2. Media Effects, Definitions
Persuasion (attitude change)
Reinforcement
Learning
Agenda setting
Priming
Framing
Historical note: Iyengar did some path-breaking work on subtle effects
like agenda-setting, priming and framing, but because of the loose and
strange way he defined these terms, it took others a decade to sort out
the differences.
3. Definitions (excludes learning and reinforcement)
Agenda setting: The amount of news coverage (e.g., # of stories) an issue (or
candidate) receives influences the degree to which the public thinks the
issue or candidate is important.
E.g., “It’s the economy stupid” in 1992 election.
Focus on “what people think about“ vs. “what people think.”
Priming: Changes in the number of stories about an issue influence the
criteria used to evaluate a political leader (An extension of agenda-setting).
E.g., “it’s the economy stupid” low H.W. Bush approval ratings.
(Issue or Emphasis) Framing: Changes in the content of a story about an
issue influence the criteria used to evaluate the issue. Framing works by
altering the importance individuals attach to certain beliefs that shape
political attitudes. Influences how people think about an issue.
Media frames Framing effects
E.g., Headlines, pictures, etc. Interpret meaning of story.
Direct Persuasion (attitude change): works by altering actual belief
content.
E.g., Negative ads negative evaluations of a candidate.
4. Agenda setting, Time-series
• News coverage more important
than objective reality in shaping
perceptions of important
problems.
• Change in the no. of news stories
on unemployment a better
predictor of change in % naming
unemployment as the most
important problem than the actual
unemployment rate.
5. Agenda setting is a function of news
coverage, which may or may not be related to
real world events
Public continued
to rate crime as
the most
important problem
for several years
after the actual
crime rate began
to fall.
7. Priming: a political consequence
of agenda setting
Priming vs. Persuasion: Priming occurs when an
individual changes the criteria on which he or she
bases an overall evaluation (e.g., basing the
evaluation on defense or energy), whereas
persuasion involves altering what an individual
thinks of the president on a given dimension
(e.g., does the president do a good or poor job
on defense policy?).
Priming does not involve changing perceptions
of how well the president is doing on an issue—it
simply alters the issues on which individuals base
their overall evaluations.
8. Percentage
20
100
50
0
30
60
70
90
10
40
80
2/10-12/01
4/4-5/01
6/14-18/01
9/11-12/01
9/20-23/01
9/11
12/7-10/01
1/15-17/02
2/24-26/02
4/28-5/1/02
5/19-20/02
7/8-9/02
7/22-23/02
9/2-5/02
10/3-6/02
11/20-24/02
1/19-22/03
2/10-12/03
3/4-5/03
Iraq
3/15-16/03
3/2 0/03 PANEL
Invades
3/22/2003
3/24/2003
3/26-27/03
4/11-13/03
5/9-12/03
general approval ratings
8/26-28/03
9/28-10/1/03
11/10-12/03
1/12-15/04
2/24-27/04
3/30-4/1/04
5/20-23/04
7/11-15/04
9/12-16/04
10/1-3/04
his job as President?
10/14-17/04
11/18-21/04
2/24-28/05
4/13-16/05
6/10-15/05
7/29-8/2/05
9/6-7/05
10/3-5/05
12/2-6/05
1/20-25/06
3/9-12/06
4/28-30/06
6/10-11/06
8/11-13/06
How can priming explain the fluctuations in George Bush’s approval?
9/15-19/06
10/27-31/06
Do you approve or disapprove of the way George W. Bush is handling
12/8-10/06
1/18-21/07
2/23-27/07
3/26-27/07
Surge
Priming example: How priming influenced G.W. Bush’s
4/20-24/07
6/26-28/07
9. Evaluations of Iraq Policy
Do you approve or disapprove of the way George W. Bush is
handling the situation with Iraq?
90 Saddam
Statue
Bush
Toppled
80 launches
invasion
Saddam
70
captured
60
Percentage
50
40
30 Abu
Ghraib
20
10 Iraq Study
Group
Report
0
10. Politicians attempt to prime the issues where
they are strongest: Issue Emphasis in Bush’s
2002 State of the Union Address
11. Impact of watching Bush’s 2002 State of the
Union Address, Experiment
People who watched the
speech were much more
likely to base their general
approval of Bush on
terrorism approval and
leadership.
14. The limits of priming effects: Who
is primed?
(Experiment, Miller and Krosnick)
To what extent are citizens mindless
“victims” of the media’s “primordial
power?”
Many people aren’t influenced, including:
People with little exposure to the news
People with little trust in the news media
15. Politicians attempt to prime and counter-prime:
Joseph Biden in a Democratic debate said of
Rudy Giuliani:
“a Giuliani speech is ‘9/11’ + subject + verb.”
16. Framing: News Frames (the stimulus)
Definitions:
Frames act like plots or story lines, lending
coherence to otherwise discrete pieces of information.
Frames organize the presentation of facts and opinion
in a news story.
How Journalists frame a story: through use of
headlines, well-placed quote or soundbite, visual
images or photos, metaphors, caricatures, and
catchphrases all may carry frames, especially useful
for TV. Frames reduce a complex issue down to 1 or 2
central points.
Note: this is the stimulus, not the effect
17. Examples of frames, news frames
Estate tax or Death tax?
Poor people or welfare?
Process, strategic, or game frame vs. policy frame
Episodic versus Thematic Frames
Clinton’s affair: a personal matter between him & his
family, or an impeachable offense?
Decrease in Medicare spending: A reduction in increases in
Medicare spending or a cut in the program?
News stories on Iraqi war:
Casualty frames relative to enemy killed?
An attack by insurgents, Al Qaida, civil war, domestic
violence, Bush popularity
18. Framing Effects
Versus other effects:
Agenda-setting and priming demonstrate how
mere media attention can subtly influence public
opinion.
Framing focuses more on media content than
mere coverage of a problem.
19. Framing Effects, Experiment
Nelson et al.
Tom Nelson and others found
different emphasis frames in
local TV news when KKK was
threatening to march in
Columbus, OH
Free speech: KKK and protestors
were determined to get out their
message
Threat of violence: KKK rallies
provoke violence between a hate
group and protestors
20. News frames on civil liberties issue
New frame influences emphasis and tolerance
Free speech a more important Public order a more important
determinant of political tolerance determinant of political tolerance
(should KKK be allowed to hold a (should KKK be allowed to hold a
rally?)more support for rally. rally?)less support for rally.
21. Framing affects the weight given to different
beliefs underlying political tolerance
Political tolerance (i.e., allowing the
expression of ideas one finds offensive) is
based on both support for free speech and
support for public order.
News frames influence the salience or weight
of free speech or public order, and thus
influence the level of political tolerance.
23. Katrina news stories: Framing with photos
Caption says he has just Caption says they are shown
been "looting a grocery
"after finding bread and
store."
soda from a local grocery
store."
24. The limits of issue framing
To what extent are citizens mindless “victims” of the
media’s “primordial power”?
Framing effects are important but are not so
mindless.
Predispositions: People can reject a frame that’s
inconsistent with their predispositions
Source: Frames from less credible sources (e.g., The
National Enquirer) have less impact than those from
credible sources (e.g., New York Times)
Competitive Framing: Frames of equal strength are
neutralize each other. Inoculation: use a weak counter-
frame to bolster the impact of the stronger frame.
25. Death Penalty Experiment: Aggregate Results
Baseline Condition Racial Argument Innocent Argument
(No Argument)
Some people say * that Some people say * that
the death penalty is unfair the death penalty is unfair
because most of the people because too many
who are executed are innocent people are being
African-Americans. executed.
Do you favor or Do you favor or oppose Do you favor or oppose
oppose the death the death penalty for the death penalty for
penalty for persons persons convicted of persons convicted of
convicted of murder? murder? murder?
Whites 65.96% b 77% b 64.28% b
% Favor - +12% Favora b - .70% Favor
Baseline
Blacks 50% 38% 34%
% Favor - -12% Favora -16% Favor
Baseline
a Difference across baseline and argument condition is statistically significant (≤.05)
b Difference across race of respondent is statistically significant (≤.05)
26. Racial argument against the death penalty actually
increases support for capital punishment among Whites
because they reject a frame that runs against their prior
beliefs.
Among
Whites, emphasizing
the racial unfairness
argument against the
death penalty creates
a backlash of greater
support for the death
penalty because it runs
against many Whites’
beliefs that the justice
system is color blind.
Among
Blacks, emphasizing
racial unfairness
moves them toward
reduced support for
capital punishment.
27. Comparison of Some Media Effects
Media Effect Type of media Specific Examples Influence
effect varies by:
Agenda setting Increase in the Importance of issue “It’s the economy, stupid” Media trust
amount of news or candidate in 1992 Political knowledge
coverage (e.g., # of changes
stories)
Priming Changes in the Criteria used to “It’s the economy, stupid”
number of stories evaluate a political in 1992
about an issue leader low approval for H.W.
Bush
(Political extension
of agenda setting)
(Issue or Emphasis) Changes in the Criteria used to E.g., Headlines, pictures, Predispositions,
Framing content (frames) of a evaluate the issue etc. Interpret “meaning” Source
story about an issue of story. Competing Frames
(Game vs Policy frame,
Episodic vs Thematic frame)
Direct Persuasion Changes in the Changes in beliefs Negative ads negative Source, Message,
(attitude change): content of a message or attitudes evaluations of a candidate Audience
Characteristics
28. Self-Selection & Selection bias
The problem for Non-experimental designs: causal inference
People are self-selecting the media they choose to consume. Newspaper
readers tend to be more educated (and thus politically informed) than TV
watchers because newspapers largely assumes greater political
information among readers, whereas TV doesn’t.
Problem with causal interpretation: An association between watching
TV and having lower political knowledge is due more to self-selection
than exposure to TV.
One solution is to conduct an experiment where we expose people to
same content but in different mediums (print and TV).
The problem for Experimental designs: generalizability
Random assignment to TV vs. print “fixes” the above problem. But
creates a different type of self-selection problem:
Iyengar (Arceneaux): In the real world, people select different shows and
different media. By randomly assigning people to different exposure
treatments, we may be exaggerating the causal impact of the treatment
in the real world.
Notes de l'éditeur
Whites more supportive of DP in every argument condition.Blacks more responsive to arguments against the death penalty than whites. Whites aren’t receptive to either argument. In fact, not only do whites appear resistant to the innocent argument, create no significant movement among whites, their support for the death penalty actually increases in the racial argument condition. Not only do they reject it, on average, about 12% move strongly in the direction opposite to the argument.