Content uploaded by Nicole Ernst
Author content
All content in this area was uploaded by Nicole Ernst on Aug 16, 2017
Content may be subject to copyright.
International Journal of Communication 11(2017), 3265–3285 1932–8036/20170005
Copyright © 2017 (Nicole Ernst, Rinaldo Kühne, and Werner Wirth). Licensed under the Creative
Commons Attribution Non-commercial No Derivatives (by-nc-nd). Available at http://ijoc.org.
Effects of Message Repetition and Negativity on
Credibility Judgments and Political Attitudes
NICOLE ERNST
University of Zurich, Switzerland
RINALDO KÜHNE
University of Amsterdam, The Netherlands
WERNER WIRTH
University of Zurich, Switzerland
Research on the truth effect has demonstrated that statements are rated as more
credible when they are repeatedly presented. However, current research indicates that
there are limits to the truth effect and that too many repetitions can decrease message
credibility. This study investigates whether message negativity contributes to this
boomerang effect and whether the interaction of credibility and negativity influences
political attitudes. These assumptions were tested in an online experiment in which the
frequency of exposure to political campaign posters and message negativity were
manipulated. The results show that negativity on political campaign posters functions as
a crucial moderator, especially in combination with high-frequency exposure. Repeatedly
presented negative posters resulted in a more negative attitude toward the presented
political issue, which was mediated by a decrease in credibility judgments.
Keywords: truth effect, campaign communication, negative campaigning, political
campaign poster
“Sixty-two thousand four hundred repetitions make one truth.” This famous quote by Aldous
Huxley—author of Brave New World—implies that the more a statement is repeated, the more credible it
is perceived to be. Hence, message repetition is an often-applied strategy in political campaigns. By
repeatedly confronting the audience with the same claim, campaigners aim to increase the persuasive
impact. The “truth effect” can explain these effects. Accordingly, recipients are more likely to hold a
statement as true if they hear the statement multiple times. Empirical findings corroborate the notion that
the repeated presentation of political messages increases message credibility and attitude change (Koch &
Zerback, 2013; Miller, 1976).
Nicole Ernst: n.ernst@ipmz.uzh.ch
Rinaldo Kühne: R.J.Kuhne@uva.nl
Werner Wirth: w.wirth@ipmz.uzh.ch
Date submitted: 2016‒10‒19
3266 Nicole Ernst, Rinaldo Kühne, and Werner Wirth International Journal of Communication 11(2017)
However, previous studies have not sufficiently considered the role of message characteristics in
the truth effect. Most studies embedded claims in neutral messages and then manipulated the number of
times participants were presented with the message (Dechêne, Stahl, Hansen, & Wanke, 2010). In
political campaigns, messages are rarely neutral; often, they are negative and may include attacks on
political opponents. This raises the question of whether the truth effect also unfolds when campaigners
repeatedly confront their voters with negative messages.
Our aim in the study was to investigate the moderating influence of message negativity on the
truth effect. For this purpose, we first discuss the truth effect and review research on the effects of
repeated exposure to political messages. Second, we address factors that can reverse the truth effect and
explain why message negativity may diminish the truth effect. Based on this literature review, we
formulated hypotheses about how the repeated exposure to campaign posters affects credibility
judgments and attitudes and how the negativity of posters moderates the effects of repetition. We then
describe the design and results of an experiment and discuss the results, implications, and directions for
future research.
The Effects of Message Repetition
Message credibility refers to “perceptions of believability, either of the source or of the source’s
message” (Metzger, Flanagin, Eyal, Lemus, & Mccann, 2003, p. 302). Research has identified four
dimensions that positively influence message credibility (Metzger et al., 2003): message structure (e.g.,
how well the message is organized), message content (e.g., how well the message is written), message
delivery (e.g., how quickly the message is communicated), and presentation style (e.g., how often the
message is repeated). One hypothesis that focuses on the relationship between a message’s presentation
style and the message’s credibility is the truth effect. According to the truth effect, the number of times a
message is presented influences how recipients evaluate the message’s credibility. For instance, when a
statement such as “Jimmy Carter was the only divorced U.S. president” is presented, the majority of
people cannot be sure whether the statement is true. To form an opinion about such a statement, people
tend to use heuristic cues, such as the statement’s source or the context in which the statement was
presented (Dechêne et al., 2010). According to the truth effect, the subjective impression that the
statement about Carter is true increases when people encounter the statement repeatedly (Hasher,
Goldstein, & Toppino, 1977; Schwartz, 1982).
This effect can be explained through two different processes. First, the truth effect can be
explained through a primarily unconscious and memory-based process. Accordingly, the repeated
presentation of a statement leads the recipient to mistakenly believe that she/he has already heard the
statement from another source (Arkes, Boehm, & Xu, 1991; Arkes, Hackett, & Boehm, 1989). Because
humans learn in the socialization process to trust statements that are expressed by several different and
independent sources, this promotes increased message credibility (Brown & Nix, 1996; Koch & Zerback,
2013). Second, the truth effect can be explained through a trust-based process. Repeated exposure to a
message is assumed to increase the “processing fluency,” which is defined as the metacognitive
experience of ease during information processing (Dechêne et al., 2010). The easier and more fluently
that information can be processed, the more credible the information appears, regardless of the
International Journal of Communication 11(2017) Effects of Message Repetition and Negativity 3267
statement’s content (Reber & Schwarz, 1999). For instance, Reber and Schwarz (1999) showed that
processing fluency can be increased by improving the visual contrast of a statement on a screen. It is
irrelevant whether the repeated statement is actually true or false (Begg, Anas, & Farinacci, 1992; Brown
& Nix, 1996) as long as the statement is sufficiently ambiguous so that the participants are uncertain
about the statement’s truth. Otherwise, the truth of the statement would be judged based on the
participants’ knowledge rather than on fluency (Dechêne et al., 2010).
The notion that repeated message exposure positively affects processing fluency is closely related
to the mere exposure effect. The exposure to a previously unknown stimulus facilitates the subsequent
processing of the stimulus (Bornstein, 1989), and this “mere repeated exposure of the individual to a
stimulus is a sufficient condition for the enhancement of his attitude toward it” (Zajonc, 1968, p. 1). Thus,
research on the truth effect emphasizes how processing fluency enhances credibility, and research on the
mere exposure effect focuses on the implications of processing fluency for the evaluation and formation of
attitudes toward stimuli.
As research on mere exposure effect indicates, message repetition not only results in higher
credibility ratings of a message but also can influence attitudes of the recipients. The mere exposure effect
suggests that message repetition can have a direct effect on attitudes. In addition, as research on attitude
change and credibility shows, messages of credible communicators typically lead to more attitude change
than messages from sources with low credibility (Choo, 1964; Hovland & Weiss, 1951). Accordingly, it is
plausible to assume that message credibility mediates the effect of message repetition on attitude change:
The more a message is repeated, the higher its credibility and the stronger its impact on attitudes will be.
The truth effect hypothesis is a useful concept to understand the effects of message repetition in
political communication. The hypothesis suggests that message repetition is a crucial strategy in political
campaigns because it can boost the credibility of political claims and election programs and increase the
persuasive impact of a candidate or party. Still, the framework has rarely been applied in political
communication research, although many studies in this field are based on the assumption that message
repetition increases persuasive impact and attitude change (Bartels, 2006).
An exception is the study by Koch and Zerback (2011), who studied the effects of repeatedly
presented campaign slogans. In their experiment, they found that if previous knowledge existed, exposure
frequency affected the credibility of the slogans. More precisely, credibility continued to increase until
three repetitions, but decreased after six statements. In line with these findings are the results of a study
by Becker and Doolittle (1975), who investigated how repeated political radio advertisements affect the
evaluation of candidates. They found that a moderate frequency of five political radio advertisements
resulted in a higher affective candidate evaluation compared with the low frequency of only two
repetitions.
Previous research on the effects of repetition on credibility judgments in the field of political
communication indicates that the truth effect has its limits. Increasing the frequency of exposure does
initially increase credibility and attitude change. However, overexposure to a political message can have
3268 Nicole Ernst, Rinaldo Kühne, and Werner Wirth International Journal of Communication 11(2017)
detrimental effects on credibility and attitude change, which implies that certain boundary conditions exist
for the truth effect to unfold.
Boundary Conditions of the Truth Effect and the Moderating Role of Message Negativity
In a meta-analysis of 51 studies, Dechêne et al. (2010) showed that the truth effect is a robust
effect that holds for different message and statement types. Research shows that the effect occurs in
trivia statements (Schwartz, 1982), opinion statements (Arkes et al., 1989; Koch & Zerback, 2013),
political election campaign slogans (Koch & Zerback, 2011), and product-related claims (Hawkins & Hoch,
1992; Johar & Roggeveen, 2007). In addition, the truth effect is confirmed whether there is a delay of
minutes (Arkes et al., 1989; Begg et al., 1992) or weeks (Hasher et al., 1977) between the repetitions of
the statement and whether the statement is presented orally (Hasher et al., 1977) or in text (Hawkins &
Hoch, 1992).
Despite the robustness of the effect, several studies indicate that the strength of the truth effect
is affected by specific moderators. One crucial, but in research mostly neglected, moderator is the
frequency of message repetition. Most studies that have established the truth effect presented the
statements up to two times (Dechêne et al., 2010). Fewer studies have employed higher repetition rates.
Hasher et al. (1977) and Gigerenzer (1984) presented statements three times, but could not prove a
significant increase in the effect after the second presentation. Arkes et al. (1991) presented the
statements six times, but found that credibility significantly increased only after the second exposure to
the statement. In each of these studies, at least one week elapsed between the presented statements.
Koch and Zerback (2011) studied the effect of one, three, and six repetitions across a few minutes and
found that the credibility increased until three repetitions, but decreased again after six repetitions.
Hawkins, Hoch, and Meyers-Levy (2001) presented a statement up to four times across a few minutes,
finding a significant decrease in the effect after the statement was presented two times. These results first
suggest that message repetition can increase credibility, but that boomerang effects can occur when a
message is presented too often. Second, the optimal number of message repetitions seems to vary across
studies. Notably, extant research does not provide a clear explanation for why the optimal number of
repetitions varies.
An explanation may be found in additional variables that have been shown to moderate the truth
effect. It has been suggested that characteristics and psychological states of the recipient (Arkes et al.,
1989; Koch & Zerback, 2011) and characteristics of the message (Unkelbach, 2007) are important
determinants. A crucial message characteristic in political communication research is negativity, which is a
commonly used and highly successful strategy in political campaigns (Ansolabehere & Iyengar, 1995). Lau
and Pomper (2002) define negative campaigning as “talking about the opponent—his or her programs,
accomplishments, qualifications, associates, and so on—with the focus, usually, on the defects of these
attributes” (p. 46). It includes all form of attacks on the opponent and describes a technique in which
politicians and parties focus on the supposed weaknesses of their opponents rather than on their own
strengths (Brettschneider, 2008; Lau & Rovner, 2009). Ansolabehere and Iyengar (1995) describe the
popularity of this tactic as follows: “In politics, the best defense is a strong offense” (p. 116). In political
news media, negativity is furthermore one of the most relevant news values (e.g., Galtung & Ruge, 1965)
International Journal of Communication 11(2017) Effects of Message Repetition and Negativity 3269
and an often-applied strategy by journalists (e.g., Esser, Engesser, Matthes, & Berganza, 2016).
Negativity is prevalent in contemporary politics and news media, and it is an ongoing debate whether
negative campaigns harm or benefit candidates, voters, and the political process itself (e.g., Geer, 2006).
Accordingly, the empirical evidence about the effects of negative political campaigns is inconclusive (Allen
& Burrell, 2002; Lau, Sigelman, & Rovner, 2007). Several studies have shown that negativity decreases
voter turnout (Ansolabehere & Iyengar, 1995) and lowers candidates’ evaluations (Pentony, 1998). Other
empirical findings indicate positive effects of negativity such as higher audience recall (Kahn & Kenney,
2000) or increased political participation (Martin, 2004).
More important, the interactive effects of message negativity and repetition on credibility
judgments have not been intensively investigated. Extant research only indicates that the repetition of
negative messages can negatively impact evaluations of the communicator and likelihood of voting
(Fernandes, 2013; Reinhard, Schindler, Raabe, Stahlberg, & Messner, 2014). For instance, Fernandes
(2013) investigated the effects of repeated negative campaign ads on candidate evaluations and likelihood
of voting. She found that evaluations of the candidate who sponsored the negative, attacking ad within a
TV program followed an inverted-U curve. After three repetitions, the evaluation of the candidate
sponsoring the ad and the likelihood of voting for that candidate increased (compared with a single ad),
but it decreased again after five repetitions if the ads were placed close together within the commercial
block. No significant effects were reported for the target actor in this massive presentation condition. If
the negative political ads were spread out across the entire TV program (three commercial blocks), the
evaluation of the candidate sponsoring the ad and the likelihood of voting for that candidate increased
with every repetition. In the spaced presentation condition, both evaluation of the target and likelihood for
voting for the target decreased.
However, we lack empirical evidence that clarifies whether such patterns also hold for credibility
judgments and whether the trajectories are different for negative and nonnegative messages. A concept
that may help to answer these questions is the attention bias hypothesis. According to this, humans pay
more attention to negative information and accordingly negative stimuli elicit higher levels of attention
(Lang, Newhagen, & Reeves, 1996; Smith et al., 2006). Attention biases have been shown to unfold in the
processing of political information: When confronted with negative political information, recipients are
more responsive and attentive toward the stimulus and process the content more intensively (Soroka &
McAdams, 2015). The attention bias hypothesis suggests that the optimal number of repetitions should be
different for negative and nonnegative messages: Due to the higher level of attention and the more
intensive processing of the information, negativity results in the earlier recognition of the persuasive
attempt, and this recognition then, according to Koch and Zerback (2013), decreases the credibility of the
communicator and/or the message. In line with this proposition, Fernandes (2013) showed that with
massive repetition of negative messages, the candidate’s evaluation, and likelihood for voting of the
sponsor decreased after too many repetitions. Thus, the trajectories of the effect of message repetition on
credibility judgments should be different for negative and nonnegative messages: Repeating a
nonnegative or negative message a few times should positively affect credibility. However, when exposure
frequency is further increased, the credibility associated with the negative message should decrease,
whereas it should further increase for the nonnegative message. In short, the optimal number of
repetitions is likely to be lower for negative than for nonnegative messages.
3270 Nicole Ernst, Rinaldo Kühne, and Werner Wirth International Journal of Communication 11(2017)
The Influence of Credibility on Persuasion and Attitude Change
Effects of message repetition and negativity on credibility judgments are of particular relevance
in political campaigns because credibility is a crucial determinant of political attitudes and behaviors. Dual
process models (Chaiken, 1980; Petty & Cacioppo, 1986) suggest that credibility is a crucial heuristic that
leads to more compliant information processing, less reactance, and more positive attitudes. Accordingly,
it has been shown that messages of credible communicators lead to more attitude change than do
messages from sources with low credibility (Choo, 1964; Hovland & Weiss, 1951; Powell, 1965). Thus, it
is plausible that message credibility mediates the effect of message repetition on attitude change.
Depending on the negativity of a message, the frequency of exposure may increase or decrease credibility
judgments, which, in turn, influence the evaluations and attitudes.
The Effects of Repeated Exposure to Neutral and Negative Campaign Posters
Political campaign posters are “large sheets printed for outdoor display” (Seidman, 2008, p. 5)
that are used to promote political candidates or communicate political messages. We chose to focus on
political campaign posters for three reasons. First, they are a traditional medium of political
communication, which plays an integral part in contemporary campaigns (Seidman, 2008). Second,
especially in election times, voters experience a high degree of poster exposure; thus, the repeated
presentation of posters in an experimental context warrants a certain degree of ecological validity. Third,
the truth effect has successfully been proven with statements on campaign posters (Koch & Zerback,
2011).
During election times, voters are generally confronted with a tremendous number of political
campaign posters, which advertise for a candidate or argue in favor of (or against) a specific political
issue. Especially in elections campaigns on complex policy issues, voters are hardly convinced by the
statement if only confronted with it once. A moderate frequency exposure is needed: Voters tend to
process the advertised information on the policy issue more fluently and rate the statement as familiar,
which eventually results in a higher credibility judgment of the presented statement. Following the truth
effect hypothesis, we assumed that being exposed to a campaign poster with a political factual statement
a moderate number of times should increase the evaluation of the message’s credibility, compared with
being exposed to the poster and the statement only once (e.g., Dechêne et al., 2010). Research in
political campaign communication has corroborated these findings (e.g., Becker & Doolittle, 1975; Koch &
Zerback, 2011), which led to our first hypothesis:
H1: Being exposed to a factual statement on a political campaign poster a moderate number of times
leads to a higher credibility of the statement than being exposed to the statement only once.
As discussed, however, there are limits to the truth effect, and a high number of repetitions can
reduce message credibility (e.g., Arkes et al., 1991; Fernandes, 2013; Hawkins et al., 2001; Koch &
Zerback, 2011). More important, we argue that the negativity of a campaign poster influences whether a
high number of repetitions has a positive or a negative effect on credibility. According to the attention bias
hypothesis, voters pay more attention to negative campaign posters than to nonnegative or neutral
International Journal of Communication 11(2017) Effects of Message Repetition and Negativity 3271
posters. Consequently, the optimal number of repetitions should be different for negative and for
nonnegative campaign posters. Because voters are more responsive and attentive and process the
political slogan of a negative campaign poster more intensively, they recognize the persuasive attempt
earlier. Consequently, a high number of repetitions, compared with a moderate number of repetitions,
does not further increase the credibility of the negative poster’s political statement, but rather decreases
the credibility. This is in line with the findings of Fernandes (2013), whose study showed that the
candidate’s evaluation and vote likelihood decreased after a moderate repetition if the negative ads were
presented closely together. In contrast, if a poster is nonnegative or neutral, it is processed less
intensively, and more repetitions are required until the persuasive attempt is recognized. Accordingly,
increasing the number of repetitions from moderate to high should further increase the credibility of the
poster’s statement. We thus formulated the following interaction hypothesis:
H2: The effect of being exposed to a factual statement on a political campaign poster a high number of
times instead of a moderate number of times depends on the poster’s negativity: (a) Increasing
exposure to a negative campaign poster decreases the statement’s credibility; (b) increasing
exposure to a nonnegative or neutral campaign poster increases the statement’s credibility.
Finally, as discussed, credibility is a crucial determinant of political attitudes, and messages with
a high credibility have a stronger impact on attitudes than messages with a low credibility (e.g., Choo,
1964; Powell, 1965). Accordingly, we assumed that credibility functions as a mediator of the effects of
repeated exposure to campaign posters about a political issue on political attitudes toward the same issue.
As noted, repeated exposure to a campaign poster that includes a message about a political issue will
increase the credibility of the poster’s message. In turn, the increased credibility of the poster’s message
will be positively associated with the attitude toward the political issue addressed on the poster.
Therefore, the following hypothesis was formulated:
H3: The effect of message repetition on the attitude toward the presented issue is mediated by the
perceived credibility of the statement. The more credible a factual statement on a campaign poster
about a political issue is perceived, the more positive is the attitude toward the political issue.
Method
An online experiment using a 3 2 between factorial design was used to test the hypotheses. The
first factor was the frequency of the campaign poster presentation (one, three, six), and the second factor was
the design of the campaign poster (neutral, negative). The participants were split into six groups and
presented with either a neutral or negative poster that showed an identical statement on a political issue one,
three, or six times. In addition, the participants were presented with five filler posters that did not promote
any political position. The filler posters were used to ensure that participants were not exclusively exposed to
the target posters and that there was a small time delay between the contacts with the target posters. With
the repetition factors of the target and filler posters, we replicated the study design of Koch and Zerback
(2011), which has been successfully employed to test the truth effect with political campaign posters. A cover
story was used to ensure that participants viewed the same campaign posters repeatedly without identifying
the purpose of the study or becoming bored. Participants were told that they would be presented with a series
3272 Nicole Ernst, Rinaldo Kühne, and Werner Wirth International Journal of Communication 11(2017)
of campaign posters in different urban settings (e.g., near a train station, in a pedestrian area) and that they
had to evaluate how well suited each setting was for placing a campaign poster.
Participants and Procedure
In total, 205 participants were randomly assigned to one of the six treatment groups. The
average age of the participants was 32 years, and 51% were women. The political positions of the
participants were equilibrated and spread within the left–right spectrum (M = 3.25, on a scale from 1 to
7). Random assignment produced group sizes ranging from 27 to 39.
The experiment was administered online in May and June 2012. Participants for the online survey
were recruited through a series of mailing lists. A link to the online survey was included in the e-mail. The
survey started with a brief introduction that presented the cover story. Participants were instructed to
evaluate the placement of a series of campaign posters in different urban settings. Subsequently,
campaign posters were presented one by one and participants had to evaluate the placement of each
poster separately. Once the participants had evaluated the placement of a poster, they could click on a
button to proceed to the evaluation of the next poster’s placement. Finally, after the whole series of
campaign posters had been presented, participants completed a questionnaire that assessed the two
dependent variables and several control variables.
Stimulus Materials
The campaign posters addressed the issue of gene technology. This issue was chosen for two
reasons. First, the issue was at the time not on the public agenda, which ensured that the participants of
the study did not have detailed knowledge or strong attitudes toward the issue. Second, there existed two
parties with contradictory positions regarding the issue, which enabled us to construct credible, negative
campaign posters on which one party attacked the other. The two parties were the Liberal Party and the
Green Party of Switzerland. The Liberal Party seeks to protect civil liberties and individual responsibility, is
committed to supporting research, and generally takes a positive position regarding gene technology. The
Green Party promotes environmentalism and green means of transportation and hence is strongly
opposed to gene technology. Thus, within the traditional positions of the two selected parties, it is
conceivable that the parties would campaign against each other on a ballot.
In a pilot study, we tested the perceived authenticity and negativity of 10 campaign posters.
More specifically, we tested which combination of sponsoring party and target party was perceived as
most authentic and which poster slogans produced the most pronounced difference in perceived
negativity. Five posters were sponsored by the Liberal Party. Three of these posters were presumably
negative and included an attack on the Green Party. Two posters were presumably neutral and did not
include any attack. Similarly, three negative posters and two neutral posters sponsored by the Green
Party were included. In addition, two filler posters, which were not sponsored by any party but just
announced the popular vote, were included in the pilot study. Thirty-three participants were instructed to
evaluate the authenticity and negativity of the posters. The goal was to first identify whether posters
sponsored by the Liberal or the Green Party were evaluated as more authentic. The results indicated that
International Journal of Communication 11(2017) Effects of Message Repetition and Negativity 3273
the posters that were sponsored by the Liberal Party were regarded as most authentic.1 Second, we
checked which pair of negative and nonnegative posters sponsored by the Liberal Party led to the largest
difference in negativity. The pair that produced the largest difference in negativity, MDifference = 4.85,
SDDifference = 1.54, t(14) = 12.19, p ≤ .01, was selected for the main study.
Both posters included a political slogan, an identical factual statement about the political issue, a
voting recommendation, and the date of the upcoming election. However, the two campaign posters
differed with regard to the negativity in the political slogan. The neutral poster included a slogan
promoting gene technology (“Promote progress—support research”) without attacking the political
opponent. In contrast, the negative poster included a slogan with an attack on the Green Party and
depicted a woman sitting in a wheelchair, keeping her hands in front of her face. The slogan beneath the
picture read, “Do you vote like the Greens? We choose human!—Gene technology saves lives.” Thus, the
negative campaign poster suggested that the Green Party did not care about disabled people because they
opposed gene technology and the potential remedies that may emerge from it.
The manipulation of message repetition consisted of presenting the respective campaign poster
(i.e., the neutral or negative poster) one, three, or six times. In addition to the target poster, participants
rated the position of five filler posters that advertised for the fictitious ballot but did not take any position.
The designed campaign posters were embedded in pictures of original poster placement
(billboards) in different urban settings in Switzerland to ensure a high ecological validity of the online
experiment. Participants were informed that they had to evaluate each placement of the campaign posters
in various urban settings to make the repeated poster presentation plausible. A question about the quality
of the poster placement was asked immediately after every poster presentation to ensure that the
participants were forced to draw attention to the poster and to prevent the participants from instantly
clicking to the next picture. The rather complicated political slogan of the Liberals (“In Switzerland, 90
genetically engineered drugs help against 153 known diseases”) was presented identically on the neutral
and attacking campaign posters. Because the experiment was online, the slogan formulation was complex
to avoid the risk of participants easily verifying or falsifying the statement by searching the Internet. To
avoid possible primacy and recency effects, we randomized the order of presentation of the target posters
and the filler posters. Stimuli examples of all poster types are depicted in the Appendix.
Measures
The survey measured the participants’ attitudes toward the presented posters, the credibility of
the factual statement, participants’ attitude toward gene technology, and a series of covariates. All
measures were based on 7-point rating scales (1 = strongly disagree, 7 = strongly agree). All items for all
assessments in the questionnaire were in German.
1 Overall, the selected posters for the main study of the Liberal Party reported higher mean values for
authenticity for the negative (M = 5.51, SD = 1.10) and neutral poster (M = 5.65, SD = 0.96) than the
highest scoring negative (M = 4.30, SD = 1.45) and neutral poster (M = 4.80, SD = 1.20) of the Green Party.
3274 Nicole Ernst, Rinaldo Kühne, and Werner Wirth International Journal of Communication 11(2017)
To check whether the manipulation of the posters’ negativity had succeeded, we measured
participants’ perceptions of negativity with four items (e.g., “The campaign poster tackles a political
party”). The items formed a reliable measure (α = .93), and a mean index for perceived negativity was
calculated (M = 3.55, SD = 1.33).
The first dependent variable, the credibility of the statement, was measured with an adapted
version of the Truth Effect Scale by Koch and Zerback (2011). The scale included three items (e.g., “I
think the statement ‘In Switzerland, 90 genetically engineered drugs help against 153 known diseases’ is
credible”). The three items formed a reliable measure (α = .92), and a mean index for credibility was
formed (M = 4.08, SD = 1.45).
The second dependent variable, participants’ attitude toward the political issue (i.e., gene
technology), was measured with three items (e.g., “There are good reasons to vote for the gene
technology initiative”). The three items formed a reliable measure (α = .72), and a mean index for a
positive attitude toward gene technology was formed (M = 4.30, SD = 1.12).
Furthermore, the questionnaire asked about political party identification, routine demographics,
and the seriousness of the respondents’ participation.
To corroborate that the indicators for the three variables (perceived negativity, credibility, and
attitude) represented three distinct constructs, we conducted a principal axis factor analysis (oblimin
rotation, eigenvalue criterion for factor extraction). The factor analysis identified the three expected
dimensions (70% explained variance). More important, each indicator exclusively loaded on the
appropriate dimension (all loadings ≥ .46) and not on any other dimension (all loadings ≤ .09) such that
no substantial cross-loadings existed. Thus, we can assume that the indicators formed distinct constructs.
Results
Treatment Check
To verify the manipulation of poster negativity, we conducted a one-way analysis of covariance
with poster type (neutral vs. negative) as the independent variable and perceived negativity as the
dependent variable. As controls, frequency, credibility, attitude attitudes toward gene technology, and
identification with the Green Party were included as covariates in the analysis. As expected, the neutral
poster was evaluated as significantly less attacking and critical (M = 2.58, SD = 1.45) than the negative
poster (M = 5.50, SD = 1.53), F(1, 199) = 189.64, p ≤ .001, η2 = .49. Thus, we can assume that the
manipulation worked as intended.
Testing of Hypotheses
To test the three hypotheses, we conducted indirect effect analyses using Hayes’s (2013) process
macro. Because our main independent variable message repetition was multicategorical, we followed the
approach outlined by Hayes and Preacher (2014) and recoded the experimental conditions into two
International Journal of Communication 11(2017) Effects of Message Repetition and Negativity 3275
dummy variables. The moderate-frequency condition (i.e., three repetitions) functioned as the reference
group. A (negatively coded) dummy for the low-frequency condition (i.e., one repetition) and a dummy for
the high-frequency condition (i.e., six repetitions) were entered as independent variables in the mediation
models. Accordingly, the two dummy variables assessed the effects of a change from low to moderate
frequency and a change from moderate to high frequency. The attitude toward gene technology was
included as the dependent variable and credibility judgment as the mediator. In both analyses, an
interaction between frequency and poster type (0 = neutral, 1 = negative) on credibility was added to the
mediation model. Identification with the Green Party and the high-frequency (Model 1) and low-frequency
dummy (Model 2) were included as covariates.2 Table 1 presents an overview of the descriptive statistics
and mean comparisons of credibility judgments and attitudes toward gene technology across the
frequency and poster type conditions.
Table 1. Overview of the Means, Standard Errors, and Group Differences.
Poster type
Frequency condition
Low Moderate High Overall
n M (SD) n M (SD) n M (SD) n M (SD)
Credibility judgment of the statement
Combined 69 3.72
(1.40)a
59 4.37
(1.44)b
73 4.22
(1.48)ab
201 4.10
(1.46)ab
Neutral 36 3.53
(1.54)a,y
27 4.22
(1.36)ab,y
39 4.55
(1.52)b,y
102 4.10
(1.54)ab,y
Negative 33 3.93
(1.21)a,y
32 4.50
(1.52)a,y
43 3.85
(1.36)a,z
99 4.09
(1.38)a,y
Attitude toward gene technology
Combined 69 4.10
(1.10)a
59 4.33
(1.13)a
73 4.49
(1.33)a
201 4.31
(1.20)a
Neutral 36 3.93
(1.14)a,y
27 4.40
(1.17)ab,y
39 4.75
(1.37)b,y
102 4.37
(1.28)ab,y
Negative 33 4.27
(1.04)a,y
32 4.28
(1.11)a,y
34 4.19
(1.24)a,y
99 4.25
(1.12)a,y
Note. All mean comparisons are based on Bonferroni post hoc tests. Means with different subscripts differ
significantly from each other (p < .05). Subscripts before the comma refer to comparisons across frequency
conditions (columns). Subscripts after the comma refer to comparisons across poster types (rows).
The first hypothesis predicted that increasing the frequency of exposure to a campaign poster
from low to moderate would positively influence the perceived credibility of the factual statement
presented on the poster. The results of the first model including the low-frequency condition as the
2 This specification was necessary because process permits only one independent variable per model
(Hayes, 2013).
3276 Nicole Ernst, Rinaldo Kühne, and Werner Wirth International Journal of Communication 11(2017)
independent variable showed that the credibility of the statement significantly increased (b = 1.09, SE =
0.33, p < .01) when the poster was presented three times (M = 4.37) instead of only one time (M =
3.72). Moreover, there was no significant interaction effect of the poster type and the low-frequency
condition dummy on the credibility judgment (b = −0.82, SE = 0.42, ns). Overall, these results
corroborate that the political statements were rated as more credible after a moderate frequency of three
repetitions compared with the low-frequency condition of only one repetition. H1 was therefore supported
(see Figure 1).
Indirect effect of neutral poster type: b = 0.35, p < .05.
Indirect effect of negative poster type: b = 0.09, ns.
Indirect effect of neutral poster type: b = 0.11, ns.
Indirect effect of negative poster type: b = −0.22, p < .05.
Figure 1. Indirect effects of low and high frequency on attitude. In each model, the indirect
effect through the other frequency dummy (low vs. high) and the identification with
the Green Party were controlled. **p < .01. ***p < .001.
The second hypothesis proposed that the effect of increasing poster exposure from a moderate to
a high frequency on the credibility of the poster’s statement would be moderated by message negativity.
The hypothesis argued that credibility would decrease in the high-frequency condition when negative
campaign poster was presented (H2a), whereas high exposure to a neutral or nonnegative campaign
poster would further increase credibility (H2b). We found that increasing the overall exposure from a
moderate- (M = 4.37) to a high-frequency condition (M = 4.22) had no direct effect on credibility (b =
0.35, SE = 0.32, ns). In line with the hypothesis, we can however report a significant interaction between
the poster type and the high-frequency dummy (b = −1.03, SE = 0.41, p < .01). The negative interaction
International Journal of Communication 11(2017) Effects of Message Repetition and Negativity 3277
suggests that the effect of increasing the frequency of exposure from three to six times was more
negative when the campaign poster was negative. To gain a better understanding of the interaction and
conclusively evaluate the hypothesis, we inspected the conditional effect of exposure for the negative and
the neutral posters. We found that the increase in repetition had a negative effect on message credibility for
the negative poster (b = −0.68, SE = 0.32, p < .05). Credibility ratings were lower for the high-frequency
(M = 3.85) than the moderate-frequency condition (M = 4.50). In comparison, the increase from the
moderate- (M = 4.22) to the high-frequency condition (M = 4.55) had no significant effect on credibility for
the neutral poster (b = 0.35, SE = 0.32, ns). Thus, H2 was partly corroborated: Being frequently exposed to
a negative campaign poster did indeed reduce the statements’ credibility, which supported H2a. However,
H2b was rejected: A high frequency of exposure to a neutral poster did not increase credibility as compared
with a moderate frequency. Credibility ratings of the neutral poster’s message were about the same and not
significantly different in the moderate- and the high-frequency conditions.
The third hypothesis predicted that the more credible a political statement about a political issue
is perceived, the more positive is the attitude toward the political issue. As expected, perceiving the
statement about gene technology as credible had a positive effect on the attitude toward the issue of gene
technology (b = 0.32, SE = 0.05, p < .001). Thus, H3 was corroborated.
The identified relationships between exposure, credibility, and attitudes, and the interaction
between negativity and the high-exposure dummy indicated that (conditional) indirect effects existed. To
quantify these effects, we inspected the results of the indirect effects analyses (see Hayes, 2013).
Significance tests of the indirect effects were based on bias-corrected bootstrap confidence intervals
(5,000 bootstrap samples). First, we found that presenting the campaign poster three times instead of one
time had a positive indirect effect through credibility on the attitude toward gene technology (b = 0.26),
95% CI [0.01, 0.58]. Second, we found that increasing the frequency of exposure from moderate to high
had a negative indirect effect on the attitude toward gene technology when the poster message was
negative (b = −0.22), 95% CI [−0.49, −0.02], but not when the message was neutral (b = 0.11), 95%
CI [−0.07, 0.35]. Thus, exposure frequency and poster negativity had a joint effect on the perceived
credibility of the political, factual statement, which, in turn, affected the political attitude.
Discussion
The aim of this study was to investigate the effects of statement repetition and message
negativity in the context of campaign communication. The results demonstrated that especially in
combination with high repetition, negativity is a crucial moderator to explain a decrease of credibility
judgments and attitude toward a political issue.
The study conducted two moderated mediation analyses. We found that the frequency of the
poster presentation had an indirect effect on the attitude toward gene technology through its effect on
message credibility. The mediation models revealed that message negativity functioned as a moderator of
these mediation mechanisms: A high number of repetitions had only an indirect negative effect on attitude
through message credibility when the poster was negative.
3278 Nicole Ernst, Rinaldo Kühne, and Werner Wirth International Journal of Communication 11(2017)
To summarize, we found evidence for the main assumption that message negativity is a crucial
moderator of the effects of message repetition on credibility. This study revealed that negative messages
can have a weaker persuasive impact. The finding is in line with current research, which has shown that
negative messages can lead to a backlash effect (Ansolabehere & Iyengar, 1995; Fridkin & Kenney, 2011;
Lau & Rovner, 2009).
The study has three theoretical implications for the truth effect hypothesis. First, the study
corroborated that a moderate message repetition positively influences credibility ratings. However, we
also found that the truth effect is limited and that message credibility decreases after too many
repetitions. Second, the study suggests that negativity is an important message characteristic, which
influences the optimal number of repetitions. This finding implies that we can further improve our
understanding of the effects of message repetition by investigating the moderating influence of additional
message characteristics as well as characteristics of the recipient. For instance, dual process theories
(e.g., Chaiken, 1980) indicate that the recipient’s involvement may affect how intensively messages are
processed, which, in turn, should influence which number of repetitions maximizes message credibility.
Third, the study is in line with previous research on the truth effect by showing that repetitions within a
short timeframe affect the persuasive impact of a stimulus. However, the finding also raises the question
about whether similar effects would arise when the time interval between repetitions is increased.
Fernandes (2013), for example, showed that candidate evaluations and likelihood of voting improved only
if ads were repeated over a larger time period. Hence, the time interval between repetitions may affect
the relationship between the number of repetitions, message credibility, and attitudes and should be
considered as an additional crucial moderator.
Our results have practical implications for communication specialists and political campaign
managers. Our study indicates that simply relying on repetition as a persuasive tool might not always be
crowned with success. Message characteristics—notably, negativity—have to be considered in the planning
of a campaign because they influence optimal exposure frequencies. Indeed, the optimal number of
repetitions may be different in real-life campaigns because of the lower attention of the audience and
interference from additional factors. Nevertheless, message negativity may still influence the effects of
repeated exposure in actual campaigns, and communication specialists should thus consider this factor
when piloting their messages. In particular, communication specialists should take into consideration that
negative or attacking campaign ads may backfire, especially if voters are confronted with identical
negative ads several times within a short time interval.
There are some limitations to this study that must be considered. One limitation involves the use
of an experiment. First, the study ignored the contextual factor of real campaigns because the poster and
the ballot were fictitious, and experimental design in general can never capture all the dynamics of
political campaigns. For example, the participant relied only on the information they received within
minutes, without having any further details about the political issue and positions of the involved parties.
Moreover, the study ignored that the external validity was restricted because participants were aware of
the experimental situation. Although the real object of the study was not revealed and a cover story was
used, participants probably paid more attention to the presented campaign poster than they would in real-
life situations, which may have elicited higher negative feelings. However, it should be noted that the
International Journal of Communication 11(2017) Effects of Message Repetition and Negativity 3279
experimental design, in return, has the advantage that it ensured the internal validity of the findings and
allowed us to establish causal relationships. Furthermore, the external validity of the stimuli was enhanced
by presenting the posters in naturalistic settings using original pictures. This presentation mode
presumably decreased the amount of attention that the participants directed toward the content of the
posters too, which should also support external validity because individuals are unlikely to pay full
attention to campaign posters on the street.
A second limitation refers to the negativity type the study used. Fridkin and Kenney (2011)
showed that the impact of negative information is multifaceted, and not all citizens are equally influenced
by negative campaign ads. Our study investigated only the effects of one party directly attacking the
position of another party. As Dardis, Shen, and Edwards (2008) demonstrated, the effects of issue-based
attacks are not identical with character-based attack ads, and as Chou and Lien (2013) examined, the
process furthermore depends on the appeal type (rational vs. emotional appeal). Next to the perception of
negativity, the design of the negative poster type could have possibly elicited other negative values or
emotions. Future research should explore whether other types of negativity have the same impact by
differentiating between direct (personal) attacks in contrast to negative arguments about a political issue
and varying the degree of emotional appeal. Moreover, the influence of different negative poster designs
on negative emotions and values and their effects on political attitudes should be investigated.
Finally, the study did not directly consider the effects of psychological reactance, which also
functions as a boundary condition for the truth effect (Koch & Zerback, 2013). Psychological reactance is a
social psychological concept that explains human behavior in response to an actual or potential loss of
freedom in the environment (J. Brehm, 1966). Individuals are assumed to be free to select a position on an
issue, but this freedom can be threatened by various pressures to adopt or change their position (S. Brehm
& Brehm, 1981). Accordingly, psychological reactance is an aversive motivational state directed toward
restoring behavioral freedoms that are threatened in order to assert autonomy. Koch and Zerback (2013)
showed that a high frequency of repetition can trigger these feelings of reactance, because recipients
interpret the high exposure as a persuasive tactic by the communicator. If the recipient identifies this intent
of persuasion, it is automatically perceived as a threat to the recipient’s freedom (J. Brehm, 1966), which will
result in psychological reactance (e.g., Moyer-Gusé, Jain, & Chung, 2012). Moreover, Meirick and Nisbett
(2011) demonstrated that negative political ads provoke higher levels of reactance compared with positive
ones. The authors demonstrated that, when confronted with persuasive political ads, negative ads promote a
stronger source derogation (one possible reactance modality that results in lower opinions about the
communicator), which results in a higher level of felt reactance when compared with positive campaign ads.
We therefore expect that message negativity would elicit higher feelings of reactance, which would result in
the decrease of the statement’s credibility. Future research should address the direct relation of
psychological reactance and message negativity and examine whether psychological reactance mediates the
effect of negative message repetition on message credibility.
To conclude, this study adds to the current research on the effects of message repetition on
credibility judgments and attitudes. Previous studies on the truth effect have not distinguished between
different message valences. This study contributes to the field by simultaneously investigating the effects
of negative and nonnegative messages. Furthermore, it demonstrates that the repeated presentation of
3280 Nicole Ernst, Rinaldo Kühne, and Werner Wirth International Journal of Communication 11(2017)
negative messages influences not only candidate evaluations and the likelihood of voting but also message
credibility and attitudes toward political issues. We also demonstrated that negativity on campaign posters
can function as a crucial moderator of repetition effects on credibility judgments. Similarly, we showed
that the valence of a stimulus can explain the inverted-U curve and the decrease of credibility ratings after
a high number of repetitions. Finally, we showed that repeatedly presented negative posters resulted in a
more negative attitude toward the presented political issue, which was mediated by a decrease in
credibility judgments.
References
Allen, M., & Burrell, N. (2002). The negativity effect in political advertising: A meta-analysis. In J. P.
Dillard & M. Pfau (Eds.), The persuasion handbook: Developments in theory and practice (pp. 83–
96). Thousand Oaks, CA: SAGE Publications.
Ansolabehere, S., & Iyengar, S. (1995). Going negative: How attack ads shrink and polarize the
electorate. New York, NY: Free Press.
Arkes, H., Boehm, L., & Xu, G. (1991). Determinants of judged validity. Journal of Experimental Social
Psychology, 27(6), 576–605.
Arkes, H., Hackett, C., & Boehm, L. (1989). The generality of the relation between familiarity and judged
validity. Journal of Behavioral Decision Making, 2(2), 81–94.
Bartels, L. (2006). Priming and persuasion in presidential campaigns. In H. E. Brady & R. Johnston (Eds.),
Capturing campaign effects (pp. 78–112). Ann Arbor, MI: University of Michigan Press.
Becker, L., & Doolittle, J. C. (1975). How repetition affects evaluations of and information seeking about
candidates. Journalism & Mass Communication Quarterly, 52(4), 611–617.
Begg, I., Anas, A., & Farinacci, S. (1992). Dissociation of processes in belief: Source recollection,
statement familiarity, and the illusion of truth. Journal of Experimental Psychology: General,
121(4), 446–458.
Bornstein, R. (1989). Exposure and affect: Overview and meta-analysis of research 1968–1987.
Psychological Bulletin, 106, 265–289.
Brehm, J. (1966). A theory of psychological reactance. New York, NY: Academic Press.
Brehm, S., & Brehm, J. (1981). Psychological reactance: A theory of freedom and control. New York, NY:
Academic Press.
International Journal of Communication 11(2017) Effects of Message Repetition and Negativity 3281
Brettschneider, F. (2008). Negative campaigning. In W. Donsbach (Ed.), The international encyclopedia of
communication (pp. 3186–3188). Malden, MA: Blackwell.
Brown, A., & Nix, L. (1996). Turning lies into truths: Referential validation of falsehoods. Journal of
Experimental Psychology: Learning, Memory, and Cognition, 22(5), 1088–1100.
Chaiken, S. (1980). Heuristic versus systematic information processing and the use of source versus
message cues in persuasion. Journal of Personality and Social Psychology, 39(5), 752–766.
Choo, T.-H. (1964). Communicator credibility and communication discrepancy as determinants of opinion
change. The Journal of Social Psychology, 64(1), 65–76.
Chou, H., & Lien, N. (2013). The effects of appeal types and candidates’ poll rankings in negative political
advertising. Asian Journal of Communication, 23(5), 489–518.
Dardis, F., Shen, F., & Edwards, H. (2008). Effects of negative political advertising on individuals’ cynicism
and self-efficacy: The impact of ad type and message exposures. Mass Communication and
Society, 11(1), 24–42.
Dechêne, A., Stahl, C., Hansen, J., & Wanke, M. (2010). The truth about the truth: A meta-analytic review
of the truth effect. Personality and Social Psychology Review, 14(2), 238–257.
Esser, F., Engesser, S., Matthes, J., & Berganza, R. (2016). Negativity. In C. de Vreese, F. Esser, & D.
Hopmann (Eds.), Comparing political journalism (pp. 71–91). London, UK: Routledge.
Fernandes, J. (2013). Effects of negative political advertising and message repetition on candidate
evaluation. Mass Communication and Society, 16(2), 268–291.
Fridkin, K., & Kenney, P. (2011). Variability in citizens’ reactions to different types of negative campaigns.
American Journal of Political Science, 55(2), 307–325.
Galtung, J., & Ruge, M. (1965). The structure of foreign news: The presentation of the Congo, Cuba and
Cyprus crises in four Norwegian newspapers. Journal of Peace Research, 2(1), 64–90.
Geer, J. (2006). In defense of negativity: Attack ads in presidential campaigns. Chicago, IL: University of
Chicago Press.
Gigerenzer, G. (1984). External validity of laboratory experiments: The frequency-validity relationship.
The American Journal of Psychology, 97(2), 185‒195.
Hasher, L., Goldstein, D., & Toppino, T. (1977). Frequency and the conference of referential validity.
Journal of Verbal Learning and Verbal Behavior, 16(1), 107–112.
3282 Nicole Ernst, Rinaldo Kühne, and Werner Wirth International Journal of Communication 11(2017)
Hawkins, S., & Hoch, S. (1992). Low-involvement learning: Memory without evaluation. Journal of
Consumer Research, 19(2), 212–225.
Hawkins, S., Hoch, S., & Meyers-Levy, J. (2001). Low-involvement learning: Repetition and coherence in
familiarity and belief. Journal of Consumer Psychology, 11(1), 1–11.
Hayes, A. (2013). Mediation, moderation, and conditional process analysis: A regression-based approach.
New York, NY: Guilford Press.
Hayes, A., & Preacher, K. (2014). Statistical mediation analysis with a multicategorical independent
variable. British Journal of Mathematical and Statistical Psychology, 67(3), 451–470.
Hovland, C., & Weiss, W. (1951). The influence of source credibility on communication effectiveness.
Public Opinion Quarterly, 15(4), 635–650.
Johar, G., & Roggeveen, A. (2007). Changing false beliefs from repeated advertising: The role of claim-
refutation alignment. Journal of Consumer Psychology, 17(2), 118–127.
Kahn, K., & Kenney, P. (2000). How negative campaigning enhances knowledge of senate elections. In J.
A. Thurber, C. J. Nelson, & D. A. Dulio (Eds.), Crowded airwaves: Campaign advertising in
elections (pp. 65–95). Washington, DC: Brookings Institution Press.
Koch, T., & Zerback, T. (2011). Wiederholung der Wahrheit: Wie die mehrmalige Präsentation politischer
Slogans die Einschätzung ihrer Glaubwürdigkeit beeinflusst [Repetition of the truth: How the
multiple presentation of political slogans influences the assessment of their credibility]. Medien-
und Kommunikationswissenschaft, 59(4), 487–504.
Koch, T., & Zerback, T. (2013). Helpful or harmful? How frequent repetition affects perceived statement
credibility. Journal of Communication, 63(6), 993–1010.
Lang, A., Newhagen, J., & Reeves, B. (1996). Negative video as structure: Emotion, attention, capacity,
and memory. Journal of Broadcasting & Electronic Media, 40(4), 460–477.
Lau, R., & Pomper, G. (2002). Effectiveness of negative campaigning in U.S. Senate elections. American
Journal of Political Science, 46(1), 47–66.
Lau, R., & Rovner, I. (2009). Negative campaigning. Annual Review of Political Science, 12, 285–306.
Lau, R., Sigelman, L., & Rovner, I. (2007). The effects of negative political campaigns: A meta-analytic
reassessment. Journal of Politics, 69(4), 1176–1209.
Martin, P. (2004). Inside the black box of negative campaign effects: Three reasons why negative
campaigns mobilize. Political Psychology, 25(4), 545–562.
International Journal of Communication 11(2017) Effects of Message Repetition and Negativity 3283
Meirick, P., & Nisbett, G. (2011). I approve this message: Effects of sponsorship, ad tone, and reactance
in 2008 presidential advertising. Mass Communication and Society, 14(5), 666–689.
Metzger, M., Flanagin, A., Eyal, K., Lemus, D., & Mccann, R. (2003). Credibility for the 21st century:
Integrating perspectives on source, message, and media credibility in the contemporary media
environment. Annals of the International Communication Association, 27(1), 293–335.
Miller, R. (1976). Mere exposure, psychological reactance and attitude change. Public Opinion Quarterly,
40(2), 229–233.
Moyer-Gusé, E., Jain, P., & Chung, A. (2012). Reinforcement or reactance? Examining the effect of an
explicit persuasive appeal following an entertainment–education narrative. Journal of
Communication, 62(6), 1010–1027.
Pentony, J. (1998). Effects of negative campaigning on vote, semantic differential, and thought listing.
Journal of Applied Social Psychology, 28(23), 2131–2149.
Petty, R., & Cacioppo, J. (1986). The elaboration likelihood model of persuasion. In L. Berkowitz (Ed.),
Advances in experimental social psychology (pp. 123–205). San Diego, CA: Academic Press.
Powell, F. (1965). Source credibility and behavioral compliance as determinants of attitude change.
Journal of Personality and Social Psychology, 2(5), 669–676.
Reber, R., & Schwarz, N. (1999). Effects of perceptual fluency on judgments of truth. Consciousness and
Cognition, 8(3), 338–342.
Reinhard, M., Schindler, S., Raabe, V., Stahlberg, D., & Messner, M. (2014). Less is sometimes more:
How repetition of an antismoking advertisement affects attitudes toward smoking and source
credibility. Social Influence, 9(2), 116–132.
Schwartz, M. (1982). Repetition and rated truth value of statements. The American Journal of Psychology,
95(3), 393–407.
Seidman, S. (2008). Posters, propaganda, and persuasion in election campaigns around the world and
through history. New York, NY: Peter Lang.
Smith, N., Larsen, J., Chartrand, T., Cacioppo, J., Katafiasz, H., & Moran, K. (2006). Being bad isn’t
always good: Affective context moderates the attention bias toward negative information. Journal
of Personality and Social Psychology, 90(2), 210–220.
Soroka, S., & McAdams, S. (2015). News, politics, and negativity. Political Communication, 32(1), 1–22.
3284 Nicole Ernst, Rinaldo Kühne, and Werner Wirth International Journal of Communication 11(2017)
Unkelbach, C. (2007). Reversing the truth effect: Learning the interpretation of processing fluency in
judgments of truth. Journal of Experimental Psychology: Learning, Memory, and Cognition, 33(1),
219–230.
Zajonc, R. (1968). Attitudinal effects of mere exposure. Journal of Personality and Social Psychology, 9(2),
1–27.
Appendix
Figure A1. Stimuli example of negative campaign poster in an urban setting.
Figure A2. Stimuli example of neutral campaign poster in an urban setting.
International Journal of Communication 11(2017) Effects of Message Repetition and Negativity 3285
Figure A3. Stimuli example of filler campaign poster in an urban setting.