How good are you at calculating risk? - Gerd Gigerenzer

645,903 views ใƒป 2020-02-25

TED-Ed


ืื ื ืœื—ืฅ ืคืขืžื™ื™ื ืขืœ ื”ื›ืชื•ื‘ื™ื•ืช ื‘ืื ื’ืœื™ืช ืœืžื˜ื” ื›ื“ื™ ืœื”ืคืขื™ืœ ืืช ื”ืกืจื˜ื•ืŸ.

ืชืจื’ื•ื: Ido Dekkers ืขืจื™ื›ื”: Sigal Tifferet
00:06
A new drug reduces the risk of heart attacks by 40%.
0
6370
4350
ืชืจื•ืคื” ื—ื“ืฉื” ืžืคื—ื™ืชื” ืกื™ื›ื•ืŸ ืœื”ืชืงืคื™ ืœื‘ ื‘-40%.
00:10
Shark attacks are up by a factor of two.
1
10720
3360
ื”ืชืงืคื•ืช ื›ืจื™ืฉื™ื ืขืœื• ืคื™ ืฉื ื™ื™ื.
00:14
Drinking a liter of soda per day doubles your chance of developing cancer.
2
14080
4866
ืฉืชื™ื™ืช ืœื™ื˜ืจ ืžืฉืงืื•ืช ืžื•ื’ื–ื™ื ื‘ื™ื•ื ืžื›ืคื™ืœื” ืืช ื”ืกื™ื›ื•ืŸ ืœืกืจื˜ืŸ.
00:18
These are all examples of relative risk,
3
18946
3330
ื›ืœ ืืœื” ื“ื•ื’ืžืื•ืช ืœืกื™ื›ื•ื ื™ื ื™ื—ืกื™ื™ื,
00:22
a common way risk is presented in news articles.
4
22276
3894
ื“ืจืš ื ืคื•ืฆื” ื‘ื” ืกื™ื›ื•ื ื™ื ืžื•ืฆื’ื™ื ื‘ื—ื“ืฉื•ืช.
00:26
Risk evaluation is a complicated tangle of statistical thinking
5
26170
4140
ื”ืขืจื›ืช ืกื™ื›ื•ื ื™ื ื”ื™ื ืกื‘ืš ืžื•ืจื›ื‘ ืฉืœ ื—ืฉื™ื‘ื” ืกื˜ื˜ื™ืกื˜ื™ืช
00:30
and personal preference.
6
30310
1688
ื•ื”ืขื“ืคื•ืช ืื™ืฉื™ื•ืช.
00:31
One common stumbling block is the difference between
7
31998
2851
ืžืฉื•ื›ื” ืื—ืช ื”ื™ื ื”ื”ื‘ื“ืœื™ื
00:34
relative risks like these and what are called absolute risks.
8
34849
4948
ื‘ื™ืŸ ืกื™ื›ื•ื ื™ื ื™ื—ืกื™ื™ื ื›ืคื™ ืฉื”ืฆื’ื ื•, ืœื‘ื™ืŸ ืžื” ืฉื ืงืจื ืกื™ื›ื•ื ื™ื ืื‘ืกื•ืœื•ื˜ื™ื™ื.
00:39
Risk is the likelihood that an event will occur.
9
39797
2910
ืกื™ื›ื•ืŸ ื”ื•ื ื”ื”ืกืชื‘ืจื•ืช ืฉืืจื•ืข ื™ืงืจื”.
00:42
It can be expressed as either a percentageโ€”
10
42707
2413
ืืคืฉืจ ืœื”ืฆื™ื’ื• ื‘ืื—ื•ื–ื™ื --
00:45
for example, that heart attacks occur in 11% of men
11
45120
3590
ืœื“ื•ื’ืžื”, ืฉื”ืชืงืคื™ ืœื‘ ืงื•ืจื™ื ืœ-11% ืžื”ื’ื‘ืจื™ื
00:48
between the ages of 60 and 79โ€”
12
48710
2750
ื‘ื™ืŸ ื’ื™ืœ 60 ืœ-79 --
00:51
or as a rateโ€” that one in two million divers along Australiaโ€™s western coast
13
51460
5813
ืื• ื›ื™ื—ืก -- ืฉืื—ื“ ืžืชื•ืš ืฉื ื™ ืžื™ืœื™ื•ืŸ ืฆื•ืœืœื ื™ื ื‘ื—ื•ืคื™ ืื•ืกื˜ืจืœื™ื”
00:57
will suffer a fatal shark bite each year.
14
57273
3170
ื™ื•ืชืงืฃ ื‘ืื•ืคืŸ ืงื˜ืœื ื™ ืขืœ-ื™ื“ื™ ื›ืจื™ืฉ, ื›ืœ ืฉื ื”.
01:00
These numbers express the absolute risk of heart attacks
15
60443
3563
ื”ืžืกืคืจื™ื ื”ืืœื” ืžื™ื™ืฆื’ื™ื ืืช ื”ืกื™ื›ื•ืŸ ื”ืื‘ืกื•ืœื•ื˜ื™ ืœื”ืชืงืคื™ ืœื‘
01:04
and shark attacks in these groups.
16
64006
2390
ื•ื”ืชืงืคื•ืช ื›ืจื™ืฉื™ื ื‘ืงื‘ื•ืฆื•ืช ื”ืืœื•.
01:06
Changes in risk can be expressed in relative or absolute terms.
17
66396
5159
ืฉื™ื ื•ื™ื™ื ื‘ืกื™ื›ื•ืŸ ื™ื›ื•ืœื™ื ืœื”ื™ื•ืช ืžื•ืฆื’ื™ื ื‘ืžื•ื ื—ื™ื ื™ื—ืกื™ื™ื ืื• ืื‘ืกื•ืœื•ื˜ื™ื™ื.
01:11
For example, a review in 2009 found that mammography screenings
18
71555
4755
ืœื“ื•ื’ืžื”, ืžืืžืจ ืกืงื™ืจื” ื‘-2009 ื’ื™ืœื” ืฉื‘ื“ื™ืงื•ืช ืžืžื•ื’ืจืคื™ื”
01:16
reduced the number of breast cancer deaths from five women in one thousand to four.
19
76310
6266
ืžืคื—ื™ืชื•ืช ืืช ื”ืžื•ื•ืช ืžืกืจื˜ืŸ ื”ืฉื“ ืžื—ืžืฉ ื ืฉื™ื ืœืืœืฃ, ืœืืจื‘ืข.
01:22
The absolute risk reduction was about .1%.
20
82576
3620
ื”ื”ืคื—ืชื” ื”ืื‘ืกื•ืœื•ื˜ื™ืช ื‘ืกื™ื›ื•ืŸ ื”ื™ืชื” ื‘ืขืจืš 0.1%.
01:26
But the relative risk reduction from 5 cases of cancer mortality to four
21
86196
4700
ืื‘ืœ ื”ื”ืคื—ืชื” ื‘ืกื™ื›ื•ืŸ ื”ื™ื—ืกื™ ืž-5 ืžืงืจื™ื ืฉืœ ืžื•ื•ืช ืžืกืจื˜ืŸ ืœ-4
01:30
is 20%.
22
90896
1666
ื”ื™ื 20%.
01:32
Based on reports of this higher number,
23
92562
2237
ื‘ื”ืชื‘ืกืก ืขืœ ื”ืžืกืคืจ ื”ื”ื™ื•ืชืจ ื’ื‘ื•ื” ื”ื–ื”,
01:34
people overestimated the impact of screening.
24
94799
3400
ืื ืฉื™ื ืžืขืจื™ื›ื™ื ื™ืชืจ ืขืœ ื”ืžื™ื“ื” ืืช ื”ืฉืคืขืชืŸ ืฉืœ ื‘ื“ื™ืงื•ืช ืกืงืจ.
01:38
To see why the difference between the two ways of expressing risk matters,
25
98199
4100
ื›ื“ื™ ืœืจืื•ืช ืœืžื” ื”ื”ื‘ื“ืœื™ื ื‘ื™ืŸ ืฉืชื™ ื”ื“ืจื›ื™ื ืœื”ื‘ืขืช ืกื™ื›ื•ืŸ ื—ืฉื•ื‘ื™ื,
01:42
letโ€™s consider the hypothetical example of a drug
26
102299
3020
ื‘ื•ืื• ื ืฉืงื•ืœ ืืช ื”ืžืงืจื” ื”ื”ื™ืคื•ืชื˜ื™ ืฉืœ ืชืจื•ืคื”
01:45
that reduces heart attack risk by 40%.
27
105319
3740
ืฉืžืคื—ื™ืชื” ื‘-40% ืืช ื”ืกื™ื›ื•ื™ ืœื”ืชืงืฃ ืœื‘.
01:49
Imagine that out of a group of 1,000 people
28
109059
2420
ื“ืžื™ื™ื ื• ืฉืžืชื•ืš ืงื‘ื•ืฆื” ืฉืœ 1000 ืื™ืฉ
01:51
who didnโ€™t take the new drug, 10 would have heart attacks.
29
111479
3740
ืฉืœื ืœืงื—ื” ืืช ื”ืชืจื•ืคื” ื”ื—ื“ืฉื”, 10 ื™ืงื‘ืœื• ื”ืชืงืฃ ืœื‘.
01:55
The absolute risk is 10 out of 1,000, or 1%.
30
115219
4540
ื”ืกื™ื›ื•ืŸ ื”ืื‘ืกื•ืœื•ื˜ื™ ื”ื•ื 10 ืžืชื•ืš 1000, ืื• 1%.
01:59
If a similar group of 1,000 people did take the drug,
31
119759
3292
ืื ืงื‘ื•ืฆื” ื“ื•ืžื” ืฉืœ 1000 ืื ืฉื™ื ืชื™ืงื— ืืช ื”ืชืจื•ืคื”,
02:03
the number of heart attacks would be six.
32
123051
2878
ืžืกืคืจ ื”ืชืงืคื™ ื”ืœื‘ ื™ื”ื™ื” ืฉื™ืฉื”.
02:05
In other words, the drug could prevent four out of ten heart attacksโ€”
33
125929
4260
ื‘ืžื™ืœื™ื ืื—ืจื•ืช, ื”ืชืจื•ืคื” ื™ื›ื•ืœื” ืœืžื ื•ืข ืืจื‘ืขื” ืžืชื•ืš ืขืฉืจื” ื”ืชืงืคื™ ืœื‘ --
02:10
a relative risk reduction of 40%.
34
130189
3098
ื”ืคื—ืชื” ื™ื—ืกื™ืช ืฉืœ ืกื™ื›ื•ืŸ ืฉืœ 40%.
02:13
Meanwhile, the absolute risk only dropped from 1% to 0.6%โ€”
35
133287
5448
ื‘ื™ื ืชื™ื™ื, ื”ืกื™ื›ื•ืŸ ื”ืื‘ืกื•ืœื•ื˜ื™ ื™ื•ืจื“ ืจืง ืž- 1% ืœ- 0.6% --
02:18
but the 40% relative risk decrease sounds a lot more significant.
36
138735
5236
ืื‘ืœ ื”ื”ืคื—ืชื” ื”ื™ื—ืกื™ืช ืฉืœ 40% ื ืฉืžืขืช ื—ืฉื•ื‘ื” ื‘ื”ืจื‘ื”.
02:23
Surely preventing even a handful of heart attacks,
37
143971
2790
ื‘ืจื•ืจ ืฉืืคื™ืœื• ืžื ื™ืขื” ืฉืœ ื›ืžื” ื”ืชืงืคื™ ืœื‘,
02:26
or any other negative outcome, is worthwhileโ€” isnโ€™t it?
38
146761
4068
ืื• ื›ืœ ืชื•ืฆืื” ืฉืœื™ืœื™ืช ืื—ืจืช ืฉื•ื•ื” ืืช ื–ื” -- ืœื?
02:30
Not necessarily.
39
150829
1733
ืœื ื‘ื”ื›ืจื—.
02:32
The problem is that choices that reduce some risks
40
152562
3597
ื”ื‘ืขื™ื” ื”ื™ื ืฉื‘ื—ื™ืจื•ืช ืฉืžืคื—ื™ืชื•ืช ื›ืžื” ืกื™ื›ื•ื ื™ื
02:36
can put you in the path of others.
41
156159
3000
ื™ื›ื•ืœื•ืช ืœื”ื’ื‘ื™ืจ ืกื™ื›ื•ื ื™ื ืื—ืจื™ื.
02:39
Suppose the heart-attack drug caused cancer in one half of 1% of patients.
42
159159
5500
ื ื ื™ื— ืฉื”ืชืจื•ืคื” ืœื”ืคื—ืชืช ื”ืชืงืคื™ ืœื‘ ื’ื•ืจืžืช ืœืกืจื˜ืŸ ื‘ื—ืฆื™ ืื—ื•ื– ืžื”ื—ื•ืœื™ื.
02:44
In our group of 1,000 people,
43
164659
2340
ื‘ืงื‘ื•ืฆื” ืฉืœ 1000 ืื ืฉื™ื,
02:46
four heart attacks would be prevented by taking the drug,
44
166999
3420
ืืจื‘ืขื” ื”ืชืงืคื™ ืœื‘ ื™ืžื ืขื• ืขืœ-ื™ื“ื™ ืœืงื™ื—ืช ื”ืชืจื•ืคื”,
02:50
but there would be five new cases of cancer.
45
170419
3872
ืื‘ืœ ื™ื”ื™ื• ื—ืžื™ืฉื” ืžืงืจื™ื ื—ื“ืฉื™ื ืฉืœ ืกืจื˜ืŸ.
02:54
The relative reduction in heart attack risk sounds substantial
46
174291
3700
ื”ื”ืคื—ืชื” ื”ื™ื—ืกื™ืช ื‘ื”ืชืงืคื™ ืœื‘ ื ืฉืžืขืช ื’ื‘ื•ื”ื”,
02:57
and the absolute risk of cancer sounds small,
47
177991
2640
ื•ื”ืกื™ื›ื•ืŸ ื”ืื‘ืกื•ืœื•ื˜ื™ ืฉืœ ืกืจื˜ืŸ ื ืฉืžืข ื ืžื•ืš,
03:00
but they work out to about the same number of cases.
48
180631
3530
ืื‘ืœ ื™ื•ืฆื ืฉื–ื” ื‘ืขืจืš ืื•ืชื• ืžืกืคืจ ืžืงืจื™ื.
03:04
In real life,
49
184161
990
ื‘ื—ื™ื™ื ื”ืืžื™ืชื™ื™ื,
03:05
everyoneโ€™s individual evaluation of risk will vary
50
185151
3090
ื›ืœ ื”ื”ืขืจื›ื•ืช ื”ื‘ื•ื“ื“ื•ืช ืฉืœ ืกื™ื›ื•ืŸ ื™ืฉืชื ื•
03:08
depending on their personal circumstances.
51
188241
2780
ื‘ื”ืชืื ืœื ืกื™ื‘ื•ืช ื”ืื™ืฉื™ื•ืช ืฉืœื”ื.
03:11
If you know you have a family history of heart disease
52
191021
2620
ืื ืืชื ื™ื•ื“ืขื™ื ืฉื™ืฉ ื‘ืžืฉืคื—ื” ืฉืœื›ื ื”ื™ืกื˜ื•ืจื™ื” ืฉืœ ืžื—ืœื•ืช ืœื‘,
03:13
you might be more strongly motivated to take a medication
53
193641
3250
ืื•ืœื™ ืชื”ื™ื” ืœื›ื ืžื•ื˜ื™ื‘ืฆื™ื” ื—ื–ืงื” ื™ื•ืชืจ ืœืงื—ืช ืืช ื”ืชืจื•ืคื”,
03:16
that would lower your heart-attack risk,
54
196891
1980
ืฉืชื•ืจื™ื“ ืืช ื”ืกื™ื›ื•ืŸ ืฉืœื›ื ืœื”ืชืงืฃ ืœื‘,
03:18
even knowing it provided only a small reduction in absolute risk.
55
198871
4888
ืืคื™ืœื• ืื ืชื“ืขื• ืฉื™ืฉ ืจืง ื”ืคื—ืชื” ืงื˜ื ื” ื‘ืกื™ื›ื•ืŸ ื”ืื‘ืกื•ืœื•ื˜ื™.
03:23
Sometimes, we have to decide between exposing ourselves to risks
56
203759
3850
ืœืคืขืžื™ื, ืื ื—ื ื• ืฆืจื™ื›ื™ื ืœื‘ื—ื•ืจ ื‘ื™ืŸ ื—ืฉื™ืคื” ืœืกื™ื›ื•ื ื™ื
03:27
that arenโ€™t directly comparable.
57
207609
2440
ืฉืื™ื ื ื ื™ืชื ื™ื ืœื”ืฉื•ื•ืื” ื™ืฉื™ืจื”.
03:30
If, for example, the heart attack drug carried a higher risk
58
210049
3401
ืื, ืœื“ื•ื’ืžื”, ืชืจื•ืคืช ื”ืชืงืฃ ื”ืœื‘ ื ื•ืฉืืช ืกื™ื›ื•ืŸ ื’ื‘ื•ื” ื™ื•ืชืจ
03:33
of a debilitating, but not life-threatening,
59
213450
2454
ืœื”ืฉืคืขื•ืช ืžื’ื‘ื™ืœื•ืช, ืื‘ืœ ืœื ืžืกื›ื ื•ืช ื—ื™ื™ื,
03:35
side effect like migraines rather than cancer,
60
215904
3200
ื›ืžื• ืžื™ื’ืจื ื•ืช ื‘ืžืงื•ื ืกืจื˜ืŸ,
03:39
our evaluation of whether that risk is worth taking might change.
61
219104
4377
ื”ื”ืขืจื›ื” ืฉืœื ื• ืื ืฉื•ื•ื” ืœื”ืกืชื›ืŸ ืื•ืœื™ ืชืฉืชื ื”.
03:43
And sometimes there isnโ€™t necessarily a correct choice:
62
223481
3580
ื•ืœืคืขืžื™ื ืื™ืŸ ื‘ื”ื›ืจื— ื”ื—ืœื˜ื” ื ื›ื•ื ื”:
03:47
some might say even a minuscule risk of shark attack is worth avoiding,
63
227061
4670
ื›ืžื” ื™ื’ื™ื“ื• ืฉื›ื“ืื™ ืœื”ื™ืžื ืข ืืคื™ืœื• ืžืกื™ื›ื•ืŸ ื–ืขื™ืจ ืœื”ืชืงืคืช ื›ืจื™ืฉื™ื.
03:51
because all youโ€™d miss out on is an ocean swim,
64
231731
2760
ื‘ื’ืœืœ ืฉื›ืœ ืžื” ืฉืชืคืกื™ื“ื• ื–ื• ืฉื—ื™ื™ื” ื‘ื™ื,
03:54
while others wouldnโ€™t even consider skipping a swim
65
234491
2753
ื‘ืขื•ื“ ืฉืื—ืจื™ื ืœื ื™ืฉืงืœื• ืืคื™ืœื• ืœื•ื•ืชืจ ืขืœ ื”ืฉื—ื™ื™ื”
03:57
to avoid an objectively tiny risk of shark attack.
66
237244
3690
ื›ื“ื™ ืœื”ื™ืžื ืข ืžืกื™ื›ื•ืŸ ืื•ื‘ื™ื™ืงื˜ื™ื‘ื™ ื–ืขื™ืจ ืœืžืชืงืคืช ื›ืจื™ืฉ.
04:00
For all these reasons, risk evaluation is tricky at baseline,
67
240934
4200
ืžื›ืœ ื”ืกื™ื‘ื•ืช ื”ืืœื•, ื”ืขืจื›ืช ืกื™ื›ื•ื ื™ื ื”ื™ื ืžืกื•ื‘ื›ืช ื‘ื‘ืกื™ืกื”,
04:05
and reporting on risk can be misleading,
68
245134
2623
ื•ื“ื™ื•ื•ื— ืขืœ ืกื™ื›ื•ื ื™ื ืขืœื•ืœ ืœื”ื˜ืขื•ืช,
04:07
especially when it shares some numbers in absolute terms
69
247757
3482
ื‘ืขื™ืงืจ ื›ืฉืžื“ื•ื•ื—ื™ื ืขืœ ื—ืœืง ืžื”ืžืกืคืจื™ื ื‘ืžื•ื ื—ื™ื ืื‘ืกื•ืœื•ื˜ื™ื™ื
04:11
and others in relative terms.
70
251239
2490
ื•ืขืœ ืื—ืจื™ื ื‘ืžื•ื ื—ื™ื ื™ื—ืกื™ื™ื.
04:13
Understanding how these measures work
71
253729
2210
ืื ืชื‘ื™ื ื• ืื™ืš ื”ืžื“ื“ื™ื ื”ืืœื” ืขื•ื‘ื“ื™ื
04:15
will help you cut through some of the confusion
72
255939
2570
ืชื”ื™ื• ืคื—ื•ืช ืžื‘ื•ืœื‘ืœื™ื
04:18
and better evaluate risk.
73
258509
2190
ื•ืชื•ื›ืœื• ืœื”ืขืจื™ืš ืกื™ื›ื•ื ื™ื ื˜ื•ื‘ ื™ื•ืชืจ.
ืขืœ ืืชืจ ื–ื”

ืืชืจ ื–ื” ื™ืฆื™ื’ ื‘ืคื ื™ื›ื ืกืจื˜ื•ื ื™ YouTube ื”ืžื•ืขื™ืœื™ื ืœืœื™ืžื•ื“ ืื ื’ืœื™ืช. ืชื•ื›ืœื• ืœืจืื•ืช ืฉื™ืขื•ืจื™ ืื ื’ืœื™ืช ื”ืžื•ืขื‘ืจื™ื ืขืœ ื™ื“ื™ ืžื•ืจื™ื ืžื”ืฉื•ืจื” ื”ืจืืฉื•ื ื” ืžืจื—ื‘ื™ ื”ืขื•ืœื. ืœื—ืฅ ืคืขืžื™ื™ื ืขืœ ื”ื›ืชื•ื‘ื™ื•ืช ื‘ืื ื’ืœื™ืช ื”ืžื•ืฆื’ื•ืช ื‘ื›ืœ ื“ืฃ ื•ื™ื“ืื• ื›ื“ื™ ืœื”ืคืขื™ืœ ืืช ื”ืกืจื˜ื•ืŸ ืžืฉื. ื”ื›ืชื•ื‘ื™ื•ืช ื’ื•ืœืœื•ืช ื‘ืกื ื›ืจื•ืŸ ืขื ื”ืคืขืœืช ื”ื•ื•ื™ื“ืื•. ืื ื™ืฉ ืœืš ื”ืขืจื•ืช ืื• ื‘ืงืฉื•ืช, ืื ื ืฆื•ืจ ืื™ืชื ื• ืงืฉืจ ื‘ืืžืฆืขื•ืช ื˜ื•ืคืก ื™ืฆื™ืจืช ืงืฉืจ ื–ื”.

https://forms.gle/WvT1wiN1qDtmnspy7