Transcript Measuring Public Opinion

PUBLIC OPINION




Definitional preliminaries
Measurements
Formation
Ideological differences in the U.S.
What constitutes a “public”?





A group that has something in common.
Members of the group:
are confronted by an issue
are divided in the ideas as how to meet the
issue
engage in discussion over the issue
What is an opinion?





An opinion is an expression of attitudes at a
particular time
Beliefs are the cognitive components: our
understanding of the way things are.
Values are ideals how things should be.
Attitudes are built upon our beliefs and values.
Cognitions (beliefs) lead to affects (feelings) and
predisposition toward certain behavior.
What is Public Opinion?

an aggregate of the individual views, attitudes,
and beliefs about a particular topic, expressed
by a significant proportion of a community.

a synthesis of the views of all or a certain segment
of society (the collective view of a population)
or
a collection of many differing or opposing views.


What is Public Opinion?

The attitudes and beliefs communicated by
ordinary citizens to decision makers

The American political scientist V.O. Key defined
public opinion in 1961 as “opinions held by
private persons which governments find it
prudent to heed.”
In order for a phenomenon to count as
public opinion, there must be:




(1) an issue,
(2) a significant number of individuals who
express opinions on the issue,
(3) some kind of a consensus among at least some
of these opinions, and
(4) this consensus must directly or indirectly exert
influence.
Why study Public Opinion?



Policy, in Democratic States, Should Rest
on Public opinion. E.g., Public Opinion
Dictates the Bounds of U.S. Foreign Policy
Respect for public opinion increases
Political Efficacy and Social Trust
Public Opinion Provides Clues About
Culture.
CURRENT POLLS

Pew Research Center

http://www.people-press.org/
Understanding society / culture
Gay Marriage







YEAR Favor
2004 30%
2006 39%
2008 38%
2010 42%
2012 47%
2013 53%
Oppose
58%
51%
49%
48%
43%
39%
Unsure
12%
10%
13%
10%
10%
8%
Learning about the U.S. in
context of other countries
Government has responsibility to
reduce income differences
UK
US
NL
NZ
CAN
Defini43
tely yes
18
39
35
28
Probably yes
38
28
40
30
33
Probably not
10
25
13
15
20
Definitely no
7
27
6
18
17
Sexual relations before marriage
UK
US
NL
DK
CAN
Always
wrong
11
30
7
6
12
Almost
always
wrong
7
11
3
3
7
Sometimes
14
20
20
9
15
Not
wrong at
all
66
38
70
83
65
Religion brings
more conflict than peace
UK
US
NL
DK
CAN
Agree
78
35
70
86
63
Neither
12
21
16
7
17
Disagree
9
43
14
6
20
There is hell
YES
UK
US
NL
DK
JAP
13
55
13
8
6
34
12
48
60
21
Maybe
NO
There is heaven
YES
UK
US
NL
DK
JAP
23
66
23
14
7
21
6
31
43
19
Maybe
NO
Believe in God
UK
US
NL
DK
JAP
NO
41
18
48
55
55
Sometimes
14
5
8
11
32
YES
with
doubts
23
15
18
20
9
YES
without
a doubt
26
63
26
14
4
People with strong faith are
intolerant of others
UK
US
NL
DK
CAN
79
49
78
79
67
Not sure 13
23
15
12
17
NO
28
7
9
16
YES
8
Europe vs. U.S. on death penalty



1997: 75% of Americans supported
death penalty
2011: 61% of Americans support death
penalty.
2008: about 30% of Europeans support
death penalty
Are in favor of the death penalty for a
person convicted of murder?
A person has the right to suicide if he/she is
tired of living

.
90
80
70
60
50
40
30
20
10
0
YES
NO
Happiness

.
60
50
40
30
20
10
0
Very Happy
Pretty Happy
Not too happy
Trust

.
60
50
40
30
20
10
0
Most people
can be trusted
Can't be too
careful
Measuring Public Opinion


Almost always measured by
aggregating individual opinions
Sometimes by observing mass behavior
(demonstrations, strikes, etc.)
The depth vs. the extent


Practical considerations: the time and
resources needed to measure public
opinion
More people = less depth
methods




By direct response (surveys, focus
groups, interviews)
By observation of behavior (election
results, consumer behavior)
By experiments
By analysis of mass media /
literature
The analysis of mass media etc.




Content analysis of the news
Content analysis of other available
records
Usually post facto (e.g., historical
research)
Usually explanatory, not
descriptive
Experiments




A mode of observation that enables
researchers to probe causal relations
Explaining why something happens (not
what happens): e.g., x happens
because of y.
Predictive value
True experiments vs. quasi-experiments
Observation of behavior



Elections
Consumer behavior (e.g., product
preferences)
Other (e.g., reading patterns, college
enrollment, crime)
A direct response

Focus groups (group dynamics /
deliberation)

In-depth interviews

Surveys (polls)
Surveys / polls

Key element of polling: sampling.

First: selection of specific population.


Populations: the entire adult population,
registered voters, whites only, Hispanic
only, college students, etc.
Second: taking a sample from the
population.
Sampling




Random sample (strict probability):
Each person has exactly the same chance
of being drawn as any other person from
the population under consideration
Probability sample / oversampling:
some units are more likely to be drawn
Confidence level / sampling error
Confidence level: the probability of
“perfect” random sample.
Usually 95%
 Sampling error / margin of error:
possible error in accuracy of results.

Sample size / confidence


Samples around 1,000 are usually
adequate for populations of 100,000 and
larger (including U.S. adult population of
about 200 million).
Sampling error for such samples is
usually around 3 – 4 percent.
What’s so special about 1000
respondents?


“Margin of error” is largely driven by the size of
the sample. The larger the sample, the smaller the
‘margin of error’
As the sample size increases the “margin of
error” falls, quite dramatically between small
sample sizes of say 100 and larger sample sizes
of say 1000. But once we get to 1000, we don’t
see much change in the “margin of error”
BAD SAMPLE



To collect random samples is very
expensive
Instead, some researchers use
convenience samples, e.g. self-selected
opinion polls, or SLOP surveys.
Margin of sampling error cannot be
estimated for a SLOP poll, no matter how
large.
Problems / challenges



Selecting samples (“cell-phone problem”)
Questionnaires
Misuse of polls (“push polls”)
Question Wording


The general principle:
every respondent should understand the
question and be able to answer it with
reliability – that is, if she were asked the
same question again, she would give the
same answer.
Open versus Closed Questions
Open: What is the most important
problem facing the country today?
Closed: Which of the following problems is
the most important one facing the
country today?
Double negative (confusing)
Provide a simple “Agree” or “Disagree” response.
On occasion, I am unable to express how interested
in politics I am.
 When the respondent answers “Agree” does it
mean he is not interested in politics? Or he is
interested but just can’t express himself well? This
question would be better worded as:
I am usually interested in politics.

Double-barreled:
Did you vote in the 2004 and 2006 elections?
 When the respondent answers “yes” does it mean
that she voted in 2004 and 2006, in 2004 but not in
2006, or in 2006 but not in 2004? This question
could be worded as two separate questions or in
the following way:
Did you vote in both 2004 and 2006, in 2004 but
not 2006, in 2006 but not in 2004, or in neither
election?
Leading:
In this question form, an initial phrase leads the
respondent by suggesting the position of an
authority with which it might be difficult for the
respondent to disagree (or agree). Leading
questions introduce a bias
Do you support President Bush’s decision to send
additional troops to Iraq?
 A better question might be:
Do you favor or oppose sending additional troops
to Iraq?

Question Order: the order of the questions
may be designed to “lead” the respondent

For example, if you ask questions about a
specific issue like the economy before
asking what the most important problem is
facing the nation, respondents will be more
likely to name the economy in that
subsequent question then they would have
been without having that context set up for
them.
Problems with questions


Imagine the U.S. is preparing for the
outbreak of an unusual disease, which is
expected to kill 600 people.
Two alternative programs have been
proposed. Assume that the exact scientific
estimates of the consequences of the
programs are as follows:
Which of these two programs do you
favor?


If Program A is adopted, 200 people will be
saved.
If program B is adopted, there is a one-third
probability that 600 people will be saved,
and a two-third probability that no people
will be saved.
Which of these two programs do you
favor?


If Program C is adopted, 400 people will
die.
If Program D is adopted, there is a one-third
probability that nobody will die, and a twothird probability that 600 people will die.
Combination of methods


Mutual verification
External consistency among different
measurements
Public opinion formation

Every opinion is a marriage of
information and predisposition:

Information forms a mental picture
Predisposition motivates some conclusion

Converting information
into public opinion




Elaboration Likelihood Model:
What happens when a person receives
a persuasive message?
Distinct routes of processing:
Central route (Elaboration occurs)
Peripheral route (Nonelaboration)
Elaboration Likelihood Model


Elaboration = the extent to which a person
thinks about the issue-relevant arguments
contained in a message.
Elaboration: complex processes; active
thinking and active counterargument.
Nonelaboration: superficial thoughts;
unthinking acceptance or unthinking
rejection
Elaboration Likelihood Model

Elaboration depends upon motivation
and ability: low motivation and low
ability leads to more peripheral
processing.
Elaboration Likelihood Model





Motivation depends on:
Involvement
Diversity of Argument
Need for Cognition
Ability depends on the rationality of a
person
Elaboration Likelihood Model





Central route:
The strength of the argument
Peripheral:
Credibility / Liking
Number of arguments
Psychological theories


Conditioning
Cognitive Dissonance
Conditioning
Classical (association).
 Operant (reinforcement)
Positive reinforcement (favorable stimulus to
cause/increase behavior)
Negative reinforcement (removal of aversive
stimulus to cause behavior)
Positive punishment (aversive stimulus to decrease
behavior)
Negative punishment (removal of favorable stimulus
to decrease behavior)

Cognitive Dissonance Theory




Any two cognitive elements will have one
of three kinds of relationships:
Irrelevant (I like ice cream and I am aware
of dangers of drunk driving)
Consonant (I like ice cream and I know
that dairy products are good for the body)
Dissonant (I like ice cream and I am aware
of the dangers of high cholesterol)
Cognitive Dissonance Theory


Inconsistency between two cognitins gives
rise to the uncomfortable psychological
state of “cognitive dissonance”
Because dissonance is so very
uncomfortable to us, we will do almost
anything to reduce the dissonance in
order to achieve consonance
Three ways in which dissonance can
be reduced
1. One might change one or more of the
cognitive elements
(a) change the original behavior: stop
eating ice cream or reduce frequency of
eating ice cream
(b) or reject the new information: it is not
true that cholesterol is bad for you
Three ways in which dissonance can
be reduced
2. New elements might by added
Eating ice cream is extremely enjoyable
Ice cream alone is not that bad; there are
worse things with cholesterol
3. One might come to see the elements as
less important
The research on cholesterol is inconclusive.
Cognitive Dissonance Theory
in decision making




The amount of dissonance depends on:
Importance of the decision
Attractiveness of the chosen alternative
Perceived attractiveness of the unchosen
alternative
A new book on the subject
Mistakes Were Made (But Not by Me):
Why We Justify Foolish Beliefs, Bad Decisions, and Hurtful Acts
by Carol Tavris and Elliot Aronson
Rational Choice (economic)


There is no common good or public
interest
Primary goal of individuals is to
maximize benefits in a self-centered and
selfish way
Self-interest as the means





Individuals have goals
They pursue goals by minimizing
costs and maximizing benefits
Complete information
Low information
Cost of participation
The Prisoners’ dilemma
Cooperate=stick to
their innocence
Defect=rat on the
other prisoner
Prisoner # 1
Coope- Defect
rate
Coope-2
Prisoner rate
-2
#2
Defect
-10
0
0
-10
-5
-5
Predispositions

People tend to resist arguments that are
inconsistent with their political
predisposition
Public Opinion Formation:
Predispositions


Political socialization: the transmission
of the political culture from one
generation to another
Political culture: A distinctive and
patterned way of thinking about how
political and economic life ought to be
carried out.
Stereotyping / Frames of reference




Cognitive frameworks consisting of
knowledge and beliefs about specific
social groups (suggesting that all members
of a group possess certain traits)
Allow to make quick judgments
Self-confirming role
Prejudice, favoritism
Predispositions: Similarities
American Political Culture





Liberty: the perception that a minimum
of limitations should be preserved
Free enterprise / competition
Political participation and its
effectiveness (efficacy)
Equality: Equal opportunity / equal
outcomes
Secular values / Religious freedom
Influences on socialization






The Family
The Peer Group
The Schools
College
Workplace
Political Parties
Predispositions: The Differences
Conservatives
Libertarians
Liberals
Populists
The median voter phenomena
Divided America ?
Elections 2000
Elections 2000 (population adjusted)
Elections 2000 by counties
Elections 2000 by counties
(population adjusted)
Elections 2000 by “close” counties (less
than 55-45%)
Elections 2000 by “close” counties (less
than 55-45%). Population adjusted