Pros and cons of public opinion polls - Jason Robert Jaffe

330,725 views ใƒป 2013-05-17

TED-Ed


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

00:00
Transcriber: Andrea McDonough Reviewer: Jessica Ruby
0
0
7000
ืชืจื’ื•ื: Ido Dekkers ืขืจื™ื›ื”: Zeeva Livshitz
00:14
We are constantly asked for our opinions.
1
14577
2588
ืื ื—ื ื• ื ืฉืืœื™ื ื›ืœ ื”ื–ืžืŸ ืœื“ืขื•ืช ืฉืœื ื•.
00:17
Which team do you think will win the Super Bowl?
2
17165
2324
ืื™ื–ื” ืงื‘ื•ืฆื” ืืชื ื—ื•ืฉื‘ื™ื ืฉืชื ืฆื— ื‘ืกื•ืคืจื‘ื•ืœ?
00:19
Who wore it better on the red carpet?
3
19489
2052
ืžื™ ืœื‘ืฉ ื‘ื’ื“ื™ื ื™ืคื™ื ื™ื•ืชืจ ืขืœ ื”ืฉื˜ื™ื— ื”ืื“ื•ื?
00:21
Who are you going to vote for for mayor?
4
21541
2254
ืœืžื™ ืืชื ืขื•ืžื“ื™ื ืœื”ืฆื‘ื™ืข ื‘ื‘ื—ื™ืจื•ืช ืœืจืืฉ ื”ืขื™ืจ?
00:23
Public opinion polls are everywhere.
5
23795
3007
ืžื™ืฉืืœื™ ื“ืขืช ืงื”ืœ ื”ื ื‘ื›ืœ ืžืงื•ื.
00:26
Important decision makers in American government
6
26802
2123
ืžืงื‘ืœื™ ื”ื—ืœื˜ื•ืช ื—ืฉื•ื‘ื•ืช ื‘ืžืžืฉืœ ื”ืืžืจื™ืงืื™
00:28
have long relied on public opinion polls
7
28925
2233
ืชืžื™ื“ ื”ืกืชืžื›ื• ืขืœ ืกืงืจื™ ื“ืขืช ืงื”ืœ
00:31
throughout elections and important legislation.
8
31158
2894
ื‘ืžื”ืœืš ื”ื‘ื—ื™ืจื•ืช ื•ื—ืงื™ืงื•ืช ื—ืฉื•ื‘ื•ืช.
00:34
The problem is public opinion isn't easy to track
9
34052
2773
ื”ื‘ืขื™ื” ื”ื™ื ืฉืงืฉื” ืœืขืงื•ื‘ ืื—ืจื™ ื“ืขืช ื”ืงื”ืœ
00:36
and, often times, isn't even right.
10
36825
2635
ื•ืคืขืžื™ื ืจื‘ื•ืช, ื–ื” ืืคื™ืœื• ืœื ื ื›ื•ืŸ.
00:39
In 1948, the <i>Chicago Daily Tribune</i> ran a now famous headline:
11
39460
4688
ื‘ 1948, ื”ืฉื™ืงืื’ื• ื“ื™ื™ืœื™ ื˜ืจื™ื‘ื™ื•ืŸ ื”ืจื™ืฅ ื›ื•ืชืจืช ืฉืžื•ื›ืจืช ื›ื™ื•ื:
00:44
"Dewey Defeats Truman,"
12
44148
1969
"ื“ื™ื•ืื™ ืžื ืฆื— ืืช ื˜ืจื•ืžืŸ,"
00:46
they cried in big, bold, black and white letters.
13
46117
3402
ื”ื ื–ืขืงื• ื‘ืื•ืชื™ื•ืช ื’ื“ื•ืœื•ืช ืžื•ื’ื“ืฉื•ืช ื‘ืฉื—ื•ืจ ื•ืœื‘ืŸ.
00:49
The problem is
14
49519
1394
ื”ื‘ืขื™ื” ื”ื™ื
00:50
that Dewey hadn't defeated Truman.
15
50913
2328
ืฉื“ื™ื•ืื™ ืœื ื ื™ืฆื— ืืช ื˜ืจื•ืžืŸ.
00:53
The <i>Tribune</i> had relied on polls to come to their conclusion.
16
53241
3832
ื”ื˜ืจื™ื‘ื™ื•ืŸ ื”ืกืชืžืš ืขืœ ืกืงืจื™ื ื›ื“ื™ ืœื”ื’ื™ืข ืœืžืกืงื ื” ืฉืœื”ื.
00:57
Whoops!
17
57073
1379
ื•ื•ืคืก!
00:58
This happens all the time because public opinion polls
18
58452
2878
ื–ื” ืงื•ืจื” ื›ืœ ื”ื–ืžืŸ ืžืคื ื™ ืฉืกืงืจื™ ื“ืขืช ืงื”ืœ
01:01
are either inaccurate or misleading.
19
61330
3161
ื”ื ืื• ืœื ืžื“ื•ื™ื™ืงื™ื ืื• ืžื˜ืขื™ื.
01:04
So, why are they wrong?
20
64491
1660
ืื–, ืœืžื” ื”ื ืฉื’ื•ื™ื™ื?
01:06
And why do we keep using them?
21
66151
2357
ื•ืœืžื” ืื ื—ื ื• ืžืžืฉื™ื›ื™ื ืœื”ืฉืชืžืฉ ื‘ื”ื?
01:08
First, let's start with an important term:
22
68508
2577
ืจืืฉื™ืช, ื‘ื•ืื• ื ืชื—ื™ืœ ืขื ืžื•ืฉื’ ื—ืฉื•ื‘:
01:11
sample.
23
71085
1616
ืžื“ื’ื.
01:12
A sample is the group of people
24
72701
1492
ืžื“ื’ื ื”ื•ื ืงื‘ื•ืฆื” ืฉืœ ืื ืฉื™ื
01:14
that respond to questions during a public opinion poll.
25
74193
3322
ืฉืขื•ื ื” ืœืฉืืœื•ืช ื‘ืžื”ืœืš ืกืงืจ ื“ืขืช ืงื”ืœ.
01:17
A poll's quality rests largely on its sample,
26
77515
2585
ืื™ื›ื•ืช ื”ืกืงืจ ื ื—ื” ื‘ืขื™ืงืจ ืขืœ ื”ืžื“ื’ื ืฉืœื•,
01:20
and a sample can be bad in a few key ways.
27
80100
2961
ื•ืžื“ื’ื ื™ื›ื•ืœ ืœื”ื™ื•ืช ื’ืจื•ืข ื‘ื›ืžื” ื“ืจื›ื™ื ืขื™ืงืจื™ื•ืช.
01:23
It can be too small,
28
83061
1594
ื”ื•ื ื™ื›ื•ืœ ืœื”ื™ื•ืช ืงื˜ืŸ ืžื“ื™,
01:24
too narrow,
29
84655
1308
ืฆืจ ืžื“ื™,
01:25
or the poll itself can be too difficult.
30
85963
2855
ืื• ืฉื”ืกืงืจ ืขืฆืžื• ื™ื›ื•ืœ ืœื”ื™ื•ืช ืงืฉื” ืžื“ื™.
01:28
Polls that are too small are bad for obvious reasons.
31
88818
2626
ืกืงืจื™ื ืงื˜ื ื™ื ืžื“ื™ ื’ืจื•ืขื™ื ืžืกื™ื‘ื•ืช ื‘ืจื•ืจื•ืช.
01:31
And while you can't possibly ask
32
91444
1672
ื•ื‘ืขื•ื“ ืฉืืชื ืœื ื™ื›ื•ืœื™ื ืœื‘ืงืฉ
01:33
every single person in America for their opinion,
33
93116
2583
ืžื›ืœ ืื“ื ื‘ืืžืจื™ืงื” ืืช ื”ื“ืขื” ืฉืœื•,
01:35
the more people you ask,
34
95699
1209
ื›ื›ืœ ืฉืชืฉืืœื• ื™ื•ืชืจ ืื ืฉื™ื,
01:36
the more accurate your prediction.
35
96908
2706
ื”ืชื—ื–ื™ืช ืฉืœื›ื ืชื”ื™ื” ืžื“ื•ื™ื™ืงืช ื™ื•ืชืจ.
01:39
Polls that are too narrow,
36
99614
1758
ืกืงืจื™ื ืฆืจื™ื ืžื“ื™,
01:41
that only ask a certain type of person a question,
37
101372
2667
ืฉืฉื•ืืœื™ื ืจืง ืกื•ื’ ืžืกื•ื™ื™ื ืฉืœ ืื ืฉื™ื ืฉืืœื”,
01:44
are bad, too.
38
104039
1627
ื’ื ื”ื ื’ืจื•ืขื™ื.
01:45
Consider a poll about whether or not
39
105666
1706
ื—ืฉื‘ื• ืขืœ ืกืงืจ ืฉืฉื•ืืœ ืื
01:47
the potato is the best vegetable in America.
40
107372
3174
ืชืคื•ื— ื”ืื“ืžื” ื”ื•ื ื”ื™ืจืง ื”ื›ื™ ื˜ื•ื‘ ื‘ืืžืจื™ืงื”.
01:50
If you only asked people in Idaho,
41
110546
2454
ืื ืืชื ืฉื•ืืœื™ื ืจืง ืื ืฉื™ื ื‘ืื™ื™ื“ื”ื•,
01:53
where the state food is the potato,
42
113000
2209
ื‘ื” ื”ืื•ื›ืœ ื”ืœืื•ืžื™ ื”ื•ื ืชืคื•ื—ื™ ืื“ืžื”,
01:55
chances are that you would get a much different answer
43
115209
2094
ื”ืกื™ื›ื•ื™ ื”ื•ื ืฉืชืงื‘ืœื• ืชืฉื•ื‘ื” ืฉื•ื ื” ืžืื•ื“
01:57
than if you asked people in the state of New Mexico,
44
117303
2526
ืžืืฉืจ ืื ืชืฉืืœื• ืื ืฉื™ื ื‘ืžื“ื™ื ืช ื ื™ื• ืžืงืกื™ืงื•,
01:59
where the state vegetable is beans.
45
119829
2813
ืฉื ื”ื™ืจืง ื”ืœืื•ืžื™ ื”ื•ื ืฉืขื•ืขื™ืช.
02:02
Getting the right kind of diversity in your sample
46
122642
2420
ืœื”ืฉื™ื’ ืืช ื”ื’ื™ื•ื•ืŸ ื”ื ื›ื•ืŸ ื‘ืžื“ื’ื
02:05
means making sure that your sample has a range
47
125062
2406
ืื•ืžืจ ืœื“ืื•ื’ ืฉืœืžื“ื’ื ืฉืœื›ื ื™ื”ื™ื” ื˜ื•ื•ื—
02:07
of ages,
48
127468
844
ืฉืœ ื’ื™ืœืื™ื,
02:08
races,
49
128312
839
ื’ื–ืขื™ื,
02:09
genders,
50
129151
664
02:09
and geographic regions,
51
129815
1287
ืžื™ื ื™ื,
ื•ืื–ื•ืจื™ื ื’ืื•ื’ืจืคื™ื™ื,
02:11
just to name a few.
52
131102
1884
ืจืง ืœื”ื–ื›ื™ืจ ื›ืžื”.
02:12
Finally, polls that are too hard can't tell you much either.
53
132986
3920
ืœื‘ืกื•ืฃ, ืกืงืจื™ื ืงืฉื™ื ืžื“ื™ ืœื ืื•ืžืจื™ื ื”ืจื‘ื”.
02:16
If you're asking people for their opinions on things
54
136906
2687
ืื ืืชื ืฉื•ืืœื™ื ืื ืฉื™ื ืืช ื”ื“ืขื” ืฉืœื”ื ืขืœ ื“ื‘ืจื™ื
02:19
about which they have no prior knowledge,
55
139593
2194
ืขืœื™ื”ื ืื™ืŸ ืœื”ื ืžื™ื“ืข ืžื•ืงื“ื,
02:21
the results will be pointless.
56
141787
2163
ื”ืชื•ืฆืื•ืช ื™ื”ื™ื• ื—ืกืจื•ืช ืžืฉืžืขื•ืช.
02:23
You're better off shaking a Magic 8 ball.
57
143950
2443
ืขื“ื™ืฃ ืœื›ื ืœืฉืงืฉืง ืืช ื›ื“ื•ืจ ื”ืงืกื.
02:26
It's not just the people you're asking
58
146393
1677
ื–ื” ืœื ืจืง ื”ืื ืฉื™ื ืฉืืชื ืฉื•ืืœื™ื
02:28
that can cause bias, though.
59
148070
1682
ืฉื™ื›ื•ืœื™ื ืœื’ืจื•ื ืœืกื˜ื™ื”, ืขื ื–ืืช.
02:29
The person doing the asking is part of the problem, too.
60
149752
3419
ื”ืื ืฉื™ื ืฉืฉื•ืืœื™ื ื”ื ื—ืœืง ืžื”ื‘ืขื™ื” ื’ื ื”ื.
02:33
That's called interviewer bias.
61
153171
3046
ื–ื” ื ืงืจื ืกื˜ื™ื™ืช ืžืจืื™ื™ืŸ.
02:36
Interviewer bias is all about the effect
62
156217
1960
ืกื˜ื™ื™ืช ืžืจืื™ื™ืŸ ื ื•ื’ืขืช ืœืืคืงื˜
02:38
that the person asking the questions
63
158177
2169
ืฉืœืื“ื ืฉืฉื•ืืœ ืืช ื”ืฉืืœื•ืช
02:40
has on the sample.
64
160346
1790
ื™ืฉ ืขืœ ื”ืžื“ื’ื.
02:42
Humans generally don't like confrontation.
65
162136
2722
ืื ืฉื™ื ื‘ื“ืจืš ื›ืœืœ ืœื ืื•ื”ื‘ื™ื ืขื™ืžื•ืช.
02:44
People worry that their answers may make them look bad.
66
164858
3448
ืื ืฉื™ื ื“ื•ืื’ื™ื ืฉื”ืชืฉื•ื‘ื•ืช ืฉืœื”ื ืื•ืœื™ ื’ื•ืจืžื•ืช ืœื”ื ืœื”ืจืื•ืช ืจืข.
02:48
Therefore, we find that people tend to give
67
168306
1817
ืœื›ืŸ, ืื ื—ื ื• ืžื•ืฆืื™ื ืฉืื ืฉื™ื ื ื•ื˜ื™ื ืœืชืช
02:50
socially desirable responses,
68
170123
2417
ืชืฉื•ื‘ื•ืช ื ื—ืฉืงื•ืช ื—ื‘ืจืชื™ืช,
02:52
not necessarily their honest opinions,
69
172540
2563
ื•ืœื ืชืžื™ื“ ืืช ื“ืขืชื ื”ืืžื™ืชื™ืช,
02:55
because they don't want to come across
70
175103
1798
ืžืคื ื™ ืฉื”ื ืœื ืจื•ืฆื™ื ืœื”ื•ืคื™ืข
02:56
as heartless,
71
176901
802
ื›ื—ืกืจื™ ืœื‘,
02:57
racist,
72
177703
602
ื’ื–ืขื ื™ื™ื,
02:58
or bigoted.
73
178305
1479
ืื• ืฆืจื™ ืื•ืคืงื™ื.
02:59
And the way we word our questions matters too.
74
179784
2667
ื•ื”ื“ืจืš ื‘ื” ืื ื—ื ื• ืžื ืกื—ื™ื ืืช ื”ืฉืืœื•ืช ืฉืœื ื• ืžืฉื ื” ื’ื ื”ื™ื.
03:02
When polls purposely sway the answers one way or the other,
75
182451
3212
ื›ืฉืกืงืจื™ื ืžื˜ื™ื ื‘ื›ื•ื•ื ื” ืืช ื”ืชืฉื•ื‘ื•ืช ืœื›ื™ื•ื•ืŸ ืื—ื“ ืื• ืื—ืจ,
03:05
it's called a push poll
76
185663
1887
ื–ื” ื ืจืื” ืกืงืจ ืžื•ื˜ื”
03:07
because it pushes people to answer a certain way.
77
187550
3345
ืžืคื ื™ ืฉื”ื•ื ืžื˜ื” ืื ืฉื™ื ืœืขื ื•ืช ื‘ื“ืจืš ืžืกื•ื™ื™ืžืช.
03:10
"Would you vote for candidate Smith?"
78
190895
1734
"ื”ืื ื”ื™ื™ืช ืžืฆื‘ื™ืข ืœืžื•ืขืžื“ ืกืžื™ืช'?"
03:12
is a perfectly normal question.
79
192629
2660
ื”ื™ื ืฉืืœื” ื ื•ืจืžืœื™ืช ืœื’ืžืจื™.
03:15
"Would you vote for candidate Smith
80
195289
1691
"ื”ืื ื”ื™ื™ืช ืžืฆื‘ื™ืข ืœืžื•ืขืžื“ ืกืžื™ืช'
03:16
if you knew that he robs senior citizens?"
81
196980
2652
ืื ื”ื™ื™ืช ื™ื•ื“ืข ืฉื”ื•ื ืขื•ืฉืง ืื–ืจื—ื™ื ื•ืชื™ืงื™ื?"
03:19
is a push poll.
82
199632
1907
ื–ื• ืฉืืœื” ืžื˜ื”.
03:21
So, if polls are open
83
201539
1205
ืื–, ืื ืกืงืจื™ื ื”ื ืคืชื•ื—ื™ื
03:22
to all sorts of manipulation and inaccuracies,
84
202744
2376
ืœื›ืœ ืžื™ื ื™ ืกื•ื’ื™ื ืฉืœ ืžื ื™ืคื•ืœืฆื™ื•ืช ื•ืื™ ื“ื™ื•ืงื™ื,
03:25
why are they still so prevalent?
85
205120
2271
ืœืžื” ื”ื ืขื“ื™ื™ืŸ ื ืคื•ืฆื™ื?
03:27
Despite their flaws, public opinion polls provide us
86
207391
2684
ืœืžืจื•ืช ื›ืœ ื”ื—ืกืจื•ื ื•ืช ืฉืœื”ื, ืกืงืจื™ ื“ืขืช ืงื”ืœ ืžืกืคืงื™ื ืœื ื•
03:30
with some sense of the thoughts and moods
87
210075
2281
ืžื•ืฉื’ ืžืกื•ื™ื™ื ืฉืœ ื”ืœื›ื™ ืžื—ืฉื‘ื” ื•ืžืฆื‘ื™ ืจื•ื—
03:32
of large groups of people.
88
212356
1919
ืฉืœ ืงื‘ื•ืฆื•ืช ื’ื“ื•ืœื•ืช ืฉืœ ืื ืฉื™ื.
03:34
They offer politicians the chance to pass legislation
89
214275
2648
ื”ื ืžืฆื™ืขื™ื ืœืคื•ืœื™ื˜ื™ืงืื™ื ืืช ื”ื”ื“ืžื ื•ืช ืœื”ืขื‘ื™ืจ ื—ืงื™ืงื”
03:36
they think a majority of Americans will support.
90
216923
2686
ืฉื”ื ื—ื•ืฉื‘ื™ื ืฉืจื•ื‘ ื”ืืžืจื™ืงืื™ื ื™ืชืžื›ื• ื‘ื”.
03:39
They help fashionistas on TV
91
219609
2102
ื”ื ืขื•ื–ืจื™ื ืœืื ืฉื™ ืื•ืคื ื” ื‘ื˜ืœื•ื•ื™ื–ื™ื”
03:41
know which star wore the dress better on the red carpet.
92
221711
3591
ืœื“ืขืช ืื™ื–ื” ื›ื•ื›ื‘ ืœื‘ืฉ ืื™ื–ื” ื‘ื’ื“ ื˜ื•ื‘ ื™ื•ืชืจ ืขืœ ื”ืฉื˜ื™ื— ื”ืื“ื•ื.
03:45
Finally, they make us,
93
225302
2654
ืœื‘ืกื•ืฃ, ื”ื ื’ื•ืจืžื™ื ืœื ื•,
03:47
the people who get polled,
94
227956
1667
ื”ืื ืฉื™ื ืฉื ืกืงืจื™ื,
03:49
feel as though our voice has been heard.
95
229623
3212
ืœื”ืจื’ื™ืฉ ื›ืื™ืœื• ื”ืงื•ืœ ืฉืœื ื• ื ืฉืžืข.
03:52
So, next time you get a phone call asking your opinion,
96
232835
2542
ืื–, ื‘ืคืขื ื”ื‘ืื” ืฉืืชื ืžืงื‘ืœื™ื ืฉื™ื—ืช ื˜ืœืคื•ืŸ ืฉืžื‘ืงืฉืช ืืช ื“ืขืชื›ื,
03:55
or if you see a poll online,
97
235377
1797
ืื• ืื ืืชื ืจื•ืื™ื ืกืงืจ ื‘ืจืฉืช,
03:57
take some time to think about who is asking
98
237174
2423
ืงื—ื• ืืช ื”ื–ืžืŸ ืœื—ืฉื•ื‘ ืžื™ ื”ื•ื ื”ืฉื•ืืœ
03:59
and why they're asking.
99
239597
1821
ื•ืœืžื” ื”ื ืฉื•ืืœื™ื.
04:01
Then, take that poll, and its results,
100
241418
2500
ืื–, ืขืฉื• ืืช ื”ืกืงืจ, ื•ืฆืคื• ื‘ืชื•ืฆืื•ืช,
04:03
with a grain of salt
101
243918
1722
ืขื ื’ืจื’ืจ ืžืœื—
04:05
or a potato.
102
245640
2463
ืื• ืชืคื•ื— ืื“ืžื”.
ื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