You are on page 1of 14

See

discussions, stats, and author profiles for this publication at:


http://www.researchgate.net/publication/263364332

The Impact of Negative


Campaign Ads

ARTICLE in JOURNAL OF APPLIED SOCIAL PSYCHOLOGY MAY 2012


Impact Factor: 0.63 DOI: 10.1111/j.1559-1816.2012.00912.x

CITATION DOWNLOADS VIEWS

1 79 32

3 AUTHORS, INCLUDING:

Icek Ajzen
University of Massachusetts
106 PUBLICATIONS 41,148
CITATIONS

SEE PROFILE

Available from: Icek Ajzen


Retrieved on: 02 July 2015
bs_bs_banner

The Impact of Negative Campaign Ads

Eric Marks, Mark Manning, and Icek Ajzen1


University of MassachusettsAmherst

We compared negative and positive ads in the context of a fictitious election. Par-
ticipants read a strong or weak message supporting one candidate (positive ad) or
derogating the opposition candidate (negative ad). The strong positive message had
a greater impact on attitudes toward the candidates than the weak positive message,
but message strength had no significant effect for negative messages, suggesting that
positive messages are centrally processed, and negative messages serve mainly as
peripheral cues. Accordingly, a strong positive message was more effective than a
weak positive message, but a weak positive message was less effective than a weak
negative message. We conclude that negative political ads are advisable only when
candidates cannot provide strong arguments to support their candidacy. jasp_912 1280..1292

From William Penns heirs criticizing Benjamin Franklins moral charac-


ter in the 1700s, to John McCain belittling Barack Obamas celebrity in 2008,
attack advertisements have been an integral part of politics since the found-
ing of the Republic. Not only do political campaigns freely utilize personal
attacks, negative forms of advertising often far exceed their positive coun-
terparts. For instance, in the 2004 Presidential race, nearly 75% of incumbent
George W. Bushs advertisements attacked his opponent, John Kerry, while
merely 29% described his own qualifications (Milbank & VandeHei, 2004).
The high prevalence of negative ads in political campaigns is at odds with
the finding that the general public mostly disapproves of personal attacks
against an opponent, and professes to be uninfluenced by such attack ads
(Garramone, 1984; Garramone, Atkin, Pinkleton, & Cole, 1990; Kurtz,
2008). Of course, political strategists tend to embrace the practice of negative
advertising because they believe that it works, and anecdotal evidence
appears to support this view. Thus, an advertising campaign in which the
so-called Swift Boat Veterans for Truth challenged Senator John Kerrys
military record is believed to have tipped the 2004 Presidential election in
favor of their candidate of choice, George W. Bush (Vlahos, 2005). Similarly,
the Willie Horton ad, which insinuated, with racial overtones, that George
H. W. Bushs opponent, Michael Dukakis, was soft on crime, is often cited as

1
Correspondence concerning this article should be addressed to Icek Ajzen, Department of
Psychology, Tobin Hall135 Hicks Way, University of Massachusetts, Amherst, MA 01003-
9271. E-mail: aizen@psych.umass.edu

1280

Journal of Applied Social Psychology, 2012, 42, 5, pp. 12801292.


2012 Wiley Periodicals, Inc.
doi: 10.1111/j.1559-1816.2012.00912.x
IMPACT OF NEGATIVE ADS 1281

the driving force behind Bushs victory in the 1988 Presidential election (CBC
News Online, 2007; Farhi, 2004).
It is also possible, however, to find anecdotal evidence that challenges the
effectiveness of attack ads. In 1993, the Canadian Progressive Conservative
Party produced an advertisement that satirized then-Prime Minster Jean
Chrtiens affliction with Bells palsy. To their dismay, the public considered
the ad to be morally reprehensible and voted in droves against the CPC Party
(CBC News Online, 2007). More recently, Senator John McCains barrage of
advertisements attacking then-Senator Barack Obamas celebrity status, lack
of experience, and questionable personal affiliations did little to help him
gain favor with voters. In fact, some political commentators have speculated
that the ads backfired, repelling voters and creating a groundswell of support
for their intended target (Cooper & Thee, 2008).
In a meta-analysis of research on the impact of negative political cam-
paigns, Lau, Sigelman, and Rovner (2007) reached the following conclusions.
First, compared to positive campaigns, negative campaigns produce a very
small but significant increase in campaign knowledge. Second, negative ads
tend to be somewhat easier to remember than comparable positive ads, but
this effect is inconsistent and neither strong nor always significant. Third,
there is no consistent effect of campaign valence on interest in the campaign.
Finally, and of greatest interest for the present purposes, the data have
provided no clear evidence to support the idea that negative campaigns are
more effective than are positive campaigns.
On the one hand, synthesis of the results of 31 data sets shows that
negative ads tend to undermine attitudes toward the attacked candidate,
making these attitudes somewhat more negative. On the other hand, synthe-
sis of 40 data sets shows that negative ads also tend to lower attitudes toward
the attacker, a backlash that would erase any benefits of the negative ads.
Perhaps most important, there is also no evidence for a systematic effect of
negative versus positive campaigns on voting intentions or actual voting
choice. Based on a synthesis of results from 43 data sets, Lau et al. (2007)
concluded that The research literature does not bear out the proposition
that negative political campaigns work in shifting votes toward those who
wage them (p. 1183). If anything, there was a slight, albeit insignificant
difference in favor of positive campaigns.
Conflicting expectations regarding the potential impact of negative cam-
paigns also emerge from an examination of the social psychological litera-
ture. As a general rule, it has been found that people are more attentive to
negative stimuli (LoBue, 2009), and they tend to weigh negative information
more heavily than equally extreme positive information in the context of
impression formation and interpersonal judgments (Feldman, 1966), such as
assessing the qualifications of job applicants (Bolster, 1961; Crissy & Regan,
1282 MARKS ET AL.

1951) or when forming opinions about the police force (Bayley & Mendel-
sohn, 1968). Studies have found the negativity effect at work in the political
sphere as well. Thus, when asked to evaluate a politician who is described as
possessing an equal number of positive and negative traits, people often rely
disproportionately on the negatives, thereby forming a negative overall
impression (Aragones, 1997; Bunker, 1996; Lau, 1982). Although different
explanations have been offered for the observed negativity bias, it is perhaps
best attributed to the observation thatat least in interpersonal contexts
negative information, because it is generally unexpected, has greater diag-
nostic value than does positive information (Skowronski & Carlston, 1989).
It is not at all clear, however, that findings obtained in an impression-
formation context can be generalized to information encountered in political
campaigns. On the one hand, it is possible that the derogatory and often
combative language featured in attack ads stands out against the backdrop of
generally cordial language in everyday conversation. As a result, ads attack-
ing an opponent may capture the audiences attention and receive more
weight than equally extreme positive ads that favor ones own candidate. On
the other hand, it is also possible that in a political context, candidates are
expected to derogate their opponents. As a result, negative information
would have relatively little diagnostic value and, hence, little impact on
attitudes toward the opposition candidate. Bradley, Angelini, and Lee (2007)
found that negative political ads tend to activate the aversive motivational
system, suggesting that negative ads may backfire. Conversely, these investi-
gators also reported better recognition memory for negative ads.
A similarly equivocal picture emerges from consideration of the elabora-
tion likelihood model (ELM; Petty & Cacioppo, 1986; Petty & Wegener,
1999), the most widely accepted theory of persuasive communication.
According to the ELM, any message can be processed along a continuum
that ranges from low-elaboration peripheral processing to high-elaboration
central processing. In its purest form, peripheral processing entails only
minimal consideration of a messages content and validity. As such, the
communication is evaluated on factors peripheral to the message itself,
such as the communicators appearance (Budesheim & DePaola, 1994), the
number of arguments contained in the message (Petty & Cacioppo, 1984), or
the recipients mood (Mathur & Chattopadhyay, 1991). Although appraising
a messages strength in this way leaves the recipient prone to error, peripheral
processing is a useful way to evaluate persuasive communications without
putting forth significant mental effort.
At the opposite end of the continuum lies central processing, where the
recipients of a persuasive message think carefully about its content, relating
it to their own beliefs and knowledge. If the message contains strong,
cogent arguments, the recipients are likely to generate thoughts that are in
IMPACT OF NEGATIVE ADS 1283

agreement with the advocated position. This favorable elaboration increases


the likelihood that the recipients will form message-congruent attitudes. On
the other hand, if a message that consists of weak, logically unsound argu-
ments is centrally processed, its flaws are easily recognized. The resulting
unfavorable elaboration will reduce the probability that the message will
persuade its recipients.
Although central processing affords recipients a greater degree of accu-
racy in identifying arguments that are strong enough to deserve serious
consideration, it requires considerably more cognitive resources than does
peripheral processing, an effort likely to be expended only if receivers have
the requisite cognitive capacity and motivation (Petty & Cacioppo, 1986).
For example, distractions and time limitations can restrict receivers capacity
for critical thought, thereby preventing them from centrally processing a
persuasive communication. Similarly, even when receivers have the required
cognitive capacity, they are unlikely to centrally process persuasive commu-
nications that do not interest them or that have no personal relevance for
them (see Petty & Cacioppo, 1986).
Applying these ideas to negative political ads, we must consider the dif-
ferent roles any variable can play in the persuasion process (Petty & Wegener,
1998). There are two potential roles that are of particular interest. First,
compared to positive information about ones own candidate, negative infor-
mation about the opposing candidate may motivate receivers to process the
information more carefully; that is, attack ads may undergo more central
processing than positive ads. As a result, negative ads containing strong
arguments should produce more attitude change (i.e., lower attitudes toward
the attacked candidate) than negative ads containing weak arguments. The
effect of argument strength should be smaller in the case of positive messages
that receive a lower level of cognitive elaboration.
A second possibility, however, is that people tend to be averse to attack
ads and, therefore, process such ads less centrally than comparable positive
messages. In this case, the negative tone of the message may nevertheless
serve as a negative peripheral cue, lowering attitudes toward the attacked
candidate, irrespective of argument strength. In contrast, positive messages
because they are centrally processedwould have a greater effect on atti-
tudes when they contain strong, rather than weak arguments.
The present study is designed to explore these alternative hypotheses.
Participants were exposed either to a positive message describing the favor-
able characteristics of the communicating candidate or a negative message
about the opposing candidate. Each message type contained either strong
arguments or weak arguments. In addition, to explore the role of cognitive
capacity, the messages were administered either under cognitive load or no
cognitive load.
1284 MARKS ET AL.

Method

Participants

Study participants were 214 students (158 females, 56 males) who were
enrolled in undergraduate psychology courses. The students ranged in age
from 17 to 26 years (M = 19.8 years, SD = 8.6).

Procedure

Upon their arrival, the participants were seated and handed an informed
consent form to read and sign. The study was described as dealing with how
voters form impressions of political candidates. Participants were given a
packet that contained an advertisement for Candidate A, one of two
candidates for Governor of a midwestern state, a questionnaire, and an
optical response form. They were instructed to read the advertisement and to
complete the attached questionnaire. The participants were informed that no
time limit would be imposed upon them and to work at a comfortable pace.
To ensure the anonymity of their responses, participants were instructed to
refrain from recording their names on any of the testing materials.
Participants who had been randomly assigned to the cognitive load con-
dition were, at this point, handed a slip of paper with an eight-digit number
printed on one side. They were asked to memorize the number and were given
45 s to do so. After 45 s had elapsed, the slips were collected. This part of the
procedure was omitted for participants in the no-load condition. All partici-
pants then read the ad and responded to the questionnaire on the attached
optical scan sheet. At the conclusion, participants in the cognitive load
condition were asked to recall the number they had memorized. Finally, all
participants were handed a debriefing statement, thanked for their participa-
tion, and dismissed.

Materials

Political ads. The participants read one of four political ads that con-
tained either strong or weak arguments in support of the originating Candi-
date A (positive ad) or in opposition to the alternative Candidate B (negative
ad). The ads were formulated in the tone and style of a press release produced
by campaign strategists for distribution to media outlets. Each ad was
approximately two pages of single-spaced text and contained six major argu-
ments dealing with the candidates positions on energy; the cost of higher
IMPACT OF NEGATIVE ADS 1285

education, healthcare, and crime; as well as the candidates leadership abili-


ties and adherence to family values. In the strong messages, each of the six
arguments was supported by relevant and credible evidence; whereas in the
weak messages, the supporting evidence was of little persuasive power. For
example, in the strong positive message, the argument dealing with energy
was formulated as follows:

As a country, and as a state, we are too heavily dependent on


unreliable and increasingly scarce foreign oil and gas. Research
indicates that the price of oil and natural gas will continue to
increase and continued use of these fossil fuels will produce
unacceptable levels of pollution and global warming.
My energy policy promotes the development of hydrogen fuel
cells, wind power, solar power, and other sources of renewable
energy for use in our state. Respected scientists and energy
experts believe that, with a concerted effort, these alternative
sources of energy can be developed within a few short years to
power cars and homes. This will drastically reduce greenhouse
gas emissions and will bring our state closer to energy indepen-
dence. Moreover, it is estimated that by 2015, renewable energy
will cost consumers nearly 25% less money than fossil fuels.

In the weak positive message, the first paragraph was the same, but the
second paragraph read as follows:

My energy policy promotes the development of hydrogen fuel


cells, wind power, solar power, and other sources of renewable
energy for use in our state. Although they are unlikely to
decrease our dependence on fossil fuels in the foreseeable
future, some scientists have argued that proper use of these
alternative energy sources could improve upon the already pic-
turesque beauty of our states landscape. Electricity-generating
windmills, for example, can be painted to resemble enormous
sunflowers. Similarly, the heat radiating from nuclear power
facilities will allow diverse populations of tropical flowers to
thrive around its perimeter.

The negative messages paralleled the positive messages. They used the
same six arguments as the positive message, but described Candidate Bs
opposition to the proposed policies and his weakness in terms of leadership
ability and family values. For example, in the strong negative message, the
1286 MARKS ET AL.

opening paragraph was the same as in the strong positive message, but the
second paragraph described Candidate Bs opposition to the development of
alternative energy sources:

Yet, my opponent opposes the development of hydrogen fuel


cells, wind power, solar power, and other sources of renewable
energy for use in our state. This is especially worrisome because
respected scientists and energy experts believe that, with a con-
certed effort, these alternative sources of energy could soon be
developed to power our cars and homes. Utilizing renewable
energy sources would drastically reduce our states greenhouse
gas emissions and would bring us closer to energy indepen-
dence. Furthermore, new research indicates that my opponents
plan for continued reliance on fossil fuels will cost consumers
almost 25% more money than renewable energy by the year
2015.

Similarly, in the weak negative message, the opening paragraph was the same
as in the weak positive message, but the second paragraph again dealt with
Candidate Bs opposition to alternative energy:

Yet, my opponent opposes the development of hydrogen fuel


cells, wind power, solar power, and other sources of renewable
energy for use in our state. Even though these alternative
sources of energy are unlikely to significantly decrease our
dependence on fossil fuels, several respected scientists believe
that utilizing them could improve upon the already picturesque
beauty of our states landscape. By advocating the continued
use of fossil fuels, my opponent is marring our beautiful state
with large, unsightly smokestacks instead of beautifying it with
windmills that can be painted to resemble enormous sunflowers.

Questionnaire. After reading one of the four campaign ads, the partici-
pants proceeded to complete a questionnaire designed to assess their attitudes
toward the two candidates. The first part of the questionnaire contained 13
attitudinal statements about Candidate A. Some of the statements dealt with
the candidates policy positions. Thus, participants were asked whether they
thought that if Candidate A were elected, state residents would benefit from
his energy policies; whether Candidate A would fight to reduce the cost of
higher education; and whether the cost of healthcare would decrease. Other
statements had to do with personal qualities, asking participants to indicate
whether they thought that Candidate A is able to relate to the concerns of the
IMPACT OF NEGATIVE ADS 1287

average American, is dishonest, is a weak leader, and whether they would


vote for the candidate. This part was followed by the same set of 13 questions
with respect to Candidate B.2 All responses were rated on a 5-point scale
ranging from 1 (strongly disagree) to 5 (strongly agree).

Results

Separate item analyses of the attitudinal items for the two candidates
revealed high levels of internal consistency, with the exception of one item
(asking whether the candidate was aggressive), which was eliminated. The
remaining 12 items had internal consistency coefficient alphas of .90 and .81
for Candidates A and B, respectively. Scores on the 12 items were averaged
to produce a measure of attitude toward each of the two candidates. This
measure served as the dependent variable in all analyses.
The attitude scores were submitted to a repeated-measures ANOVA, with
message strength (weak vs. strong), message valence (negative vs. positive),
and cognitive load (no vs. yes) as between-subjects factors; and candidate (A
vs. B) as a repeated-measures within-subjects factor. Because of missing data
for 2 participants, the sample size was reduced to 212, with between 24 and 28
participants per condition. The ANOVA revealed few significant effects as a
result of cognitive load, and because this manipulation was exploratory and
not central to our hypotheses, it will not be considered further.3 The mean
attitudes and standard deviations are presented in Table 1.
The results revealed significant main effects for the factors of message
strength, message valence, and candidate. As would be expected because the
messages were designed either to support Candidate A or to derogate Can-
didate B, attitudes toward Candidate A (M = 3.40) were more favorable than
were attitudes toward Candidate B (M = 2.82), F(1, 208) = 92.03, p < .001. In
addition, strong messages produced more favorable attitudes toward the two
candidates (M = 3.26) than did weak messages (M = 2.96), F(1, 208) = 49.50,
p < .001; and positive messages produced more favorable attitudes
(M = 3.19) than did negative messages (M = 3.03), F(1, 208) = 12.27,
p < .001. The ANOVA also revealed a significant message strength by
message valence interaction, F(1, 208) = 38.30, p < .001. Message strength
had a significant effect on mean candidate ratings, t(109) = 9.58, p < .001,

2
The final section of the questionnaire dealt with six more general issues, such as how
comfortable the participants felt reading the statement and whether they found the statement
offensive. As these questions were not germane to the studys hypotheses and did not reveal any
interesting effects, they will not be considered further.
3
There were no significant effects resulting from gender.
1288 MARKS ET AL.

Table 1

Means of Attitudes Toward the Two Candidates


Under Different Experimental Conditions

Candidate A Candidate B

M SD M SD
Weak message
Negative 3.37 0.65 2.66 0.41
Positive 2.72 0.75 3.07 0.40
Strong message
Negative 3.57 0.73 2.53 0.44
Positive 3.94 0.43 3.01 0.39
Note. N = 213.

only for positive messages. Its effect was not significant for negative mes-
sages, t(100) = 0.59, ns.
These effects involving message strength and valence, however, are not
particularly meaningful because the effects of these variables would be
expected to differ for the two candidates. More interesting, therefore, are the
interactions with the candidate factor. The message strength by candidate
interaction was highly significant, F(1, 208) = 44.84, p < .001. Model param-
eters reveal that strong messages produced significantly more favorable atti-
tudes toward Candidate A (M = 3.74; b = 1.22, SE = .12, z = 9.91, p < .05)
than did weak messages (M = 3.04), but message strength had no significant
effect (b = -.06, SE = 1.08, z = .79, ns) on attitudes toward Candidate B
(Ms = 2.78 and 2.87, respectively). This finding suggests that participants
centrally processed message arguments in forming their attitudes toward
Candidate A, the originator of the messages, and were thus more influenced
by strong arguments than by weak arguments. However, they appear to have
largely disregarded message content when forming their attitudes toward the
opposing candidate.
The message valence by candidate interaction was also significant, F(1,
208) = 23.33, p < .001. Although the effect was relatively small, positive mes-
sages produced more favorable attitudes toward Candidate A (M = 3.33)
than toward Candidate B (M = 3.04), t(109) = 3.17, p < .01. A much stronger
effect was produced by the negative messages, which resulted in relatively
unfavorable attitudes toward Candidate B (M = 2.60) and relatively favor-
able attitudes toward Candidate A (M = 3.46), t(101) = 9.14, p < .001. The
IMPACT OF NEGATIVE ADS 1289

Figure 1. Attitudes toward Candidates A and B as a function of message strength and message
valence. Error bars indicate standard errors.

difference in attitudes toward the two candidates was significantly greater for
negative messages than for positive messages, t(210) = 4.09, p < .001. This
result suggests that, overall, negative messages gave the originating candidate
a greater advantage than did positive messages.
The results most relevant for the questions raised in this study can be
found in the three-way interaction among message strength, message valence,
and candidate (see Figure 1). It can be seen in Figure 1 that the strong
messages, whether positive or negative, produced large discrepancies in favor
of Candidate A. In each case, the difference was highly significant for the
strong negative messages, t(50) = 7.49, p < .001; and for the strong positive
messages, t(56) = 11.46, p < .001.
The weak messages, however, reveal a different pattern of influence. The
weak positive message in which Candidate A provided unconvincing support
for his arguments was counterproductive, resulting in significantly less favor-
able attitudes toward Candidate A than toward Candidate B, t(52) = -2.69,
p < .01. In contrast, the weak negative message produced attitudes signifi-
cantly in favor of Candidate A, t(50) = 5.50, p < .01. These findings show that
when a candidate can provide strong supportive evidence for his arguments,
his message will be effective, whether it deals positively with his own positions
and qualifications or negatively with those of his opponent. However, when
the supportive evidence is weak, the candidate is better off using a negative
attack ad.
1290 MARKS ET AL.

Discussion

The results of the present study offer a number of interesting, though


tentative conclusions regarding the possible impact of negative campaign
ads. First, not surprisingly, messages with strong supportive evidence had an
advantage over messages with weak supportive arguments. Second, overall,
the messages had more of an impact on attitudes toward their originator
(Candidate A) than on his opponent (Candidate B). Third, there was no
evidence to suggest that negative messages derogating an opponent are cen-
trally processed. On the contrary, the results show that positive, rather than
negative messages were processed centrally: A strong positive message had a
significantly greater impact on attitudes toward the candidates than did a
weak positive message; there was no significant difference between strong and
weak negative messages. This is the first indication that negative messages
can be quite effective, even when they contain weak arguments although,
overall, positive messages were more effective than were negative messages.
Finally, the three-way interaction between message strength, message
valence, and candidate revealed that most message types gave Candidate A
an advantage over Candidate B, but this was not the case for a weak positive
message. When Candidate A provided weak arguments in support of his
candidacy, the message actually lowered attitudes toward him, giving Can-
didate B the advantage.
In sum, we can conclude that in our study, positive messages were cen-
trally processed, given an advantage to a positive message with strong argu-
ments. Negative messages were not centrally processed and seem to have
served mainly as a negative peripheral cue. Because participants did not pay
much attention to the content of such messages, the negative message was
quite effective, even when it contained weak arguments.
We must, however, be careful when trying to generalize these findings to
real political contexts. First, the use of a convenience sample of college
students limits our ability to generalize the findings to the populations at
large. Second, we used messages in the form of hypothetical press releases to
convey information about the candidates in a hypothetical election. In real-
life contests, press releases are only one form of promoting a candidate or
derogating the opponent. Many negative appeals are presented in short
television commercials that do not require a great deal of processing. Also,
real-world negative advertising often employs third-party surrogates or
sources presented as independent to attack the opponent. This may serve to
insulate the attacker, who can maintain a positive message while at the same
time employing negative attack ads. Moreover, voters are exposed to a large
number of ads from both camps. All of these factors may affect the impact of
negative versus positive messages. More research is needed to follow up on
IMPACT OF NEGATIVE ADS 1291

our finding that positive messages may be processed more centrally than are
negative messages and to test the generalizability of this finding to real-world
situations.

References

Aragones, E. (1997). Negativity effect and the emergence of ideologies.


Journal of Theoretical Politics, 9, 189210.
Bayley, D. H., & Mendelsohn, H. (1968). Minorities and the police: Confron-
tation in America. New York: Free Press.
Bolster, B. I. (1961). The reaction of interviewers to favorable and unfavor-
able information. Journal of Applied Psychology, 45, 97103.
Bradley, S. D., Angelini, J. R., & Lee, S. (2007). Psychophysiological and
memory effects of negative political ads: Aversive, arousing, and well
remembered. Journal of Advertising, 36, 115127.
Budesheim, T. L., & DePaola, S. J. (1994). Beauty or the beast? The effects of
appearance, personality, and issue information on evaluations of political
candidates. Personality and Social Psychology Bulletin, 20, 339348.
Bunker, T. (1996). Right on the money. Brandweek, 37, 3033.
CBC News Online. (2007, January 29). Political attack ads. Retrieved
December 1, 2008, from www.cbc.ca/news/interactives/who-attackads/
Cooper, M., & Thee, M. (2008, October 14). Poll Says McCain is hurting his
bid by using attacks. Retrieved February 20, 2009, from www.
nytimes.com/2008/10/15/us/politics/15poll.html?partner=rssnyt
Crissy, W. J., & Regan, J. (1951). Halo in the employment interview. Journal
of Applied Psychology, 35, 338341.
Farhi, P. (2004, September 7). Two political ads share more than fame and
controversy. Retrieved February 19, 2009, from www.washingtonpost.
com/wp-dyn/articles/A1151-2004Sep6.html
Feldman, S. (1966). Motivational aspects of attitudinal elements and their
place in cognitive interaction. In S. Feldman (Ed.), Cognitive consistency
(pp. 76114). San Diego, CA: Academic Press.
Garramone, G. M. (1984). Voter responses to negative political ads. Jour-
nalism Quarterly, 61, 250259.
Garramone, G. M., Atkin, C. K., Pinkleton, B. E., & Cole, R. T. (1990).
Effects of negative political advertising on the political process. Journal of
Broadcasting and Electronic Media, 34, 299311.
Kurtz, H. (2008, January 6). People may dislike attack ads, but the messages
tend to stick. Retrieved February 25, 2009, from www.washingtonpost.
com/wp-dyn/content/article/2008/01/05/AR2008010502196.html
Lau, R. R. (1982). Negativity in political perception. Political Behavior, 4,
353377.
1292 MARKS ET AL.

Lau, R. R., Sigelman, L., & Rovner, I. B. (2007). The effects of negative
political campaigns: A meta-analytic reassessment. Journal of Politics, 69,
11761209.
LoBue, V. (2009). More than just another face in the crowd: Superior detec-
tion of threatening facial expressions in children and adults. Developmen-
tal Science, 12, 305313.
Mathur, M., & Chattopadhyay, A. (1991). The impact of moods generated
by television programs on responses to advertising. Psychology and Mar-
keting, 8, 5977.
Milbank, D., & VandeHei, J. (2004, May 31). From Bush, unprecedented
negativity. Washington Post, p. A01.
Petty, R. E., & Cacioppo, J. T. (1984). The effects of involvement on
responses to argument quantity and quality: Central and peripheral
routes to persuasion. Journal of Personality and Social Psychology, 46,
6981.
Petty, R. E., & Cacioppo, J. T. (1986). Communication and persuasion:
Central and peripheral routes to attitude change. New York: Springer-
Verlag.
Petty, R. E., & Wegener, D. T. (1998). Attitude change: Multiple roles for
persuasion variables. In D. T. Gilbert, S. T. Fiske, & G. Lindzey (Eds.),
The handbook of social psychology (4th ed., Vol. 1, pp. 323390). Boston:
McGraw-Hill.
Petty, R. E., & Wegener, D. T. (1999). The elaboration likelihood model:
Current status and controversies. In S. Chaiken & Y. Trope (Eds.), Dual-
process theories in social psychology. (pp. 4172). New York: Guilford.
Skowronski, J. J., & Carlston, D. E. (1989). Negativity and extremity biases
in impression formation: A review of explanations. Psychological Bulletin,
105, 131142.
Vlahos, K. B. (2005, February 16). Conservatives laud Swift Boat veterans.
Retrieved February 20, 2009, from www.foxnews.com/story/0,2933,
147728,00.html

You might also like