
that makes it more valuable

110
00:04:16,250 --> 00:04:18,670
than having fun
at a party, maybe.

111
00:04:18,670 --> 00:04:19,630
Let's call it an 8.

112
00:04:23,470 --> 00:04:24,800
OK.

113
00:04:24,800 --> 00:04:30,590
So, expected utility model
would look at this and say,

114
00:04:30,590 --> 00:04:35,690
this option tree has a 100%
chance of getting a value of 8.

115
00:04:35,690 --> 00:04:39,920
So, the value of this branch
of my decision tree is 8.

116
00:04:43,616 --> 00:04:49,330
The expected value is 8.

117
00:04:49,330 --> 00:04:52,570
The value of this branch
of my decision tree

118
00:04:52,570 --> 00:04:58,471
is going to be 3/4 but--

119
00:04:58,471 --> 00:05:00,220
remember we assign
probabilities to these.

120
00:05:00,220 --> 00:05:02,230
This outcome is a
lot more likely.

121
00:05:02,230 --> 00:05:14,427
So, it would be 6 plus a
quarter of minus 7, which is--

122
00:05:14,427 --> 00:05:17,896
AUDIENCE: That's negative 7.

123
00:05:17,896 --> 00:05:19,020
ABBY NOYCE: So, what is it?

124
00:05:19,020 --> 00:05:25,710
It's 3/4 of 6, which is 4 and
1/2, plus a quarter of minus 7,

125
00:05:25,710 --> 00:05:30,150
which is minus-- so,
this is 4 and 1/2.

126
00:05:30,150 --> 00:05:35,010
This is minus 1.75, right?

127
00:05:35,010 --> 00:05:39,446
So 4 and 1/2, 3 and 1/2, 2 and--

128
00:05:39,446 --> 00:05:40,136
AUDIENCE: .75

129
00:05:40,136 --> 00:05:41,052
ABBY NOYCE: Thank you.

130
00:05:41,052 --> 00:05:41,860
3/4, yes.

131
00:05:44,440 --> 00:05:50,260
So, according to an
expected utility model,

132
00:05:50,260 --> 00:05:53,950
if this is how you weight
all of these things--

133
00:05:53,950 --> 00:05:56,650
the goodness here,
the badness over here

134
00:05:56,650 --> 00:05:59,320
and the probabilities
of each of these--

135
00:05:59,320 --> 00:06:03,370
then which of these is going
to be the more rational outcome

136
00:06:03,370 --> 00:06:05,282
to pick?

137
00:06:05,282 --> 00:06:06,250
AUDIENCE: Study.

138
00:06:06,250 --> 00:06:07,960
ABBY NOYCE: Stay home and study.

139
00:06:07,960 --> 00:06:12,330
So, on the other hand, the
fact that people probably--

140
00:06:12,330 --> 00:06:15,700
that may not be the outcome
that everybody would

141
00:06:15,700 --> 00:06:17,890
choose in this situation--

142
00:06:17,890 --> 00:06:22,000
probably shows that
people either assess

143
00:06:22,000 --> 00:06:24,250
these probabilities different
or weight the outcomes

144
00:06:24,250 --> 00:06:25,160
differently.

145
00:06:25,160 --> 00:06:28,372
But this is a model of how you
can expect people to behave.

146
00:06:28,372 --> 00:06:30,800
AUDIENCE: I think that we
forgot to take into account the

147
00:06:30,800 --> 00:06:33,310
being bored while we study.

148
00:06:33,310 --> 00:06:36,910
ABBY NOYCE: So, you think
that studying should be--

149
00:06:36,910 --> 00:06:37,870
stay home and study--

150
00:06:37,870 --> 00:06:39,411
AUDIENCE: It has a
short-term outcome

151
00:06:39,411 --> 00:06:40,640
and a long-term outcome.

152
00:06:40,640 --> 00:06:43,900
ABBY NOYCE: So, the total
outcome should be maybe lower.

153
00:06:43,900 --> 00:06:46,173
You think our assigned
values are off.

154
00:06:46,173 --> 00:06:46,756
AUDIENCE: Yes.

155
00:06:46,756 --> 00:06:50,095
AUDIENCE: What are we
considering "well" on the exam?

156
00:06:50,095 --> 00:06:53,180
ABBY NOYCE: Up to you.

157
00:06:53,180 --> 00:06:56,130
AUDIENCE: 105%.

158
00:06:56,130 --> 00:06:57,400
ABBY NOYCE: Sure.

159
00:06:57,400 --> 00:06:59,520
Anyway, this is a model.

160
00:06:59,520 --> 00:07:01,710
So, in cases where
you think that people

161
00:07:01,710 --> 00:07:03,719
are going to make
rational decisions,

162
00:07:03,719 --> 00:07:05,010
this is a model that gets used.

163
00:07:05,010 --> 00:07:07,590
So, like, economists think
about decision-making

164
00:07:07,590 --> 00:07:10,110
in terms of maximizing
your expected outcome

165
00:07:10,110 --> 00:07:12,850
and things like that.

166
00:07:12,850 --> 00:07:15,750
So, the thing is that
decisions that are actually

167
00:07:15,750 --> 00:07:18,540
made by real live human
beings, who are not

168
00:07:18,540 --> 00:07:20,790
perfectly rational
creatures, as you've probably

169
00:07:20,790 --> 00:07:22,560
noticed at some point--

170
00:07:22,560 --> 00:07:24,960
people make decisions
that are different.

171
00:07:24,960 --> 00:07:29,942
So, here's an example, kind
of a hypothetical question

172
00:07:29,942 --> 00:07:31,650
that is a good example
of how this works.

173
00:07:31,650 --> 00:07:35,812
Suppose the nation is preparing
for the outbreak of a disease.

174
00:07:35,812 --> 00:07:37,520
We know that there is
an epidemic coming.

175
00:07:37,520 --> 00:07:40,259
And it's expected
to kill 200 people.

176
00:07:40,259 --> 00:07:42,300
And the government and
the national health people

177
00:07:42,300 --> 00:07:47,400
and the CDC are like, OK, there
are two options that we could--

178
00:07:47,400 --> 00:07:49,440
kind of like 24--
there are two options

179
00:07:49,440 --> 00:07:51,080
that we could do
to deal with this.

180
00:07:51,080 --> 00:07:57,450
In one option, we are certain
to save 200 of these 600 people.

181
00:07:57,450 --> 00:08:00,780
And in the other option,
there's a one third probability

182
00:08:00,780 --> 00:08:03,780
that we'll save all of them,
but there's a 2/3 probability

183
00:08:03,780 --> 00:08:07,200
that all 600 people will
die from this disease.

184
00:08:07,200 --> 00:08:09,616
Which of these options
would you support?

185
00:08:09,616 --> 00:08:11,035
AUDIENCE: The first one.

186
00:08:11,035 --> 00:08:14,832
[INTERPOSING VOICES]

187
00:08:14,832 --> 00:08:17,040
ABBY NOYCE: You may not even
be one of the 600 people

188
00:08:17,040 --> 00:08:18,670
who's expected to die in this.

189
00:08:18,670 --> 00:08:21,990
You may be safely ensconced
in your, I don't know,

190
00:08:21,990 --> 00:08:23,940
your apocalyptic
little hideaway.

191
00:08:23,940 --> 00:08:27,090
But who thinks that option one
looks like a more appealing

192
00:08:27,090 --> 00:08:30,260
option here?

193
00:08:30,260 --> 00:08:31,750
AUDIENCE: I'd run away.

194
00:08:31,750 --> 00:08:33,375
ABBY NOYCE: Who thinks
option two looks

195
00:08:33,375 --> 00:08:35,220
like a more appealing option?

196
00:08:35,220 --> 00:08:37,350
You are forced to choose.

197
00:08:37,350 --> 00:08:38,549
Pick one or the other.

198
00:08:38,549 --> 00:08:39,080
Option one?

199
00:08:44,450 --> 00:08:45,520
Option one.

200
00:08:45,520 --> 00:08:48,010
Put your hands up where I can
see them so I can count you.

201
00:08:48,010 --> 00:08:49,270
Make a decision.

202
00:08:49,270 --> 00:08:51,490
One, two, three,
four, five, six.

203
00:08:51,490 --> 00:08:53,860
Option two?

204
00:08:53,860 --> 00:08:55,350
And three.

205
00:08:55,350 --> 00:08:59,170
Now, look at our
expected utility model.

206
00:08:59,170 --> 00:09:01,450
These come out the
same if you count.

207
00:09:01,450 --> 00:09:06,790
So, OK, one third of
600, if you multiply

208
00:09:06,790 --> 00:09:08,990
the probability times the
number, is 200 people.

209
00:09:08,990 --> 00:09:12,190
So, according to this model,
the expected utility of these

210
00:09:12,190 --> 00:09:14,380
is the same.

211
00:09:14,380 --> 00:09:17,620
And yet, a majority of people
will like option one better.

212
00:09:17,620 --> 00:09:18,820
Why?

213
00:09:18,820 --> 00:09:21,220
Well, let's get to that.

214
00:09:21,220 --> 00:09:23,050
Here's another
scenario, same idea.

215
00:09:23,050 --> 00:09:25,000
We're preparing for the
outbreak of a disease.

216
00:09:25,000 --> 00:09:28,260
There's been two
options presented.

217
00:09:28,260 --> 00:09:32,040
The first option will result
in 400 of these people

218
00:09:32,040 --> 00:09:34,180
dying, no matter what we do.

219
00:09:34,180 --> 00:09:37,570
The second option--
well, there's

220
00:09:37,570 --> 00:09:40,300
a 33% chance that
nobody will die at all,

221
00:09:40,300 --> 00:09:41,930
that everyone will be saved.

222
00:09:41,930 --> 00:09:45,760
And a 2/3 probability that
600 people are going to die.

223
00:09:45,760 --> 00:09:47,512
Which of these options
do you like better?

224
00:09:47,512 --> 00:09:48,810
AUDIENCE: Isn't
that the same thing?

225
00:09:48,810 --> 00:09:50,476
ABBY NOYCE: Who likes
option one better?

226
00:09:50,476 --> 00:09:51,760
Raise your hands.

227
00:09:51,760 --> 00:09:53,080
Who likes option two better?

228
00:09:53,080 --> 00:09:55,060
Raise your hands.

229
00:09:55,060 --> 00:09:56,759
So, that's it.

230
00:09:56,759 --> 00:09:58,550
All right, so by seeing
them both together,

231
00:09:58,550 --> 00:10:01,611
you guys notice that these
are exactly the same scenario,

232
00:10:01,611 --> 00:10:02,110
right?

233
00:10:02,110 --> 00:10:03,970
The numbers are identical.

234
00:10:03,970 --> 00:10:05,900
But if you present them
to people separately,

235
00:10:05,900 --> 00:10:07,960
what we'll find is that
in the first case, when

236
00:10:07,960 --> 00:10:11,800
this is phrased as how many
people are going to be saved,

237
00:10:11,800 --> 00:10:13,900
people like certainty.

238
00:10:13,900 --> 00:10:16,220
People are risk-averse.

239
00:10:16,220 --> 00:10:19,810
They don't want to pick
a risky situation when

240
00:10:19,810 --> 00:10:21,970
the question is couched
in terms of how much we're

241
00:10:21,970 --> 00:10:23,880
going to gain.

242
00:10:23,880 --> 00:10:25,630
When the question
is couched in terms

243
00:10:25,630 --> 00:10:28,510
of how much you might lose,
when we're looking at deaths,

244
00:10:28,510 --> 00:10:31,600
people start being
more attracted

245
00:10:31,600 --> 00:10:33,370
to the riskier possibility.

246
00:10:33,370 --> 00:10:36,400
When you have a chance of
something bad happening,

247
00:10:36,400 --> 00:10:38,900
but a chance that it might
not, people will go for risk.

248
00:10:38,900 --> 00:10:40,900
Whereas if we phrase it
as a chance of something

249
00:10:40,900 --> 00:10:43,060
good happening or a
chance that it might not,

250
00:10:43,060 --> 00:10:45,310
people will go for certainty.

251
00:10:45,310 --> 00:10:49,470
People like certainty for good
things and risk for bad things.

252
00:10:49,470 --> 00:10:51,332
We're risk-averse
when things are

253
00:10:51,332 --> 00:10:52,540
discussed in terms of losses.

254
00:10:56,710 --> 00:10:59,560
Does this make sense?

255
00:10:59,560 --> 00:11:01,840
Think of it as, like, gambling.

256
00:11:01,840 --> 00:11:06,460
If I tell you, you can
either win $50 for sure

257
00:11:06,460 --> 00:11:10,165
or you have a 75%
chance of winning

258
00:11:10,165 --> 00:11:14,514
$150 and a 25% chance of
winning nothing, which of these

259
00:11:14,514 --> 00:11:15,430
are you going to pick?

260
00:11:15,430 --> 00:11:18,972
50 certain dollars or a gamble
on getting more than that?

261
00:11:18,972 --> 00:11:20,280
AUDIENCE: More.

262
00:11:20,280 --> 00:11:22,860
ABBY NOYCE: Who would
pick 50 certain dollars?

263
00:11:22,860 --> 00:11:24,990
Who would pick the gamble?

264
00:11:29,059 --> 00:11:31,780
AUDIENCE: A 75% chance of what?

265
00:11:31,780 --> 00:11:34,817
ABBY NOYCE: $150.

266
00:11:34,817 --> 00:11:37,150
Expected utility model says
that these are equal, right?

267
00:11:40,227 --> 00:11:42,060
But when it's couched
as things you can get,

268
00:11:42,060 --> 00:11:43,920
people tend to be--

269
00:11:43,920 --> 00:11:46,420
and there's going to be
variations in this for amounts

270
00:11:46,420 --> 00:11:48,669
and for personality-- people
tend to be more attracted

271
00:11:48,669 --> 00:11:49,730
to the certain thing.

272
00:11:49,730 --> 00:11:52,063
If I spin that the other way
around and I said, well you

273
00:11:52,063 --> 00:11:54,984
could either lose
$50 for certain

274
00:11:54,984 --> 00:11:56,650
or you could take a
gamble where there's

275
00:11:56,650 --> 00:12:02,224
a 25% chance of losing nothing
or a 75% chance of losing $150,

276
00:12:02,224 --> 00:12:03,640
which one are you
going to prefer?

277
00:12:03,640 --> 00:12:04,681
AUDIENCE: The second one.

278
00:12:04,681 --> 00:12:05,650
ABBY NOYCE: The gamble.

279
00:12:05,650 --> 00:12:07,210
Who thinks they
prefer the gamble?

280
00:12:07,210 --> 00:12:10,540
Who thinks they prefer
to lose $50 for certain?

281
00:12:10,540 --> 00:12:14,860
Again, the expected utility
math says that they're the same.

282
00:12:14,860 --> 00:12:16,633
But people like the risk.

283
00:12:16,633 --> 00:12:19,531
AUDIENCE: Because there's a
difference between losing $50

284
00:12:19,531 --> 00:12:21,463
and not gaining $50.

285
00:12:26,250 --> 00:12:27,810
Losing $50 means you lose money.

286
00:12:27,810 --> 00:12:31,280
Not gaining $50 means you
just don't-- nothing happens.

287
00:12:31,280 --> 00:12:34,770
ABBY NOYCE: That's part of it.

288
00:12:34,770 --> 00:12:36,310
So, somebody put
together something

289
00:12:36,310 --> 00:12:39,040
that's similar with a
study, that they said, so,

290
00:12:39,040 --> 00:12:41,450
imagine you're at a
casino and you either--

291
00:12:41,450 --> 00:12:43,480
and they set it up as
two possible scenarios.

292
00:12:43,480 --> 00:12:46,890
You win $200 and
you have a chance

293
00:12:46,890 --> 00:12:49,419
of gaining some more beyond
that, or you win $400

294
00:12:49,419 --> 00:12:51,460
and you have a chance of
losing some beyond that,

295
00:12:51,460 --> 00:12:53,376
so that all the scenarios
kind of average out.

296
00:12:53,376 --> 00:12:55,547
So, you ended up with
$250 or something.

297
00:12:55,547 --> 00:12:57,130
And I don't remember
the exact numbers

298
00:12:57,130 --> 00:12:58,300
and I could probably
work it out but it

299
00:12:58,300 --> 00:12:59,740
would take me 10 minutes.

300
00:12:59,740 --> 00:13:03,700
And what they found is that
even when your net outcome would

301
00:13:03,700 --> 00:13:06,850
have been the same across either
the win $200 and then win some

302
00:13:06,850 --> 00:13:11,710
more versus win $400 and then
lose some, people are, again,

303
00:13:11,710 --> 00:13:13,210
more attracted-- they're--

304
00:13:13,210 --> 00:13:16,960
a certain thing, when they have
a chance of gaining something,

305
00:13:16,960 --> 00:13:19,660
they'd rather take a
smaller, certain gain,

306
00:13:19,660 --> 00:13:22,570
than a riskier, larger gain.

307
00:13:22,570 --> 00:13:25,240
And the opposite is true when
they stand to lose something.

308
00:13:25,240 --> 00:13:29,790
They'd rather take a
riskier possible larger loss

309
00:13:29,790 --> 00:13:33,340
but possible not losing
anything, versus a smaller

310
00:13:33,340 --> 00:13:34,270
certain loss.

311
00:13:34,270 --> 00:13:35,770
So, even when the amounts are--

312
00:13:35,770 --> 00:13:39,970
so, our perception of
the change is different.

313
00:13:39,970 --> 00:13:45,540
It's like we feel like gaining
$50 is less bad than losing

314
00:13:45,540 --> 00:13:49,330
$50, that one of these is a
bigger change than the other.

315
00:13:52,578 --> 00:13:54,246
AUDIENCE: Is it
like the [INAUDIBLE]

316
00:13:54,246 --> 00:13:56,454
when you have nothing to
lose, like kind of go for it

317
00:13:56,454 --> 00:13:58,521
type thing?

318
00:13:58,521 --> 00:14:00,270
ABBY NOYCE: It's
probably related to that,

319
00:14:00,270 --> 00:14:07,650
although I'd say that that
might more come from this risk

320
00:14:07,650 --> 00:14:09,540
attraction to risky--

321
00:14:09,540 --> 00:14:11,056
when we're losing something.

322
00:14:11,056 --> 00:14:12,930
So, you're like, eh,
what am I going to lose?

323
00:14:17,520 --> 00:14:19,130
Actually I suspect
that it's more like

324
00:14:19,130 --> 00:14:20,880
if you feel like you
have nothing to lose,

325
00:14:20,880 --> 00:14:23,290
then you're not risking as much.

326
00:14:23,290 --> 00:14:25,350
So, what might seem
like a risky possibility

327
00:14:25,350 --> 00:14:28,241
to someone who's got a vested
interest just isn't as much

328
00:14:28,241 --> 00:14:28,740
of a risk.

329
00:14:28,740 --> 00:14:30,281
Like the bad things
that could happen

330
00:14:30,281 --> 00:14:32,100
are less bad, in that case.

331
00:14:32,100 --> 00:14:33,420
But it's probably related.

332
00:14:33,420 --> 00:14:37,200
There's a lot of studies looking
at how people assess risk.

333
00:14:37,200 --> 00:14:39,690
And the outcome is that in
a lot of ways we're not--

334
00:14:39,690 --> 00:14:41,856
you can set up a lot of
scenarios in which we're not

335
00:14:41,856 --> 00:14:43,260
very good at it.

336
00:14:43,260 --> 00:14:44,910
Probably because
the kinds of risks

337
00:14:44,910 --> 00:14:47,100
that we evolved to
deal with, the risks

338
00:14:47,100 --> 00:14:49,200
you have to deal with as
a primate running around

339
00:14:49,200 --> 00:14:51,780
on the Savannah trying to
not get eaten by a lion,

340
00:14:51,780 --> 00:14:53,670
are very different
from the kinds of risks

341
00:14:53,670 --> 00:14:57,420
you have to assess as a
human being living in a city

342
00:14:57,420 --> 00:15:00,840
and dealing with cars
and finances and all

343
00:15:00,840 --> 00:15:03,150
of these other things
that we try and make

344
00:15:03,150 --> 00:15:04,230
sense out of every day.

345
00:15:04,230 --> 00:15:06,800
So, our intuitive
math for these things

346
00:15:06,800 --> 00:15:08,550
doesn't deal well with
a lot of situations

347
00:15:08,550 --> 00:15:10,550
that we find ourselves in.

348
00:15:10,550 --> 00:15:11,530
Yeah?

349
00:15:11,530 --> 00:15:15,078
AUDIENCE: If you
start off with $200,

350
00:15:15,078 --> 00:15:18,894
don't people feel that,
Oh, I already have $200.

351
00:15:18,894 --> 00:15:20,325
So, I can only gain.

352
00:15:20,325 --> 00:15:22,710
I can't lose.

353
00:15:22,710 --> 00:15:26,940
Like, they have something
could push off and go forward.

354
00:15:26,940 --> 00:15:29,860
Because if you have
$400, they're like,

355
00:15:29,860 --> 00:15:33,869
oh, I can lose more
than $200 or something.

356
00:15:33,869 --> 00:15:35,410
ABBY NOYCE: Maybe,
although I suspect

357
00:15:35,410 --> 00:15:37,110
you'd see the same
outcomes no matter

358
00:15:37,110 --> 00:15:38,940
what your base amounts were.

359
00:15:38,940 --> 00:15:41,430
I suspect you'd see the
same thing with $10 and $20

360
00:15:41,430 --> 00:15:45,816
and values ending up
somewhere in the middle.

361
00:15:45,816 --> 00:15:48,310
And I suspect you'd see it--

362
00:15:48,310 --> 00:15:53,960
I think it's a factor of how
the questions are phrased,

363
00:15:53,960 --> 00:15:55,710
whether it's set up
as a situation of loss

364
00:15:55,710 --> 00:16:00,010
or a situation of gain, versus
where the baselines actually

365
00:16:00,010 --> 00:16:00,510
are.

366
00:16:04,880 --> 00:16:05,470
All right.

367
00:16:05,470 --> 00:16:08,010
So, humans are kind of
generally bad at numbers.

368
00:16:08,010 --> 00:16:09,560
We're good at some
other things too.

369
00:16:09,560 --> 00:16:11,860
So, the rules of
thumb that people

370
00:16:11,860 --> 00:16:14,020
use when they're
trying to decide--

371
00:16:14,020 --> 00:16:16,690
and this is less
decision making per se--

372
00:16:16,690 --> 00:16:19,300
but when we're trying to
figure out how to make

373
00:16:19,300 --> 00:16:21,850
sense out of the world.

374
00:16:21,850 --> 00:16:25,450
Who here has taken some amount
of CS, done some programming?

375
00:16:25,450 --> 00:16:27,770
One-- some amount
of computer science,

376
00:16:27,770 --> 00:16:30,970
some amount of programming
in school or for a hobby.

377
00:16:30,970 --> 00:16:31,660
A few, OK.

378
00:16:31,660 --> 00:16:33,760
So, when you're
writing software,

379
00:16:33,760 --> 00:16:35,830
you're interested in
algorithms, right?

380
00:16:35,830 --> 00:16:38,380
So, an algorithm is a series
of steps that will always

381
00:16:38,380 --> 00:16:41,710
get you a right answer.

382
00:16:41,710 --> 00:16:44,650
People, humans, human
brains, don't seem

383
00:16:44,650 --> 00:16:45,970
to be so much on algorithms.

384
00:16:45,970 --> 00:16:47,670
We tend to take a
lot of shortcuts.

385
00:16:47,670 --> 00:16:49,420
We tend to use what
are called heuristics.

386
00:16:49,420 --> 00:16:51,460
And your heuristic is
like a general strategy

387
00:16:51,460 --> 00:16:54,160
that will usually get
you the right answer.

388
00:16:54,160 --> 00:16:55,970
It's good enough.

389
00:16:55,970 --> 00:17:00,010
But if you're interested in
how decision-making works,

390
00:17:00,010 --> 00:17:03,430
how people make some kinds of
judgment calls or things like

391
00:17:03,430 --> 00:17:05,725
that, then-- just
like how, in vision,

392
00:17:05,725 --> 00:17:08,349
it's interesting to look at the
places where your visual system

393
00:17:08,349 --> 00:17:11,470
makes mistakes, looking at
things like optical illusions

394
00:17:11,470 --> 00:17:13,670
in order to understand
how it works--

395
00:17:13,670 --> 00:17:15,910
people who study
decision-making look at places

396
00:17:15,910 --> 00:17:20,050
where these heuristics seem to
lead us to making bad decisions

397
00:17:20,050 --> 00:17:23,740
or decisions that are wrong
relative to the facts or we

398
00:17:23,740 --> 00:17:27,440
will come to a conclusion
that is not actually true.

399
00:17:27,440 --> 00:17:30,430
So, I want to talk about
three major heuristics

400
00:17:30,430 --> 00:17:35,560
that we seem to use in order to
understand how our world works.

401
00:17:35,560 --> 00:17:39,880
First up is what's called
the representative heuristic.

402
00:17:39,880 --> 00:17:44,950
So, suppose you toss
a coin six times.

403
00:17:44,950 --> 00:17:49,720
Which of these potential
outcomes is most likely?

404
00:17:49,720 --> 00:17:53,970
Heads, heads, heads, tails,
tails, tails, or tails, heads,

405
00:17:53,970 --> 00:17:58,120
tails, heads, tails, heads,
or heads, tails, tails, heads,

406
00:17:58,120 --> 00:18:04,090
tails, heads, or heads, heads,
heads, heads, heads, heads?

407
00:18:04,090 --> 00:18:06,212
Which of these is most likely?

408
00:18:06,212 --> 00:18:08,180
AUDIENCE: It's the
same, no matter what?

409
00:18:08,180 --> 00:18:12,115
ABBY NOYCE: It is the
same no matter what.

410
00:18:12,115 --> 00:18:15,190
How many people have kind
of an intuitive perception

411
00:18:15,190 --> 00:18:16,690
that, for example,
the last of these

412
00:18:16,690 --> 00:18:18,940
is less likely if
you looked at that?

413
00:18:18,940 --> 00:18:20,920
Who would kind of, even
if you know the math,

414
00:18:20,920 --> 00:18:23,037
know better from a
math perspective.

415
00:18:23,037 --> 00:18:24,370
But it feels less likely, right?

416
00:18:24,370 --> 00:18:25,995
If you flipped a coin
six times and got

417
00:18:25,995 --> 00:18:28,450
all heads you'd be like, whoa.

418
00:18:28,450 --> 00:18:30,520
And yet, it's no
less likely than

419
00:18:30,520 --> 00:18:33,530
any other specific pattern.

420
00:18:33,530 --> 00:18:35,440
But what happens
is that because we

421
00:18:35,440 --> 00:18:38,680
know that we're taking a
random sample from a pretty

422
00:18:38,680 --> 00:18:43,270
random decision space, we
think that it's more likely

423
00:18:43,270 --> 00:18:47,230
if the sample is similar to
that population from which it's

424
00:18:47,230 --> 00:18:48,780
selected.

425
00:18:48,780 --> 00:18:54,420
So, this is kind of the,
if it looks like a duck

426
00:18:54,420 --> 00:18:56,560
and it walks like a duck
and it quacks like a duck,

427
00:18:56,560 --> 00:18:59,020
it's probably a duck, heuristic.

428
00:18:59,020 --> 00:19:00,480
If the sample that
we're looking at

429
00:19:00,480 --> 00:19:04,860
seems to match the population
that it's drawn from,

430
00:19:04,860 --> 00:19:07,700
then this is probably true.

431
00:19:07,700 --> 00:19:10,290
So, representativeness
heuristics, for all

432
00:19:10,290 --> 00:19:12,870
that we're going to talk about
the places where they break,

433
00:19:12,870 --> 00:19:15,842
heuristics are generally useful.

434
00:19:15,842 --> 00:19:16,800
I mean, think about it.

435
00:19:16,800 --> 00:19:17,549
They've got to be.

436
00:19:17,549 --> 00:19:19,094
We wouldn't hang
on to strategies

437
00:19:19,094 --> 00:19:20,760
for getting around
in the world, if they

438
00:19:20,760 --> 00:19:23,410
weren't reasonably good at it.

439
00:19:23,410 --> 00:19:26,040
You'd get wiped out
of the gene pool.

440
00:19:26,040 --> 00:19:27,720
It just wouldn't stick around.

441
00:19:27,720 --> 00:19:29,400
So, representativeness
heuristics

442
00:19:29,400 --> 00:19:32,310
are good for some things.

443
00:19:32,310 --> 00:19:37,320
They fail when dealing
with randomness-- people

444
00:19:37,320 --> 00:19:39,360
are generally bad
with randomness.

445
00:19:39,360 --> 00:19:41,250
They also fail when
you're thinking

446
00:19:41,250 --> 00:19:43,690
about complicated
categories-- categories

447
00:19:43,690 --> 00:19:46,210
that human beings
tend to fall into.

448
00:19:52,042 --> 00:20:05,910
So, If I was to describe for
you my friend, Marie, and Marie

449
00:20:05,910 --> 00:20:08,850
is a sophomore in college.

450
00:20:08,850 --> 00:20:10,950
And she tends to
wear a lot of black,

451
00:20:10,950 --> 00:20:13,530
like a black t-shirt with
a long sleeve fishnet

452
00:20:13,530 --> 00:20:17,760
shirt underneath it,
baggy black skater pants

453
00:20:17,760 --> 00:20:20,070
with some D-rings
and straps on them.

454
00:20:20,070 --> 00:20:21,967
You guys know the look, right?

455
00:20:21,967 --> 00:20:23,550
Which of these two
things do you think

456
00:20:23,550 --> 00:20:27,420
is more likely to
be true about Marie?

457
00:20:27,420 --> 00:20:35,220
Marie is an economics major
or Marie is an economics

458
00:20:35,220 --> 00:20:38,190
major and one of
the folks who puts

459
00:20:38,190 --> 00:20:41,340
on Rocky Horror at our
university every Halloween.

460
00:20:41,340 --> 00:20:45,562
Option A or option B,
which is more likely?

461
00:20:45,562 --> 00:20:46,770
So, who thinks it's option A?

462
00:20:46,770 --> 00:20:47,830
Raise your hands.

463
00:20:47,830 --> 00:20:48,930
Who thinks it's option B?

464
00:20:48,930 --> 00:20:49,638
Raise your hands.

465
00:20:52,560 --> 00:20:54,800
Which is odd because
the math says

466
00:20:54,800 --> 00:20:57,710
that if I've got a conjunction
of two things, where Marie

467
00:20:57,710 --> 00:21:01,730
is both an economics major and
helps put on the Rocky Horror

468
00:21:01,730 --> 00:21:05,360
Show, that can't be more
likely than simply the fact

469
00:21:05,360 --> 00:21:06,960
that she's an economics major.

470
00:21:06,960 --> 00:21:10,100
So, this is a case where
the fact that Marie,

471
00:21:10,100 --> 00:21:13,040
this small sample
of one, matches

472
00:21:13,040 --> 00:21:17,079
our perception of people
who do Rocky Horror better

473
00:21:17,079 --> 00:21:19,370
than our perception of people
who are economics majors.

474
00:21:19,370 --> 00:21:24,110
That makes it seem like one of
these things is more likely.

475
00:21:24,110 --> 00:21:25,470
Something to watch out for.

476
00:21:25,470 --> 00:21:28,070
So, this is, in part--

477
00:21:28,070 --> 00:21:30,676
oh, one more one more kind of--

478
00:21:30,676 --> 00:21:32,300
here, think about
some stats questions.

479
00:21:32,300 --> 00:21:36,050
So, picture a town that has
two hospitals, a big one

480
00:21:36,050 --> 00:21:38,690
and a little one.

481
00:21:38,690 --> 00:21:40,370
They both have an OBGYN unit.

482
00:21:40,370 --> 00:21:41,960
They both have babies.

483
00:21:41,960 --> 00:21:44,750
About 45 babies a day are
born in the big hospital

484
00:21:44,750 --> 00:21:47,610
and about 15 in the little one.

485
00:21:47,610 --> 00:21:50,450
And you guys probably know that
pretty much half of all babies

486
00:21:50,450 --> 00:21:51,711
are boys, right?

487
00:21:51,711 --> 00:21:54,210
But there's going to be some
day to day fluctuation in this.

488
00:21:54,210 --> 00:21:56,543
So, you might have one day
there's more boys than girls,

489
00:21:56,543 --> 00:21:58,250
one day there's more
girls than boys.

490
00:21:58,250 --> 00:22:01,680
But over time, it would
average out to 50%.

491
00:22:01,680 --> 00:22:03,830
So, suppose that these
hospitals are keeping

492
00:22:03,830 --> 00:22:09,290
a record of how often
there were more than 60%

493
00:22:09,290 --> 00:22:12,321
of the babies born were boys.

494
00:22:12,321 --> 00:22:13,820
Which hospital do
you think is going

495
00:22:13,820 --> 00:22:16,500
to have more days
in a given year

496
00:22:16,500 --> 00:22:20,780
where there are
more than 60% boys?

497
00:22:20,780 --> 00:22:24,200
More than 60% of the babies
born that day were boys.

498
00:22:24,200 --> 00:22:27,500
Larger or smaller
or about the same?

499
00:22:33,470 --> 00:22:35,120
Who thinks it's larger?

500
00:22:35,120 --> 00:22:36,470
Smaller?

501
00:22:36,470 --> 00:22:39,440
About the same?

502
00:22:39,440 --> 00:22:41,370
That's what most
people tend to say.

503
00:22:41,370 --> 00:22:45,650
But this is one of those
things where you have

504
00:22:45,650 --> 00:22:48,390
what's called the
small sample fallacy.

505
00:22:48,390 --> 00:22:51,410
So, the sample size is--

506
00:22:51,410 --> 00:22:53,840
the fact that the smaller
hospital has a smaller sample

507
00:22:53,840 --> 00:22:58,160
size, it's going to see
skews from the average

508
00:22:58,160 --> 00:22:59,270
more frequently.

509
00:22:59,270 --> 00:23:01,310
If I toss a coin three
times, you probably

510
00:23:01,310 --> 00:23:02,851
wouldn't be shocked
if they turned up

511
00:23:02,851 --> 00:23:04,910
all heads or all tails, right?

512
00:23:04,910 --> 00:23:06,770
If I tossed the coin
20 times, and it

513
00:23:06,770 --> 00:23:09,080
was all heads or all tails,
you'd be a little bit more

514
00:23:09,080 --> 00:23:09,830
taken aback.

515
00:23:14,590 --> 00:23:16,510
If I tossed it 100
times and got all hands,

516
00:23:16,510 --> 00:23:17,843
it would be really weird, right?

517
00:23:21,290 --> 00:23:22,940
Getting three heads for--

518
00:23:25,650 --> 00:23:27,290
there's, what,
like a 12.5% chance

519
00:23:27,290 --> 00:23:29,877
of getting three heads and
something much smaller--

520
00:23:29,877 --> 00:23:32,210
much, much much smaller than
that-- chance of getting 20

521
00:23:32,210 --> 00:23:34,370
heads, 1 over 2 to the 20.

522
00:23:37,000 --> 00:23:41,490
So, when you take
a small sample,

523
00:23:41,490 --> 00:23:44,260
the smaller your sample is
relative to the population

524
00:23:44,260 --> 00:23:47,080
you're drawing from,
then the more likely

525
00:23:47,080 --> 00:23:51,170
your sample is to
be really different.

526
00:23:51,170 --> 00:23:54,102
Suppose I went to your high
school and just closed my eyes,

527
00:23:54,102 --> 00:23:56,560
had the list of all the kids
in your grade, closed my eyes,

528
00:23:56,560 --> 00:23:57,910
and picked three--

529
00:23:57,910 --> 00:24:00,070
what are the chances
they'd all be girls?

530
00:24:00,070 --> 00:24:01,840
Not huge but it
wouldn't be shocking

531
00:24:01,840 --> 00:24:04,670
if I picked out three kids at
random and they were all girls.

532
00:24:04,670 --> 00:24:06,940
What if I picked out 50?

533
00:24:06,940 --> 00:24:08,020
It's a lot less likely.

534
00:24:11,297 --> 00:24:13,630
The larger the sample that
we're taking, the more likely

535
00:24:13,630 --> 00:24:16,630
it is to be representative
of the population

536
00:24:16,630 --> 00:24:17,920
we're drawing from.

537
00:24:17,920 --> 00:24:20,440
So, one of the places
where people make mistakes

538
00:24:20,440 --> 00:24:21,910
with this representative
heuristic,

539
00:24:21,910 --> 00:24:24,500
is when they're looking
at small sample sizes

540
00:24:24,500 --> 00:24:28,390
but not taking that into account
when they try and consider what

541
00:24:28,390 --> 00:24:30,280
the larger population would be.

542
00:24:30,280 --> 00:24:31,870
This is the same
mistake people make

543
00:24:31,870 --> 00:24:34,732
if they know three people
who fall into a category

544
00:24:34,732 --> 00:24:36,940
and therefore make some kind
of broad generalization,

545
00:24:36,940 --> 00:24:39,910
like, I know three
guys who are engineers,

546
00:24:39,910 --> 00:24:42,400
therefore all
engineers are guys.

547
00:24:42,400 --> 00:24:43,850
No, it doesn't work that way.

548
00:24:43,850 --> 00:24:46,000
But you've got to
watch out for it.

549
00:24:46,000 --> 00:24:47,320
All right.

550
00:24:47,320 --> 00:24:52,970
Another pattern people use is
what's called the availability

551
00:24:52,970 --> 00:24:53,470
heuristic.

552
00:24:53,470 --> 00:24:55,190
So, read this list of names.

553
00:24:58,250 --> 00:25:00,000
And look at me when
you're done so I know.

554
00:25:06,505 --> 00:25:07,130
Just read them.

555
00:25:07,130 --> 00:25:09,505
I'm not going to quiz you on
them or anything, I promise.

556
00:25:12,740 --> 00:25:13,880
All right.

557
00:25:13,880 --> 00:25:16,226
Notice anything about them?

558
00:25:16,226 --> 00:25:18,100
Anything that calls
itself to your attention?

559
00:25:18,100 --> 00:25:19,391
Yeah, some of them are authors.

560
00:25:19,391 --> 00:25:21,060
Good.

561
00:25:21,060 --> 00:25:23,211
AUDIENCE: I only recognize
the author names.

562
00:25:23,211 --> 00:25:24,460
ABBY NOYCE: Here's a question.

563
00:25:24,460 --> 00:25:26,702
Were there more women's
names or men's names?

564
00:25:26,702 --> 00:25:27,368
AUDIENCE: Men's.

565
00:25:27,368 --> 00:25:27,862
AUDIENCE: Men's.

566
00:25:27,862 --> 00:25:28,945
AUDIENCE: I couldn't tell.

567
00:25:30,999 --> 00:25:33,290
ABBY NOYCE: Who thinks there
were more guys' names than

568
00:25:33,290 --> 00:25:34,310
women's names?

569
00:25:34,310 --> 00:25:36,770
Who thinks there were more
women's names than men's names?

570
00:25:36,770 --> 00:25:37,800
Got to pick one.

571
00:25:37,800 --> 00:25:38,510
I'm going to make you pick one.

572
00:25:38,510 --> 00:25:39,510
AUDIENCE: [INAUDIBLE]

573
00:25:39,510 --> 00:25:41,470
ABBY NOYCE: Nope.

574
00:25:41,470 --> 00:25:43,250
Forced choice, you've
got to pick one.

575
00:25:43,250 --> 00:25:46,150
Who thinks there were more
women than men listed there?

576
00:25:46,150 --> 00:25:47,462
AUDIENCE: [INAUDIBLE]

577
00:25:47,462 --> 00:25:49,670
ABBY NOYCE: Who thinks there
were more men than women

578
00:25:49,670 --> 00:25:50,170
listed?

579
00:26:00,540 --> 00:26:02,071
So, there's more
guys than women.

580
00:26:02,071 --> 00:26:03,570
On the other hand,
you guys probably

581
00:26:03,570 --> 00:26:05,190
noticed that the
women's names are

582
00:26:05,190 --> 00:26:06,824
names you recognized, right?

583
00:26:06,824 --> 00:26:07,740
At least some of them.

584
00:26:07,740 --> 00:26:09,230
The guys' names, probably not.

585
00:26:09,230 --> 00:26:11,470
The guys' names are just
kind of pulled at random.

586
00:26:11,470 --> 00:26:14,940
So, the availability heuristic
is a guideline we use,

587
00:26:14,940 --> 00:26:17,610
where we judge frequency
of an occurrence

588
00:26:17,610 --> 00:26:21,480
based on how easy it is for
us to pull up an example,

589
00:26:21,480 --> 00:26:24,060
pull up relevant examples of it.

590
00:26:24,060 --> 00:26:26,320
This is kind of the
mistake that's being made--

591
00:26:26,320 --> 00:26:28,590
who here knows somebody
who is afraid of flying,

592
00:26:28,590 --> 00:26:30,720
doesn't like to fly,
possibly because they

593
00:26:30,720 --> 00:26:33,303
think that they're going to die
in some kind of terrible plane

594
00:26:33,303 --> 00:26:34,950
crash?

595
00:26:34,950 --> 00:26:36,880
Who here knows somebody
who's afraid to fly,

596
00:26:36,880 --> 00:26:39,680
who drives or rides in
cars that people drive?

597
00:26:43,650 --> 00:26:44,820
What's with that?

598
00:26:44,820 --> 00:26:47,130
We all know that you're way
more likely to die terribly

599
00:26:47,130 --> 00:26:48,796
in a car accident
than a plane accident,

600
00:26:48,796 --> 00:26:52,740
by like a factor of 100, right?

601
00:26:52,740 --> 00:26:56,410
But a lot of people feel
like dying in a plane crash

602
00:26:56,410 --> 00:26:57,240
is more likely.

603
00:26:57,240 --> 00:27:00,150
And one of the reasons for
this is that car accidents

604
00:27:00,150 --> 00:27:02,535
don't generally make the news.

605
00:27:02,535 --> 00:27:03,660
Car accidents happen.

606
00:27:03,660 --> 00:27:06,780
They might show up on page four
of the newspaper or something.

607
00:27:06,780 --> 00:27:11,910
Whereas plane crashes are a
week-long media extravaganza.

608
00:27:11,910 --> 00:27:14,965
They're on the news 24/7
if a plane goes down.

609
00:27:14,965 --> 00:27:16,340
You've probably
encountered this.

610
00:27:16,340 --> 00:27:17,756
So, one of the
things that happens

611
00:27:17,756 --> 00:27:21,290
is if you're trying to assess
the relative frequency of car

612
00:27:21,290 --> 00:27:25,740
accidents versus
plane crashes, then

613
00:27:25,740 --> 00:27:29,250
you might have an easier time
retrieving the plane crash

614
00:27:29,250 --> 00:27:30,477
incidents from your memory.

615
00:27:30,477 --> 00:27:31,560
You heard more about them.

616
00:27:31,560 --> 00:27:33,060
You thought more about them.

617
00:27:33,060 --> 00:27:34,770
They were big news
when they happened.

618
00:27:34,770 --> 00:27:38,280
And so, you perceive
plane crashes

619
00:27:38,280 --> 00:27:41,940
as being more frequent
than car crashes.

620
00:27:41,940 --> 00:27:46,260
So, this availability heuristic
is good for lots of things.

621
00:27:46,260 --> 00:27:49,770
If I ask you, do
you think there are

622
00:27:49,770 --> 00:27:52,491
more guys or girls in
your high school class,

623
00:27:52,491 --> 00:27:53,490
or is it about the same?

624
00:28:02,115 --> 00:28:03,990
It's probably pretty
close to about the same.

625
00:28:03,990 --> 00:28:07,510
But one of the ways that we
answer this kind of question

626
00:28:07,510 --> 00:28:09,760
is by thinking through the
members of our high school

627
00:28:09,760 --> 00:28:13,270
class that we know, and saying,
how many of them are girls?

628
00:28:13,270 --> 00:28:15,960
How many of them are guys?

629
00:28:15,960 --> 00:28:19,925
And so, if the number of girls
and guys that you can recall

630
00:28:19,925 --> 00:28:21,550
is roughly equal,
you're going to guess

631
00:28:21,550 --> 00:28:23,410
that they're about the same.

632
00:28:23,410 --> 00:28:27,460
So, two things that affect
how available something is

633
00:28:27,460 --> 00:28:30,010
in memory, are what's called--

634
00:28:30,010 --> 00:28:32,150
one of them is recency,
so, how recently we

635
00:28:32,150 --> 00:28:33,670
thought about something.

636
00:28:33,670 --> 00:28:36,434
So, if you ask
people, what do you

637
00:28:36,434 --> 00:28:38,350
think your odds are of
dying in a plane crash,

638
00:28:38,350 --> 00:28:41,140
and ask people to
estimate this--

639
00:28:41,140 --> 00:28:43,260
and you ask them
to do it and then--

640
00:28:43,260 --> 00:28:46,550
well probably if there
was a big plane crash

641
00:28:46,550 --> 00:28:48,690
and it's in the news,
and a week later you

642
00:28:48,690 --> 00:28:50,440
ask people to estimate
what their odds are

643
00:28:50,440 --> 00:28:53,630
of dying in a plane crash,
they'll give you one number.

644
00:28:53,630 --> 00:28:56,770
And suppose six months later,
you find the same people

645
00:28:56,770 --> 00:29:00,530
and ask them to give you
another estimate, which

646
00:29:00,530 --> 00:29:02,780
of these numbers do you
think is going to be bigger?

647
00:29:02,780 --> 00:29:03,781
AUDIENCE: The first one?

648
00:29:03,781 --> 00:29:06,071
ABBY NOYCE: Probably the
first one, the one immediately

649
00:29:06,071 --> 00:29:07,060
after whatever it is.

650
00:29:07,060 --> 00:29:09,730
Because the recency
of the event makes

651
00:29:09,730 --> 00:29:11,170
it more available to them.

652
00:29:11,170 --> 00:29:12,250
It's easier to remember.

653
00:29:12,250 --> 00:29:14,930
It's right there on
top of your memory.

654
00:29:14,930 --> 00:29:17,530
And so, your perception of how
frequently the stuff happens

655
00:29:17,530 --> 00:29:19,442
is skewed by that.

656
00:29:19,442 --> 00:29:22,700
AUDIENCE: I heard
that's how people choose

657
00:29:22,700 --> 00:29:24,399
the numbers that
they gamble on when

658
00:29:24,399 --> 00:29:25,690
they're playing, say, roulette.

659
00:29:25,690 --> 00:29:29,155
They say, hey, this number
has appeared frequently.

660
00:29:29,155 --> 00:29:32,310
I'm going to bet on this one
because it's a "hot" number.

661
00:29:32,310 --> 00:29:34,120
ABBY NOYCE: Yeah,
you'll see that.

662
00:29:34,120 --> 00:29:35,050
One of these things--

663
00:29:35,050 --> 00:29:37,720
it's called the
gambler's fallacy

664
00:29:37,720 --> 00:29:40,300
in something that's entirely
random like roulette,

665
00:29:40,300 --> 00:29:42,070
like flipping a coin.

666
00:29:42,070 --> 00:29:44,560
Where-- but there's
this kind of idea

667
00:29:44,560 --> 00:29:48,130
because it really feels like
whatever has just happened

668
00:29:48,130 --> 00:29:50,740
is going to affect
what happens next.

669
00:29:50,740 --> 00:29:52,915
Probably not true.

670
00:29:52,915 --> 00:29:54,545
AUDIENCE: OK, is
there a game where

671
00:29:54,545 --> 00:29:58,461
you have a bullet in the gun but
it's only one bullet and then

672
00:29:58,461 --> 00:30:00,710
you spin the barrel and then
you try to kill yourself?

673
00:30:00,710 --> 00:30:01,930
AUDIENCE: I think
that's Russian Roulette.

674
00:30:01,930 --> 00:30:02,386
ABBY NOYCE: Russian Roulette.

675
00:30:02,386 --> 00:30:04,210
AUDIENCE: It is a game, though.

676
00:30:04,210 --> 00:30:07,000
ABBY NOYCE: I'm not sure
I'd call it a game, but yes.

677
00:30:11,214 --> 00:30:12,880
More often you'll
hear it in the context

678
00:30:12,880 --> 00:30:15,100
of being used as a
metaphor, like such and so

679
00:30:15,100 --> 00:30:17,190
is equivalent to playing--

680
00:30:17,190 --> 00:30:30,970
[SIDE CONVERSATION]

681
00:30:30,970 --> 00:30:35,020
ABBY NOYCE: So, the
other thing that

682
00:30:35,020 --> 00:30:39,100
affects this availability
idea is how familiar you

683
00:30:39,100 --> 00:30:39,980
are with the outcome.

684
00:30:39,980 --> 00:30:44,290
So, again, and this is the
thing where the level of--

685
00:30:44,290 --> 00:30:47,410
how many people
think that more kids

686
00:30:47,410 --> 00:30:51,012
get kidnapped in the United
States today than 50 years ago?

687
00:30:51,012 --> 00:30:52,410
AUDIENCE: I don't know.

688
00:30:52,410 --> 00:30:54,280
I've never really
thought about it.

689
00:30:54,280 --> 00:30:55,420
ABBY NOYCE: This is a
more interesting question

690
00:30:55,420 --> 00:30:57,880
to ask grownups, people who
actually were paying attention

691
00:30:57,880 --> 00:31:00,700
to the news 20, 30 years ago.

692
00:31:00,700 --> 00:31:05,045
Because the thing is that
it's actually pretty stable--

693
00:31:05,045 --> 00:31:07,170
kidnapping rates in this
country are pretty stable.

694
00:31:07,170 --> 00:31:09,640
And yet, over time,
you'll see things

695
00:31:09,640 --> 00:31:11,500
like parents get more concerned.

696
00:31:11,500 --> 00:31:14,470
They don't let their kids
play on their own as much.

697
00:31:14,470 --> 00:31:18,930
They don't let their kids get
places on their own as much.

698
00:31:18,930 --> 00:31:21,370
You'll see these patterns,
where if you ask your parents,

699
00:31:21,370 --> 00:31:22,850
were they allowed to
go around town alone?

700
00:31:22,850 --> 00:31:24,433
Were they allowed
to take the T alone?

701
00:31:24,433 --> 00:31:27,160
Were they allowed to ride
their bike to school?

702
00:31:27,160 --> 00:31:29,010
You'll see things where
30 years ago, people

703
00:31:29,010 --> 00:31:30,260
were allowed to do this stuff.

704
00:31:30,260 --> 00:31:31,218
And today, kids aren't.

705
00:31:31,218 --> 00:31:34,990
And one of the things I
think that affects this--

706
00:31:34,990 --> 00:31:36,310
is and this is probably--

707
00:31:36,310 --> 00:31:38,726
I'm going to try not to let
this devolve into my "the news

708
00:31:38,726 --> 00:31:41,164
media is out to get you"
rant, which is separate.

709
00:31:41,164 --> 00:31:42,580
But I think one
of the things that

710
00:31:42,580 --> 00:31:46,360
affects this is that there's
been this transition,

711
00:31:46,360 --> 00:31:49,360
so that things like,
for example, a missing

712
00:31:49,360 --> 00:31:51,711
kid goes from being
local news in the city

713
00:31:51,711 --> 00:31:53,710
where it happened in the
immediate surroundings,

714
00:31:53,710 --> 00:31:56,440
to national news.

715
00:31:56,440 --> 00:32:00,700
So, our perception
of how many of these

716
00:32:00,700 --> 00:32:02,950
are going on increases.

717
00:32:02,950 --> 00:32:04,450
And because we know
more about them,

718
00:32:04,450 --> 00:32:07,720
we are more familiar
with the incidents.

719
00:32:07,720 --> 00:32:11,214
And so, this increased
familiarity, again,

720
00:32:11,214 --> 00:32:12,880
makes it easier for
us to think of them.

721
00:32:12,880 --> 00:32:15,070
We judge them as being
more risky, more frequent

722
00:32:15,070 --> 00:32:15,912
occurrences.

723
00:32:21,324 --> 00:32:23,800
[INAUDIBLE]

724
00:32:23,800 --> 00:32:26,860
One more heuristic
that we tend to use,

725
00:32:26,860 --> 00:32:28,130
that I wanted to bring up--

726
00:32:28,130 --> 00:32:31,190
this is what's called the
anchoring and adjustment

727
00:32:31,190 --> 00:32:31,690
heuristic.

728
00:32:31,690 --> 00:32:35,890
So, when we're trying to
estimate the size of something

729
00:32:35,890 --> 00:32:38,140
or figure out, I don't
know, how much should it

730
00:32:38,140 --> 00:32:41,870
cost me to fly to Denver,
we'll make an approximation.

731
00:32:41,870 --> 00:32:43,690
And then as we get
new information,

732
00:32:43,690 --> 00:32:46,810
we'll adjust that approximation.

733
00:32:46,810 --> 00:32:49,360
But what often
tends to happen, is

734
00:32:49,360 --> 00:32:53,350
that that first number,
that anchor number, we

735
00:32:53,350 --> 00:32:54,472
kind of treat it too much.

736
00:32:54,472 --> 00:32:55,930
Even if it was just
a random guess,

737
00:32:55,930 --> 00:32:59,560
even if it had nothing to do,
whatsoever, with whatever we're

738
00:32:59,560 --> 00:33:06,609
trying to estimate, it tends
to kind of affect the changes

739
00:33:06,609 --> 00:33:07,150
that we make.

740
00:33:07,150 --> 00:33:09,910
We won't make big enough changes
away from that original anchor.

741
00:33:15,616 --> 00:33:16,990
So, one example
of this would be,

742
00:33:16,990 --> 00:33:20,340
suppose I want to fly to
Denver for this weekend-- which

743
00:33:20,340 --> 00:33:22,917
I do, but I'm not going
to because I'm poor

744
00:33:22,917 --> 00:33:24,750
and Denver is expensive
to get to because it

745
00:33:24,750 --> 00:33:26,910
takes a plane to it.

746
00:33:26,910 --> 00:33:29,790
Colorado.

747
00:33:29,790 --> 00:33:35,620
And suppose I said, OK, I
think flying to Denver--

748
00:33:35,620 --> 00:33:37,620
Suppose I set aside some
money to fly to Denver,

749
00:33:37,620 --> 00:33:40,110
and I start looking
for airplane tickets.

750
00:33:40,110 --> 00:33:45,420
And the first airplane tickets
I look at are like $1,200.

751
00:33:45,420 --> 00:33:48,120
And I go, oh, boy,
no not so much.

752
00:33:48,120 --> 00:33:52,290
And I start doing some hunting,
some looking for cheaper fares.

753
00:33:52,290 --> 00:33:55,440
And, eventually, I hunt
and I find smaller fares

754
00:33:55,440 --> 00:33:57,060
and smaller fares.

755
00:33:57,060 --> 00:34:00,420
And by the time I find
fares that are $700,

756
00:34:00,420 --> 00:34:03,220
it looks like a
really good deal.

757
00:34:03,220 --> 00:34:06,990
So, I buy them because $700 is a
lot less than what I originally

758
00:34:06,990 --> 00:34:08,190
saw.

759
00:34:08,190 --> 00:34:13,080
On the other hand,
let's still say

760
00:34:13,080 --> 00:34:16,409
if I budgeted $500
to do this, then $700

761
00:34:16,409 --> 00:34:19,020
is more than I had
anticipated to spend.

762
00:34:19,020 --> 00:34:25,010
But because I had one idea
set as this anchor of $1,200,

763
00:34:25,010 --> 00:34:27,794
then adjustments from
that feel significant,

764
00:34:27,794 --> 00:34:28,710
even when they're not.

765
00:34:28,710 --> 00:34:30,126
Even if I hadn't
still gotten down

766
00:34:30,126 --> 00:34:33,270
to what I'd planned
to spend on tickets,

767
00:34:33,270 --> 00:34:34,929
it felt like it was
a much better deal

768
00:34:34,929 --> 00:34:36,179
than what I'd originally seen.

769
00:34:36,179 --> 00:34:39,289
So, this is the anchoring
an adjustment heuristic.

770
00:34:39,289 --> 00:34:40,830
And so, what's
interesting with this,

771
00:34:40,830 --> 00:34:44,730
is that people do this even
when the information that

772
00:34:44,730 --> 00:34:48,360
seems to being set as an
anchor has nothing, whatsoever,

773
00:34:48,360 --> 00:34:51,159
to do with what they're
trying to estimate.

774
00:34:51,159 --> 00:34:58,620
So, Tversky and Kahneman asked
people to estimate percentages.

775
00:34:58,620 --> 00:35:01,050
What percentage of UN
delegates are from Africa?

776
00:35:01,050 --> 00:35:04,200
What percentage of states
in the United States

777
00:35:04,200 --> 00:35:06,810
have more than one
state university?

778
00:35:06,810 --> 00:35:09,390
What percentage of-- probably
a lot higher now than it was

779
00:35:09,390 --> 00:35:10,950
in 1974--

780
00:35:10,950 --> 00:35:13,920
what percentage
of, I don't know,

781
00:35:13,920 --> 00:35:16,260
people in the United
States live in a city?

782
00:35:16,260 --> 00:35:17,940
Various percentages.

783
00:35:17,940 --> 00:35:22,140
And they would ask the
participant this question,

784
00:35:22,140 --> 00:35:24,900
and they had a wheel, like a
mini Wheel of Fortune wheel,

785
00:35:24,900 --> 00:35:27,510
with just the numbers
from 0 to 100 arranged

786
00:35:27,510 --> 00:35:29,410
in order around the rim.

787
00:35:29,410 --> 00:35:31,860
And so they would
ask the question,

788
00:35:31,860 --> 00:35:35,100
they would spin the
wheel, and then the wheel

789
00:35:35,100 --> 00:35:37,230
would stop on something.

790
00:35:37,230 --> 00:35:39,420
And then they'd have the
participants indicate

791
00:35:39,420 --> 00:35:41,430
their answer by
moving the wheel so

792
00:35:41,430 --> 00:35:43,110
that the pointer was
on the percentage

793
00:35:43,110 --> 00:35:44,235
that they thought was true.

794
00:35:47,000 --> 00:35:47,900
We got this?

795
00:35:47,900 --> 00:35:48,500
So, they've got a wheel.

796
00:35:48,500 --> 00:35:49,590
They spin the wheel.

797
00:35:49,590 --> 00:35:52,349
So, if I ask you, what
percentage of UN delegates

798
00:35:52,349 --> 00:35:53,390
are from African nations.

799
00:35:53,390 --> 00:35:55,850
I spin the wheel,
the wheel hits 10,

800
00:35:55,850 --> 00:35:58,100
you think it's probably
higher than 10-- more than 10%

801
00:35:58,100 --> 00:35:59,780
of countries are in Africa.

802
00:35:59,780 --> 00:36:02,130
So, you're going to
move it up from there.

803
00:36:02,130 --> 00:36:04,220
And what's interesting--
what they found,

804
00:36:04,220 --> 00:36:08,150
is that this entirely
randomly chosen number--

805
00:36:08,150 --> 00:36:09,980
I spun this wheel
right in front of you,

806
00:36:09,980 --> 00:36:12,440
you saw me do it you
know it's random--

807
00:36:12,440 --> 00:36:14,630
affects the outcomes
that people give.

808
00:36:14,630 --> 00:36:19,280
You know, it's if wheel
stops on a low number, most

809
00:36:19,280 --> 00:36:22,310
of the people are moving
it up from that low number.

810
00:36:22,310 --> 00:36:25,760
The estimates that they
give are lower than

811
00:36:25,760 --> 00:36:29,330
if it lands on a high number
and people are moving it down,

812
00:36:29,330 --> 00:36:30,503
consistently.

813
00:36:30,503 --> 00:36:32,435
AUDIENCE: Is there
a set interval?

814
00:36:36,590 --> 00:36:38,390
ABBY NOYCE: I don't
know if they compared--

815
00:36:38,390 --> 00:36:39,765
the way I would
probably do this,

816
00:36:39,765 --> 00:36:41,270
is I would compare
a group where I

817
00:36:41,270 --> 00:36:42,894
didn't have any
wheel-spinning nonsense

818
00:36:42,894 --> 00:36:45,230
and just had them
guess a percentage,

819
00:36:45,230 --> 00:36:47,030
and took the average
of those, so that I

820
00:36:47,030 --> 00:36:50,120
could see, what does the
general population believe

821
00:36:50,120 --> 00:36:52,505
the number of African
nations in the UN

822
00:36:52,505 --> 00:36:54,590
to be, and compare from there.

823
00:36:54,590 --> 00:36:58,400
But I couldn't find any data
that they had done this.

824
00:36:58,400 --> 00:37:01,970
The example that the book
in which I was reading

825
00:37:01,970 --> 00:37:04,550
about this experiment gave--

826
00:37:04,550 --> 00:37:07,370
they gave an example of
the wheel landing on 10.

827
00:37:07,370 --> 00:37:10,352
Participants would
give an answer of 25%.

828
00:37:10,352 --> 00:37:14,950
The wheel landing on 60,
people had to move it down.

829
00:37:14,950 --> 00:37:16,740
They'd give an answer
of more like 45%.

830
00:37:16,740 --> 00:37:19,190
So, a reasonably big range
between these two but I

831
00:37:19,190 --> 00:37:21,868
don't have a good
solid answer for it.

832
00:37:21,868 --> 00:37:23,365
AUDIENCE: Could
it be because they

833
00:37:23,365 --> 00:37:26,360
wouldn't want to move it so much
from where it was randomly--

834
00:37:26,360 --> 00:37:28,190
ABBY NOYCE: --because
it was so heavy.

835
00:37:28,190 --> 00:37:28,984
Well, yeah.

836
00:37:28,984 --> 00:37:30,650
That's pretty much
what we're saying, is

837
00:37:30,650 --> 00:37:34,220
that once a number has
been set on, where--

838
00:37:34,220 --> 00:37:36,800
not even consciously, but
we're reluctant to move too far

839
00:37:36,800 --> 00:37:37,730
away from it.

840
00:37:37,730 --> 00:37:39,782
So, we tend to want to
make small changes rather

841
00:37:39,782 --> 00:37:40,490
than big changes.

842
00:37:43,760 --> 00:37:46,514
It's like, suppose your
favorite ice cream place

843
00:37:46,514 --> 00:37:47,930
kept the cost of
an ice cream cone

844
00:37:47,930 --> 00:37:49,730
at $2 for years and
years and years.

845
00:37:49,730 --> 00:37:52,010
And then all of a
sudden, it went up to $4.

846
00:37:52,010 --> 00:37:53,360
You'd be like, what?

847
00:37:53,360 --> 00:37:57,252
The price of my ice cream just
doubled, I don't think so.

848
00:37:57,252 --> 00:37:58,365
AUDIENCE: Ridiculous!

849
00:37:58,365 --> 00:37:59,990
ABBY NOYCE: You'd be
ticked off, right?

850
00:37:59,990 --> 00:38:01,760
You'd probably buy
less ice cream.

851
00:38:01,760 --> 00:38:04,709
But if it went up gradually
over the same amount of time,

852
00:38:04,709 --> 00:38:06,500
like, instead of being
the same for so long

853
00:38:06,500 --> 00:38:10,130
and then going up suddenly,
if it went up by $0.25

854
00:38:10,130 --> 00:38:13,310
a year or something, you'd
be a lot less averse that.

855
00:38:13,310 --> 00:38:16,280
So, that's another example
where big changes in some kind

856
00:38:16,280 --> 00:38:18,249
of numbers are disconcerting.

857
00:38:18,249 --> 00:38:19,040
We don't like them.

858
00:38:19,040 --> 00:38:20,540
Little changes we're
pretty OK with.

859
00:38:31,961 --> 00:38:32,955
All right.

860
00:38:35,940 --> 00:38:38,280
So, kind of take-away
things from this.

861
00:38:38,280 --> 00:38:41,010
So, decisions that we make,
judgments that we make,

862
00:38:41,010 --> 00:38:43,860
all of this are influenced
by a lot of things

863
00:38:43,860 --> 00:38:45,890
that we aren't
directly aware of.

864
00:38:45,890 --> 00:38:46,980
One of the things that's
interesting in all

865
00:38:46,980 --> 00:38:48,896
of this heuristic study,
is when you point out

866
00:38:48,896 --> 00:38:51,370
to people that people tend
to use these heuristics,

867
00:38:51,370 --> 00:38:54,094
that they tend to make these
kinds of statistical decisions,

868
00:38:54,094 --> 00:38:56,010
that this is a piece of
information you should

869
00:38:56,010 --> 00:38:59,410
take into account,
they get better at it.

870
00:38:59,410 --> 00:39:01,290
So, there's at least--

871
00:39:01,290 --> 00:39:04,380
so, we use these heuristics
and kind of naively

872
00:39:04,380 --> 00:39:07,410
we don't think about what
the flaws in them are,

873
00:39:07,410 --> 00:39:10,800
but people can learn to
take that into account

874
00:39:10,800 --> 00:39:12,780
and make better decisions
and more accurate

875
00:39:12,780 --> 00:39:17,260
estimates by being aware of how
their heuristic systems work.

876
00:39:22,249 --> 00:39:22,749
OK.

877
00:39:42,740 --> 00:39:44,390
Shifting gears a
little bit, I want

878
00:39:44,390 --> 00:39:48,480
to talk about one
particular study.

879
00:39:48,480 --> 00:39:51,330
So, we were talking
a minute ago--

880
00:39:51,330 --> 00:39:54,090
for the first part of
this-- about the strategies

881
00:39:54,090 --> 00:39:56,090
that people use to make
decisions in their life.

882
00:39:58,730 --> 00:40:01,700
These guys are
looking at what some

883
00:40:01,700 --> 00:40:03,980
of the underlying
neuroscience is in this.

884
00:40:03,980 --> 00:40:06,400
And they had people
do a very simple task.

885
00:40:06,400 --> 00:40:10,160
They said, we're
going to have you

886
00:40:10,160 --> 00:40:13,370
have a button in each hand
and whenever you feel like it,

887
00:40:13,370 --> 00:40:15,320
push one of them.

888
00:40:15,320 --> 00:40:17,290
So, they had people do this.

889
00:40:17,290 --> 00:40:20,330
And they had them
sit-in an FMRI tube.

890
00:40:20,330 --> 00:40:22,910
And they wanted
to know, is there

891
00:40:22,910 --> 00:40:25,550
any way that we can measure
what kind of decision people

892
00:40:25,550 --> 00:40:29,330
are going to take before they're
aware that they've decided it?

893
00:40:29,330 --> 00:40:30,761
So, here you are.

894
00:40:30,761 --> 00:40:32,510
You're sitting in the
FMRI tube and you're

895
00:40:32,510 --> 00:40:37,310
seeing a screen which is
showing a stream of letters

896
00:40:37,310 --> 00:40:40,337
for 500 milliseconds,
for half a second, each.

897
00:40:40,337 --> 00:40:42,170
So, it'd show a letter,
show another letter,

898
00:40:42,170 --> 00:40:44,450
show another letter,
show another letter.

899
00:40:44,450 --> 00:40:45,880
And you're just
hanging out there.

900
00:40:45,880 --> 00:40:47,590
And whenever you
feel like it, you

901
00:40:47,590 --> 00:40:49,370
are asked to do two things--

902
00:40:49,370 --> 00:40:54,020
to press either the
left or the right button

903
00:40:54,020 --> 00:40:56,720
and also to make a
note of which letter

904
00:40:56,720 --> 00:41:00,290
was being shown at the moment
that you made the decision.

905
00:41:00,290 --> 00:41:03,140
So, you were asked to decide
and press right away--

906
00:41:03,140 --> 00:41:06,230
not to decide and then wait for
a minute, just make a decision,

907
00:41:06,230 --> 00:41:10,440
push the button, and notice
which letter was being shown.

908
00:41:10,440 --> 00:41:13,610
So, what they did is after
you pushed the button,

909
00:41:13,610 --> 00:41:17,120
you got a screen
and you were asked

910
00:41:17,120 --> 00:41:20,240
to indicate which
letter had been

911
00:41:20,240 --> 00:41:23,700
displayed when you had made the
decision to press the button.

912
00:41:23,700 --> 00:41:29,600
So, if I'm sitting here
watching the screens

913
00:41:29,600 --> 00:41:34,430
and in this case, when the
moment when the Q is shown,

914
00:41:34,430 --> 00:41:37,190
is when I decide to
press the button.

915
00:41:37,190 --> 00:41:40,040
And no matter how fast I go
from decision to button press--

916
00:41:40,040 --> 00:41:41,660
it might be a little bit slow--

917
00:41:41,660 --> 00:41:43,940
so what's actually
being displayed

918
00:41:43,940 --> 00:41:48,430
when the computer detects the
button press, might be the V.

919
00:41:48,430 --> 00:41:52,410
And after I press the button,
the computer's going to ask me,

920
00:41:52,410 --> 00:41:57,354
of these three previous slides,
which one was showing when you

921
00:41:57,354 --> 00:41:58,520
decided to press the button?

922
00:41:58,520 --> 00:42:01,100
And if the Q was showing when
I decided to press the button,

923
00:42:01,100 --> 00:42:02,400
I'd indicate that.

924
00:42:02,400 --> 00:42:05,882
I'd say it was the Q and
move on and do it again.

925
00:42:05,882 --> 00:42:08,090
So, is what they're asking
subjects to do make sense?

926
00:42:10,801 --> 00:42:13,050
Decide to push a button,
remember which letter it was,

927
00:42:13,050 --> 00:42:14,877
tell us which letter it was.

928
00:42:14,877 --> 00:42:16,460
And what these guys
wanted to find out

929
00:42:16,460 --> 00:42:17,270
was a couple of things.

930
00:42:17,270 --> 00:42:19,436
There's been research in
the literature for a while,

931
00:42:19,436 --> 00:42:22,490
for about 20 years
now, showing that when

932
00:42:22,490 --> 00:42:24,620
people decide to make
a motor movement,

933
00:42:24,620 --> 00:42:26,010
like pressing a button--

934
00:42:26,010 --> 00:42:30,170
which is a motor response;
it's a very small one compared

935
00:42:30,170 --> 00:42:32,940
to a lot of other motor things--

936
00:42:32,940 --> 00:42:36,886
then at least some period
in advance of that,

937
00:42:36,886 --> 00:42:38,510
maybe half a second
in advance of that,

938
00:42:38,510 --> 00:42:41,650
there's activity
in the brain that

939
00:42:41,650 --> 00:42:43,400
seems to indicate that
a decision has been

940
00:42:43,400 --> 00:42:45,852
made before the
participant is aware

941
00:42:45,852 --> 00:42:47,060
that they've made a decision.

942
00:42:47,060 --> 00:42:50,510
Before you say, I've decided,
before you can think that,

943
00:42:50,510 --> 00:42:54,650
your brain apparently already
knows that you've decided.

944
00:42:54,650 --> 00:42:55,900
Creepy, yeah.

945
00:42:55,900 --> 00:42:59,570
So, these guys said, OK,
the previous rounds of this

946
00:42:59,570 --> 00:43:03,170
had only asked people
to push a single button.

947
00:43:03,170 --> 00:43:04,910
They didn't have a choice.

948
00:43:04,910 --> 00:43:07,640
So, we want to see if we can
predict the choice that people

949
00:43:07,640 --> 00:43:07,810
make.

950
00:43:07,810 --> 00:43:09,018
Here they've got two options.

951
00:43:09,018 --> 00:43:11,060
You can push either the
button in your left hand

952
00:43:11,060 --> 00:43:12,714
or the button in
your right hand.

953
00:43:12,714 --> 00:43:14,630
And they want to see if
they can predict which

954
00:43:14,630 --> 00:43:17,960
choice people make
before the people are

955
00:43:17,960 --> 00:43:20,700
aware that they've done it.

956
00:43:20,700 --> 00:43:23,900
So, what did they find?

957
00:43:23,900 --> 00:43:25,070
Look, pretty brain pictures.

958
00:43:25,070 --> 00:43:27,010
OK.

959
00:43:27,010 --> 00:43:29,870
So, they had people
do this in an FMRI.

960
00:43:29,870 --> 00:43:31,980
They looked at their
patterns of brain activity

961
00:43:31,980 --> 00:43:35,570
in particular regions
over a temporal scale.

962
00:43:35,570 --> 00:43:37,200
So, each of these--
as you can see,

963
00:43:37,200 --> 00:43:38,950
there's different
particular brain regions

964
00:43:38,950 --> 00:43:41,360
that are being graphed here.

965
00:43:41,360 --> 00:43:45,410
each of these graphs has
time along the bottom axis

966
00:43:45,410 --> 00:43:47,060
in seconds.

967
00:43:47,060 --> 00:43:49,310
This red line is
the point at which

968
00:43:49,310 --> 00:43:52,820
subjects reported
that they had decided

969
00:43:52,820 --> 00:43:55,880
to make a decision using that--

970
00:43:55,880 --> 00:43:58,170
which letter was showing
when you decided--

971
00:43:58,170 --> 00:43:59,130
that method.

972
00:43:59,130 --> 00:44:05,240
And then this is showing how
accurately the vertical axis

973
00:44:05,240 --> 00:44:05,770
is showing.

974
00:44:05,770 --> 00:44:07,520
So, they looked at the
particular patterns

975
00:44:07,520 --> 00:44:09,560
of activation in each
of these regions.

976
00:44:09,560 --> 00:44:14,390
And they said, can we find a
pattern here that predicts--

977
00:44:14,390 --> 00:44:16,490
Using the information
in this brain region,

978
00:44:16,490 --> 00:44:20,180
at this point in time, can we
predict which choice you made?

979
00:44:20,180 --> 00:44:22,670
So, this is the 50% line here.

980
00:44:22,670 --> 00:44:24,580
So, that's chance.

981
00:44:24,580 --> 00:44:26,330
And then anything up
above that-- so, this

982
00:44:26,330 --> 00:44:28,040
would be like 75% accuracy.

983
00:44:28,040 --> 00:44:31,010
That's pretty good.

984
00:44:31,010 --> 00:44:32,960
So, the filled-in
black dots are places

985
00:44:32,960 --> 00:44:35,310
where they're significantly
better than chance

986
00:44:35,310 --> 00:44:37,847
at determining which
choice was made.

987
00:44:37,847 --> 00:44:39,680
The white ones are
places where they're not,

988
00:44:39,680 --> 00:44:42,700
where it's just
the same as chance.

989
00:44:42,700 --> 00:44:44,100
So, what did they find out?

990
00:44:44,100 --> 00:44:47,590
Well, they found out
that after subjects

991
00:44:47,590 --> 00:44:50,620
made this conscious decision
to push the button, then

992
00:44:50,620 --> 00:44:54,315
all of these motor cortex areas
are really good at predicting

993
00:44:54,315 --> 00:44:56,440
whether you're going to
press it with the left hand

994
00:44:56,440 --> 00:44:58,580
or the right hand, which
isn't too surprising.

995
00:44:58,580 --> 00:45:01,604
Motor cortex is involved in all
of these kinds of decisions.

996
00:45:04,330 --> 00:45:08,080
Left motor cortex, right motor
cortex, the supplemental motor

997
00:45:08,080 --> 00:45:10,030
area and the
pre-supplemental motor area,

998
00:45:10,030 --> 00:45:13,030
which are, again, as you guys
might remember from last week,

999
00:45:13,030 --> 00:45:16,270
are areas that seem
to deal with planning

1000
00:45:16,270 --> 00:45:20,320
activities and goal-directed
motion and stuff.

1001
00:45:20,320 --> 00:45:21,790
And so, all of
these are areas that

1002
00:45:21,790 --> 00:45:25,240
can identify what choice you're
making after you've consciously

1003
00:45:25,240 --> 00:45:26,350
made it.

1004
00:45:26,350 --> 00:45:29,650
What's more interesting
is these ones,

1005
00:45:29,650 --> 00:45:32,320
which show some
amount of ability

1006
00:45:32,320 --> 00:45:35,200
to predict what choice you
make before you're consciously

1007
00:45:35,200 --> 00:45:37,270
aware that you made it.

1008
00:45:37,270 --> 00:45:40,200
So, even way out
here-- look at this.

1009
00:45:40,200 --> 00:45:43,840
This is frontopolar
cortex, lateral and medial.

1010
00:45:43,840 --> 00:45:46,990
So, way out here.

1011
00:45:46,990 --> 00:45:49,840
Eight seconds
before participants

1012
00:45:49,840 --> 00:45:54,370
report that they've made
a decision, just looking

1013
00:45:54,370 --> 00:45:56,830
at activity in the
frontopolar cortex,

1014
00:45:56,830 --> 00:46:00,302
these researchers could
predict with 60% accuracy

1015
00:46:00,302 --> 00:46:02,260
whether you were going
to push the right button

1016
00:46:02,260 --> 00:46:03,490
or the left button.

1017
00:46:03,490 --> 00:46:06,610
Eight seconds-- that's like
forever in neuroscience terms--

1018
00:46:06,610 --> 00:46:10,270
before you know that
you made that decision.

1019
00:46:10,270 --> 00:46:13,900
I can tell you which decision
you're going to make.

1020
00:46:13,900 --> 00:46:15,742
AUDIENCE: That's creepy.

1021
00:46:15,742 --> 00:46:16,450
ABBY NOYCE: Yeah.

1022
00:46:19,330 --> 00:46:23,791
Slightly sooner in
posterior cingulate cortex.

1023
00:46:23,791 --> 00:46:28,540
It's not quite as good
in medial frontopolar.

1024
00:46:28,540 --> 00:46:31,090
But they said if they combined
the information from all three

1025
00:46:31,090 --> 00:46:33,470
of these areas then their
accuracy actually went up.

1026
00:46:33,470 --> 00:46:36,730
It went up above
that 60% to like 75%,

1027
00:46:36,730 --> 00:46:38,920
3/4 accuracy at predicting
whether subjects

1028
00:46:38,920 --> 00:46:43,000
were going to push the
left or the right button

1029
00:46:43,000 --> 00:46:48,006
before they thought they'd made
that decision, which is cool.

1030
00:46:48,006 --> 00:46:50,754
AUDIENCE: Well, couldn't
they have a 50% chance of

1031
00:46:50,754 --> 00:46:52,050
[INAUDIBLE]?

1032
00:46:52,050 --> 00:46:54,600
ABBY NOYCE: Right, so
it's getting things

1033
00:46:54,600 --> 00:46:56,040
above 50% that are relevant.

1034
00:46:58,680 --> 00:47:03,900
So, again, the filled- incircles
on the graphs are time points

1035
00:47:03,900 --> 00:47:07,950
at which the difference between
their accuracy and chance is

1036
00:47:07,950 --> 00:47:10,664
statistically
significant, which--

1037
00:47:10,664 --> 00:47:12,330
it's not just random
that they got this,

1038
00:47:12,330 --> 00:47:13,950
that it actually
means something.

1039
00:47:13,950 --> 00:47:18,300
Whereas the white ones, where
it's very close to chance,

1040
00:47:18,300 --> 00:47:19,692
are not significant.

1041
00:47:27,080 --> 00:47:32,706
Questions, comments, creepy?

1042
00:47:32,706 --> 00:47:35,081
AUDIENCE: I think there was
an article in, like, The Wall

1043
00:47:35,081 --> 00:47:36,330
Street Journal about that.

1044
00:47:36,330 --> 00:47:37,913
ABBY NOYCE: Yeah,
there's been results

1045
00:47:37,913 --> 00:47:40,170
like this for quite a while.

1046
00:47:40,170 --> 00:47:43,574
It's kind of an ongoing
topic of research.

1047
00:47:43,574 --> 00:47:45,240
Like I said, the first
people doing this

1048
00:47:45,240 --> 00:47:47,800
were doing it in the
'80s, in the mid '80s.

1049
00:47:47,800 --> 00:47:50,460
But as the methods get
better and as new techniques

1050
00:47:50,460 --> 00:47:51,132
get looked at--

1051
00:47:51,132 --> 00:47:53,340
so I think this is the first
one like this where they

1052
00:47:53,340 --> 00:47:56,240
have the two-choice option.

1053
00:47:56,240 --> 00:47:58,710
This is a paper from like May
of this year, I want to say.

1054
00:47:58,710 --> 00:47:59,370
This is new.

1055
00:47:59,370 --> 00:48:01,082
This is shiny new.

1056
00:48:01,082 --> 00:48:04,554
AUDIENCE: What happens
if you visualize

1057
00:48:04,554 --> 00:48:08,040
pressing with your left hand
as you do it with your right?

1058
00:48:08,040 --> 00:48:09,844
ABBY NOYCE: I have no idea.

1059
00:48:09,844 --> 00:48:11,260
You'd probably
mess up their data.

1060
00:48:14,820 --> 00:48:17,549
Only do this to researchers
with whom you are not friends,

1061
00:48:17,549 --> 00:48:19,840
at least without telling them
that's what you're doing.

1062
00:48:22,720 --> 00:48:32,110
AUDIENCE: That's a good
test to do [INAUDIBLE]

1063
00:48:32,110 --> 00:48:33,540
ABBY NOYCE: Maybe.

1064
00:48:33,540 --> 00:48:34,860
Yeah.

1065
00:48:34,860 --> 00:48:36,630
And on the other hand,
we know that when

1066
00:48:36,630 --> 00:48:39,270
you visualize doing something,
at least to some extent,

1067
00:48:39,270 --> 00:48:41,436
the pieces of cortex that
would actually be involved

1068
00:48:41,436 --> 00:48:43,480
in doing it, are active.

1069
00:48:43,480 --> 00:48:45,732
So, if I ask you to visualize
a banana, the same parts

1070
00:48:45,732 --> 00:48:47,190
of your cortex that
would be active

1071
00:48:47,190 --> 00:48:49,920
if you were actually looking
at a, banana are active.

1072
00:48:49,920 --> 00:48:52,732
Or if I ask you to imagine
standing up and doing

1073
00:48:52,732 --> 00:48:54,690
a jumping jack, then the
pieces of motor cortex

1074
00:48:54,690 --> 00:48:56,800
that would be involved
in that, are active.

1075
00:49:01,940 --> 00:49:03,120
All right.

1076
00:49:03,120 --> 00:49:04,700
So, this brings
us around to some

1077
00:49:04,700 --> 00:49:09,850
of the big philosophical
questions in this field.

1078
00:49:09,850 --> 00:49:11,900
So, we've been talking
pretty much all

1079
00:49:11,900 --> 00:49:14,270
through this course
with this underlying

1080
00:49:14,270 --> 00:49:19,327
assumption that our subjective
experience of the world--

1081
00:49:19,327 --> 00:49:21,410
our perceptions, our
understandings, our memories,

1082
00:49:21,410 --> 00:49:23,200
all this--

1083
00:49:23,200 --> 00:49:26,270
can be traced directly
to physical stuff

1084
00:49:26,270 --> 00:49:28,700
going on in the brain.

1085
00:49:28,700 --> 00:49:33,020
And we're pretty sure that
physical stuff has to be

1086
00:49:33,020 --> 00:49:35,120
caused by other physical stuff.

1087
00:49:35,120 --> 00:49:38,660
You can't have
something that is only

1088
00:49:38,660 --> 00:49:41,030
mental without a
physiological basis that

1089
00:49:41,030 --> 00:49:42,410
can then cause physical stuff.

1090
00:49:42,410 --> 00:49:43,977
So, physical
changes in the brain

1091
00:49:43,977 --> 00:49:45,810
have to be caused by
other physical changes.

1092
00:49:45,810 --> 00:49:48,200
Some of the stuff where
we've looked at the chemistry

1093
00:49:48,200 --> 00:49:50,930
or the cellular-level
changes underlying things--

1094
00:49:50,930 --> 00:49:53,570
that's this idea that
physical changes are caused

1095
00:49:53,570 --> 00:49:56,629
by other physical events.

1096
00:49:56,629 --> 00:49:58,670
And this starts bringing
us into all of these fun

1097
00:49:58,670 --> 00:50:01,070
questions about free will.

1098
00:50:01,070 --> 00:50:05,090
So, if you decide
to do something

1099
00:50:05,090 --> 00:50:07,432
based on the state your
brain is in right now

1100
00:50:07,432 --> 00:50:09,890
and the state your brain is in
right now can be traced back

1101
00:50:09,890 --> 00:50:14,390
to every state your brain
has always been in then

1102
00:50:14,390 --> 00:50:17,180
do you get to make
a free decision

1103
00:50:17,180 --> 00:50:20,720
or are all of your
decisions predetermined

1104
00:50:20,720 --> 00:50:22,430
by the experiences
you've previously had?

1105
00:50:29,248 --> 00:50:31,683
AUDIENCE: This class has made
me stop trusting everything.

1106
00:50:31,683 --> 00:50:34,610
AUDIENCE: That's a
logical argument.

1107
00:50:34,610 --> 00:50:36,539
ABBY NOYCE: I'm OK with
that as an outcome.

1108
00:50:36,539 --> 00:50:37,580
AUDIENCE: I would say no.

1109
00:50:37,580 --> 00:50:42,060
But I would say both
answers could be arguments

1110
00:50:42,060 --> 00:50:49,040
because if you said yes,
you could think that it's--

1111
00:50:49,040 --> 00:50:50,440
I don't know.

1112
00:50:50,440 --> 00:50:52,460
ABBY NOYCE: So
here's a question.

1113
00:50:52,460 --> 00:50:57,250
Suppose you are on the jury in
a criminal case for a guy who's

1114
00:50:57,250 --> 00:51:00,910
been accused of beating
up a bunch of people,

1115
00:51:00,910 --> 00:51:04,920
like a serial assault, like
assault on a number of people,

1116
00:51:04,920 --> 00:51:07,030
has injured people.

1117
00:51:07,030 --> 00:51:12,580
And what his attorney
says is that this guy--

1118
00:51:12,580 --> 00:51:15,490
you get lots of witnesses saying
that up until five years ago

1119
00:51:15,490 --> 00:51:17,770
this guy was gentle
as a lamb, never hurt

1120
00:51:17,770 --> 00:51:21,130
anybody in his entire life.

1121
00:51:21,130 --> 00:51:24,097
And his attorney says
that he got a brain tumor

1122
00:51:24,097 --> 00:51:25,930
and he now has a brain
tumor that has caused

1123
00:51:25,930 --> 00:51:30,190
this change in his behavior.

1124
00:51:30,190 --> 00:51:34,080
Is this guy responsible
for what he has done?

1125
00:51:34,080 --> 00:51:35,097
Should he be convicted?

1126
00:51:35,097 --> 00:51:35,680
AUDIENCE: Yes.

1127
00:51:35,680 --> 00:51:36,115
AUDIENCE: No.

1128
00:51:36,115 --> 00:51:36,698
AUDIENCE: Yes.

1129
00:51:36,698 --> 00:51:37,247
AUDIENCE: No.

1130
00:51:37,247 --> 00:51:37,830
AUDIENCE: Yes.

1131
00:51:49,582 --> 00:51:51,706
AUDIENCE: --because
he got a brain tumor,

1132
00:51:51,706 --> 00:51:54,008
then it's not his
fault. He didn't

1133
00:51:54,008 --> 00:51:55,499
choose to get a brain tumor.

1134
00:51:55,499 --> 00:51:57,487
AUDIENCE: Yeah, but he
still hurt other people.

1135
00:52:03,645 --> 00:52:04,270
ABBY NOYCE: OK.

1136
00:52:08,860 --> 00:52:12,887
What happens if we say, OK, you
have to be treated for this.

1137
00:52:12,887 --> 00:52:15,220
We remove the brain tumor,
and this guy goes back to his

1138
00:52:15,220 --> 00:52:20,140
previous exemplary,
gentle-as-a-lamb lifestyle?

1139
00:52:20,140 --> 00:52:22,764
We postponed the trial
for a year for him

1140
00:52:22,764 --> 00:52:23,680
to get this treatment.

1141
00:52:23,680 --> 00:52:24,530
Then what?

1142
00:52:24,530 --> 00:52:26,030
You going to throw
this guy in jail?

1143
00:52:26,030 --> 00:52:26,650
He's got kids.

1144
00:52:26,650 --> 00:52:27,527
He's got family.

1145
00:52:27,527 --> 00:52:28,360
He's got a good job.

1146
00:52:28,360 --> 00:52:31,855
He's a staple of his community.

1147
00:52:31,855 --> 00:52:34,105
But he beat up 10 people and
put them in the hospital.

1148
00:52:38,070 --> 00:52:39,890
These aren't easy questions.

1149
00:52:39,890 --> 00:52:42,160
And this is kind of one of
the more clear-cut cases.

1150
00:52:42,160 --> 00:52:48,640
But we're also starting to see--

1151
00:52:48,640 --> 00:52:51,610
and I think we're going
to see more of this--

1152
00:52:51,610 --> 00:52:52,810
for example, OK.

1153
00:52:52,810 --> 00:52:55,070
So, that's an extreme case.

1154
00:52:55,070 --> 00:52:56,230
What about

1155
00:52:56,230 --> 00:52:58,014
AUDIENCE: People
who plead insanity

1156
00:52:58,014 --> 00:53:01,210
after doing something
really stupid?

1157
00:53:01,210 --> 00:53:03,460
ABBY NOYCE: Insanity plea
is an interesting thing.

1158
00:53:03,460 --> 00:53:09,190
Legally it's got to be based
on a professional's judgment

1159
00:53:09,190 --> 00:53:11,470
that the person did not know
that what they were doing

1160
00:53:11,470 --> 00:53:12,886
was wrong at the
time they did it.

1161
00:53:18,436 --> 00:53:19,810
So, the classic
good case of this

1162
00:53:19,810 --> 00:53:22,300
is, for example, somebody who
is severely mentally retarded

1163
00:53:22,300 --> 00:53:26,140
and doesn't know that it's
wrong to take a candy bar,

1164
00:53:26,140 --> 00:53:28,300
for an example of this.

1165
00:53:28,300 --> 00:53:30,248
AUDIENCE: But what if
somebody's a sociopath,

1166
00:53:30,248 --> 00:53:32,690
who doesn't realize that
it's wrong to kill people?

1167
00:53:32,690 --> 00:53:36,725
ABBY NOYCE: So, that's where
you don't send this person

1168
00:53:36,725 --> 00:53:37,600
to the death penalty.

1169
00:53:37,600 --> 00:53:39,430
You send them to
involuntary committal.

1170
00:53:39,430 --> 00:53:41,200
You move them--
so, the state has

1171
00:53:41,200 --> 00:53:46,910
the right to-- if somebody
is a danger to others,

1172
00:53:46,910 --> 00:53:49,990
the state has the right
to get a judge to say,

1173
00:53:49,990 --> 00:53:52,450
this person can be locked up.

1174
00:53:52,450 --> 00:53:55,180
This person can be confined
to a treatment center,

1175
00:53:55,180 --> 00:53:57,310
until they are no longer
such a danger, which

1176
00:53:57,310 --> 00:53:59,922
may be indefinite.

1177
00:54:08,900 --> 00:54:10,430
So, what about
something that's kind

1178
00:54:10,430 --> 00:54:14,240
of like the brain tumor issue.

1179
00:54:14,240 --> 00:54:17,889
So, we know that, for
example, in a lot of ways,

1180
00:54:17,889 --> 00:54:19,430
experiences in early
childhood really

1181
00:54:19,430 --> 00:54:22,475
strongly influence a lot
of how your brain develops.

1182
00:54:25,070 --> 00:54:28,070
For example, we
know that kids who

1183
00:54:28,070 --> 00:54:33,560
are abused as little
kids tend to grow up

1184
00:54:33,560 --> 00:54:36,740
to be more aggressive, to
have a hard time forming

1185
00:54:36,740 --> 00:54:39,710
relationships, to
have, in general,

1186
00:54:39,710 --> 00:54:42,620
have more issues, all
around, than kids who grew up

1187
00:54:42,620 --> 00:54:46,240
in stable, healthy families.

1188
00:54:46,240 --> 00:54:52,770
So, what happens when you
have some young adult who

1189
00:54:52,770 --> 00:54:57,330
does something violent,
who hurt somebody

1190
00:54:57,330 --> 00:55:02,010
and the case that their attorney
makes is he couldn't help it.

1191
00:55:02,010 --> 00:55:04,960
His brain was wired to
do this by his past.

1192
00:55:04,960 --> 00:55:07,800
Where people start saying
that you lose-- that because

1193
00:55:07,800 --> 00:55:11,160
of how our brains are shaped
by experience that happens

1194
00:55:11,160 --> 00:55:15,960
to them, and then our
future behavior is defined

1195
00:55:15,960 --> 00:55:18,870
by these experiences,
how much control

1196
00:55:18,870 --> 00:55:22,980
does one really have
over what one does?

1197
00:55:22,980 --> 00:55:24,480
It starts getting kind of iffy.

1198
00:55:24,480 --> 00:55:26,796
And it's kind of dark to
think about, in a lot of ways.

1199
00:55:26,796 --> 00:55:28,170
AUDIENCE: But
then, wouldn't that

1200
00:55:28,170 --> 00:55:30,794
mean that no one's responsible
for anything that they do?

1201
00:55:30,794 --> 00:55:31,460
ABBY NOYCE: Yes.

1202
00:55:31,460 --> 00:55:35,246
AUDIENCE: They can go around
killing everybody, knowing--

1203
00:55:35,246 --> 00:55:37,620
ABBY NOYCE: This is the slippery
slope argument for this,

1204
00:55:37,620 --> 00:55:42,000
is that it's definitely possible
to start out thinking about

1205
00:55:42,000 --> 00:55:44,820
this and come to the conclusion
that there is no such thing

1206
00:55:44,820 --> 00:55:47,850
as personal responsibility, that
you cannot be held responsible

1207
00:55:47,850 --> 00:55:50,790
for anything you do because it's
all just your brain doing it

1208
00:55:50,790 --> 00:55:53,310
and it's all defined
by personal experience,

1209
00:55:53,310 --> 00:55:55,380
previous experience.

1210
00:55:55,380 --> 00:56:02,020
It's all mechanical, that there
isn't a choice option involved.

1211
00:56:02,020 --> 00:56:05,625
AUDIENCE: That's why I think
that, legally, they should not

1212
00:56:05,625 --> 00:56:06,632
take into account--

1213
00:56:06,632 --> 00:56:08,840
they should assume that
everyone does have free will.

1214
00:56:08,840 --> 00:56:12,330
Because otherwise,
it wouldn't work.

1215
00:56:12,330 --> 00:56:15,697
Nothing would actually work.

1216
00:56:15,697 --> 00:56:17,530
ABBY NOYCE: Yes, this
is an excellent point.

1217
00:56:17,530 --> 00:56:18,900
That your assumption
of free will-- not

1218
00:56:18,900 --> 00:56:20,654
just for the legal
system, but for, like,

1219
00:56:20,654 --> 00:56:22,320
getting out of bed
in the morning, dude.

1220
00:56:22,320 --> 00:56:24,195
You've got to believe
that you have free will

1221
00:56:24,195 --> 00:56:27,080
or it all falls apart.

1222
00:56:27,080 --> 00:56:30,020
I don't know.

1223
00:56:30,020 --> 00:56:31,490
AUDIENCE: --slave to your brain.

1224
00:56:31,490 --> 00:56:32,480
AUDIENCE: --robots.

1225
00:56:32,480 --> 00:56:32,810
ABBY NOYCE: So,
what's the difference

1226
00:56:32,810 --> 00:56:34,638
between you and your brain?

1227
00:56:34,638 --> 00:56:36,195
AUDIENCE: I don't know.

1228
00:56:36,195 --> 00:56:37,570
ABBY NOYCE: Is
there a difference

1229
00:56:37,570 --> 00:56:38,694
between you and your brain?

1230
00:56:38,694 --> 00:56:40,194
AUDIENCE: It's kind of squishy.

1231
00:56:40,194 --> 00:56:41,610
ABBY NOYCE: It's
a good thing it's

1232
00:56:41,610 --> 00:56:43,510
inside that nice solid skull.

1233
00:56:43,510 --> 00:56:46,330
Is there a difference?

1234
00:56:46,330 --> 00:56:48,340
Part of the big question
that comes down here--

1235
00:56:48,340 --> 00:56:53,230
is there a difference between
your self and your brain?

1236
00:56:53,230 --> 00:56:57,130
Is that a distinction
that we can make validly?