International Journal of Communication 10(2016), 2576–2595 1932–8036/20160005
Copyright © 2016 (Franziska Marquart, Jörg Matthes, & Elisabeth Rapp). Licensed under the Creative
Commons Attribution Non-commercial No Derivatives (by-nc-nd). Available at http://ijoc.org.
Selective Exposure in the Context of Political Advertising:
A Behavioral Approach Using Eye-Tracking Methodology
FRANZISKA MARQUART
JÖRG MATTHES
ELISABETH RAPP
University of Vienna, Austria
Selective exposure refers to the tendency of individuals to attend to information that is
in line with their political views. This study advocates a behavioral approach to selective
exposure research by introducing eye-tracking as a straightforward measure of selection
processes. We tested participants’ selective exposure to political poster advertisements
from one left-wing and one right-wing party. Individuals were simultaneously exposed to
on-screen ads by the two different parties, and their eye movements were unobtrusively
recorded. Findings indicate that the political orientation of the participants explained
selective exposure in terms of the time taken to look at each ad. Implications for
selective exposure research are discussed.
Keywords: selective exposure, political advertising, eye-tracking, political predispositions
The scholarly debate on selective exposure in the field of communication research has prospered
considerably in the past years (e.g., Garrett, 2013; Knobloch-Westerwick, 2015; Sunstein, 2001). As part
of this discussion, it has been argued that if individuals only attend to information that is in line with their
own views, this may result in the proliferation of echo chambers (Jamieson & Cappella, 2008), in which
one’s opinion is amplified by media and personal networks consistent with their own position and
conflicting points of view are less likely to be encountered. Such echo chambers could possibly threaten
the diversity of opinions, with damaging consequences for deliberative democracies, but research mostly
provides evidence for the fact that the majority of citizens still rely on a balanced diet of different news
sources and information outlets (Chaffee, Saphir, Graf, Sandvig, & Hahn, 2001; Garrett, 2013).
Scholars have found accumulating evidence for the occurrence of selective exposure (e.g.,
Garrett, 2009, 2013; Knobloch-Westerwick, 2015), but research has been concerned mostly with the
active selection of information, for example, by asking individuals retrospectively about their media use or
recording their content choice. In contrast, little is known about selection at a comparably early stage of
exposure. That is, how do people assign their (visual) attention according to their own (political) self and
the information they are presented with? If we understand attention as a part of selective exposure and a
Franziska Marquart: franziska.marquart@univie.ac.at
Jörg Matthes: joerg.matthes@univie.ac.at
Elisabeth Rapp: rapp.elisabeth@gmail.com
Date submitted: 2015-07-22
International Journal of Communication 10(2016) Selective Exposure in Political Advertising 2577
precondition for any type of processing or effects (i.e., understanding, memory, learning, or persuasion)
to occur, this issue clearly demands further investigation.
In this study, we asked whether individuals actively expose themselves to political messages that
are in line with their own ideological predispositions. Specifically, we tested participants’ selective
exposure to poster advertisements from two opposing parties. By recording participants’ eye movements,
we specified the actual amount of time people award to different political ads, and tested our proposed
effects with regard to posters by a left-wing (Green Party) and a right-wing (Austrian Freedom Party)
party. For reasons outlined below, we believe the implementation of eye-tracking in selective exposure
research to be an important methodological contribution, as it may help in determining whether selective
exposure takes place at a very early level of content perception and selection. By testing this assumption
in the context of poster advertisements, we further aimed to take a closer look at media content that
differs from information normally tested in selective exposure research: We argue that messages that are
looked at in passing demand less time and processing, but are hard to avoid. As a consequence, eyetracking
is especially well suited to address processes of selective exposure. To our knowledge, this is the
first study to address the effects of political predispositions on selective exposure to political advertising
with the help of eye-tracking methodology. Our findings are discussed with regard to the broader
implications for selective exposure research.
Selective Exposure Theory
During campaigns, voters are confronted with a large array of political advertisements, both from
candidates they sympathize with as well as those they dislike. In a surrounding where the amount of
incoming information outnumbers individuals’ cognitive capacity to process every message, people
selectively expose themselves to some content while dismissing other. The attendance to and/or
avoidance of (political) messages have been termed selective exposure. This process, in line with
Festinger’s (1957) theory of cognitive dissonance, is theorized to reinforce preexisting attitudes and
beliefs. Most authors find support for the hypothesis that individuals pay greater attention to consistent
information, but research has on some occasions been inconclusive (e.g., Donsbach, 2009; KnoblochWesterwick
& Meng, 2009, 2011; Matthes, 2012; Sears & Freedman, 1967).
In an attempt to embrace the diversity of research in the area, Knobloch-Westerwick (2015)
proposed a framework addressing the different phases of selective exposure. The selective exposure selfand
affect management (SESAM; Knobloch-Westerwick, 2015) model postulates that individuals’ selection
of media content is determined by the relevant self-concept that is activated when they are exposed to it:
Depending on the working self accessible at the time of exposure, people attend differently to media.
Importantly, self-concepts can be automatically activated by media stimuli and subsequently guide
message exposure and attention: “Media messages represent social contexts that render particular selfconcepts
relevant, which in turn influence what messages are selected and how these affect media users”
(Knobloch-Westerwick, 2015, p. 967). The model asserts that attention to a given stimulus leads to the
recognition of content elements and determines which working self will be activated. This, in turn, affects
exposure to (or avoidance of) and further processing of the message. On the one hand, the SESAM model
conceptualizes selective exposure as a process in which media users turn to messages because they help
2578 F. Marquart, J. Matthes, & E. Rapp International Journal of Communication 10(2016)
them maintain said self-concepts. On the other hand, merely encountering media content or new
information may also activate relevant self-concepts (Knobloch-Westerwick, 2015). The dynamictransactional
character of this process (Früh & Schönbach, 1982) points to the interrelation of an
individual and her surrounding, including the media message. It also emphasizes how the different stages
of attention, selection, and message processing are intertwined: After a phase of—potential
preconscious—attention, people identify the broad subject of the stimulus and anticipate possible
outcomes of message exposure. They determine which self-concept is relevant, which further guides the
selective exposure process, including prospective media impacts, interpretation, and response to the
message (Knobloch-Westerwick, 2015). In a similar manner, McGuire (2013) characterized the first steps
of a persuasion process as exposure to the content, followed by attention, liking/maintaining interest in
the message, and comprehension. The discussed frameworks highlight the dynamic nature and connection
of attention and processing stages in the interaction with media messages.
In line with these arguments, we believe exposure to political advertising, similar to other forms
of media content, to be guided by (a) individuals’ prior attitudes and predispositions as well as (b)
characteristics of the message itself (i.e., the advertising party; Taber & Lodge, 2006). Exposure to
attitude-consistent messages may increase the accessibility of partisanship in people, thus reinforcing
their political self-concept (e.g., Knobloch-Westerwick & Meng, 2011). Along these lines, findings by
Matthes and Marquart (2015) show that exposure to like-minded political advertising could accelerate the
timing of citizens’ voting decisions and increase political participation. Accordingly, the nature of a media
stimulus in terms of its closeness to the individual may determine message exposure and (subsequent)
effects. Although political content in general as well as campaign communication in particular have been
studied in the context of selective exposure before (e.g., Garrett, 2009; Iyengar, Hahn, Krosnick, &
Walker, 2008), no research so far has addressed the perception of political poster advertisements in this
regard.
Political Poster Advertisements
Given the relevance of posters for political advertising in the European context (e.g., Kaid, 2012;
Seidman, 2008), it is important to determine whether citizens expose themselves to these ads. During
election campaigns, parties compete for voters’ attention, and streets are virtually plastered with posters.
Often, these ads are installed next to each other, and exposure to multiple posters by various parties can
hardly be avoided (Matthes & Marquart, 2015; Plasser & Plasser, 2002). The specific characteristics of
poster ads that are mostly attended to in passing and without investing much cognitive energy make them
important media messages for studying the processes of selective exposure. Political posters have been
found to successfully communicate a party’s or candidate’s issue position (Seidman, 2008), most notably
because of their high visibility in the public sphere (Kaid, 2012). Given their comparably low complexity,
the limited amount of space adjudged to textual information, and their reliance on visual content, political
posters aim at gaining voters’ attention for a very brief amount of time (Dumitrescu, 2010). This short
attention span is used to convey the relevant information (candidate, party, or issue positions) as fast as
possible and before—much as in conventional, nonpolitical advertising—individuals feel the need to avoid
the persuasive content. Due to posters’ relative unostentatious nature, voters report feeling less annoyed
International Journal of Communication 10(2016) Selective Exposure in Political Advertising 2579
by them compared with other forms of political advertisements, and the ads still have the potential to
affect attitudes and vote choice (e.g., Johnston & Pattie, 1998; Seidman, 2008; van den Bulck, 1993).
Within the scope of this study and following the SESAM model (Knobloch-Westerwick, 2015), we
conceptualize selective exposure to political posters as a continuous process that involves attention to the
content, the activation of the relevant (here: political) self-concept by means of contact with the ad, and
the subsequent guiding of exposure and information processing. However, as outlined above, the
threshold of being exposed to such ads (be they like-minded or not) is low compared with forms of
advertising that require more active engagement, such as TV spots or candidates’ websites: Especially
during national campaigns, the majority of voters notice poster ads on the streets (Schmitt-Beck &
Wolsing, 2010). For that reason, asking people whether they have seen them potentially yields
inconclusive results. Instead, we propose to employ eye-tracking technology to test how much voters
actually expose themselves to such ads. Again, selective exposure in the context of political posters is a
dynamic process: When encountering the ads in a given social context (e.g., while driving past them on
the street), individuals’ political self-concepts are activated, and will subsequently guide the (length and
intensity of) exposure to the posters. Whereas studies in selective exposure in the past mostly focused on
questioning individuals about their media use, researchers have increasingly turned their attention to
behavioral measurements.
Measuring Selective Exposure
Clay, Barber, and Shook (2013), in a recent review on measuring selective exposure (see also
Feldman, Stroud, Bimber, & Wojcieszak, 2013; Ryan & Brader, 2015), identified four main categories of
measurement techniques. First, survey participants are questioned retrospectively about their behavior,
for instance, with regard to news consumption habits (e.g., Garrett, Carnahan, & Lynch, 2013; Matthes &
Marquart, 2015). For example, Chaffee and colleagues (2001) administered a cross-sectional survey
measuring attention (i.e., recall) to media content about political candidates and testing selective
exposure by means of political orientation and preference for specific candidates during an election.
Although such recall measures mostly pertain to ongoing events, and therefore are highly relevant in
terms of external validity, interpretation of their findings is not always conclusive (Bennett & Iyengar,
2008). Specifically, problems of selective attention and retention apply, which is to say that individuals
have a biased memory when it comes to remembering messages that are in line or conflict with their own
attitudes. In addition, ideological consistency may be a desirable goal in self-presentation, thus leading to
a distortion in self-report measurements (Clay et al., 2013).
Second to addressing behavior retrospectively, questioning individuals about their intentional
behavior with regard to selective exposure is a common technique: Participants are given a choice of, for
instance, news articles from which to select, and asked about their interest in reading those articles (for a
discussion, see also Feldman et al., 2013). The problems arising from such measurements are similar to
the arguments outlined above. In addition, measures of intentional behavior mostly leave individuals to
choose one message or another (i.e., as a dichotomous variable), limiting the transferability of findings.
Moreover, researchers gain no information about how recipients attend to this material: When reading a
2580 F. Marquart, J. Matthes, & E. Rapp International Journal of Communication 10(2016)
newspaper, one may simply scan counterattitudinal information, but closely read attitude-consistent
paragraphs, or vice versa.
Third, studies recently have begun using behavioral measures to determine the actual selective
exposure people engage in (see Garrett & Stroud, 2014; Jang, 2014; Knobloch-Westerwick, 2015; Prior,
2013; Vraga, Bode, & Troller-Renfree, 2016). An important advantage of such measurements is that they
take into account the differences in actual exposure time allocated to the information. Such behavioral
measures include the (unobtrusive) tracking of the time spent on reading news articles or the recording of
interactive behavior in selecting pieces of information from websites (Jang, 2014; Knobloch-Westerwick &
Meng, 2009; Valentino, Banks, Hutchings, & Davis, 2009) or CD-ROMs (Iyengar et al., 2008). Behavioral
measurements, however, often rely on fake content (e.g., mock websites), with questions arising
pertaining to the generalizability of findings. In addition, it can be argued that the time people spend on
websites does not necessarily relate to the amount of reading one does when viewing such a website
(e.g., Knobloch-Westerwick & Meng, 2009). Fourth, Clay and colleagues (2013) discuss the collection of
aggregate data in large population samples (e.g., market data and population level estimates). Naturally,
individual explanations for media selection become secondary to general explanatory factors in such larger
studies, making it increasingly difficult to clarify motivations for attending to specific media.
Against the background of the discussed measurements techniques, we deem the recording of
participants’ actual behavior the most appropriate technique when researching selective exposure, as it
allows for the identification of actual choices without having to rely on self-reports (Knobloch-Westerwick,
2015). In the present study, we operationalized attitude-consistent and -inconsistent messages as political
advertisements promoted by a party that individuals either support (i.e., like and sympathize with) or
dislike, and measured selective exposure to these ads by recording participants’ eye movements when
presented with two poster ads simultaneously.
Eye Movement as an Indicator for Selective Exposure to Political Posters
Eye-tracking is still not very established in communication research (Sandberg, Gidlöf, &
Holmberg, 2011; but see Vraga et al., 2016). This may come as a surprise given that any “visual exposure
(potential and actual) to a communication effort is a necessary but not a sufficient condition for a
cognitive, attitudinal or behavioral effect to occur” (Sandberg et al., 2011, p. 24). In their study of
children’s exposure to online advertising, Sandberg and colleagues (2011) measured participants’ fixations
to determine the amount of time they adjudged different ads (see also Jansen, Nederkoorn, & Mulkens,
2005). In the realm of consumer research, scholars have identified a “positive feedback loop” (Gidlöf,
Wallin, Dewhurst, & Holmqvist, 2013, p. 2) between what individuals look at and what they choose, which
is to say that the more time participants award an item, the more likely they are to prefer it over
alternative offers in a subsequent situation (see also Janiszewski, Kuo, & Tavassoli, 2013; Orquin &
Mueller Loose, 2013). Similar findings stem from studies in the field of marketing and consumer research
(e.g., Bebko, Sciulli, & Bhagat, 2014; Pieters & Warlop, 1999; Schotter, Gerety, & Rayner, 2012), showing
that exposure (measured as the amount of time dedicated to a stimulus via eye-tracking) relates to
important outcome variables such as attitudes or behavior. Investigating the visual exposure to
(nonpolitical) poster ads in photographs displaying street sceneries, Maughan, Gutnikov, and Stevens
International Journal of Communication 10(2016) Selective Exposure in Political Advertising 2581
(2007) also established a link between the amount of fixations an ad received and its following evaluation
in terms of liking. However, caution is warranted with regard to the causal interpretation of findings in this
context (i.e., liking determining exposure or vice versa). In the field of political communication, hardly any
comparable studies have been conducted so far, which may be because the effort in collecting and
analyzing eye movements from large samples is enormous, as people need to be exposed to the stimuli in
individual sessions in the laboratory.
Yet, information about the actual amount of time dedicated to different content is of vital
importance, as it frees scholars from the burden of having to rely on self-report measures. Eye
movements can be considered indicators of preferences at a very early stage of media reception (e.g.,
Higgins, Leinenger, & Rayner, 2014), thus preceding any possible effects. Although exposure time has
been used as a dependent variable in determining selective exposure before (e.g., Knobloch-Westerwick &
Meng, 2009), to our knowledge, no studies have employed the analysis of eye movements in this area.
Yet, we believe that the documentation of perceptual processes in the context of political advertising
would provide a valuable addition to the selective exposure literature. Using eye movements as indicators
is a very direct and unbiased way of observing selective exposure. More specifically, eye-tracking observes
single fixations that usually last 200 to 400 milliseconds, which is especially important when addressing
media content that does not require much time and/or cognitive energy for processing. On the level of
such short exposure, measuring study participants’ time spent on a stimulus simply by recording, for
example, how long the content was displayed on screen might yield inconclusive results. In contrast to
that, eye-tracking can be considered an unbiased and straightforward measure of exposure.
Taken together, we identify two important shortcomings in the present literature: First, no
investigations have addressed selective exposure to political poster advertising in a setting that is similar
to a regular environment, that is, posters of several parties competing for potential voters’ attention on
the street. Second, self-report measures in selective exposure research may be susceptible to
measurement errors, for which reason we suggest using a behavioral eye-tracking measure to determine
the actual amount of time devoted to political ads from different parties. Both points are addressed in the
research presented here.
Hypotheses
As this study was set in a Western European context, political advertising by numerous parties
was relevant to our context. However, we decided to focus on two distinct parties (and their corresponding
ads) that are positioned on the opposite ends of the political spectrum: the Austrian Green Party
(designated left wing) and the Austrian Freedom Party (designated right wing). Our experimental design
resembled the placement of political ads in the streets, with several posters placed next to each other. As
people walk by, they quickly look at those advertisements for a couple of seconds. During this time frame,
they can hardly process the full information displayed on the posters, as it is rather unlikely that people
stop walking and start reading until every ad’s content is fully understood (although this does happen,
occasionally; see Lessinger, Moke, & Holtz-Bacha, 2003). By contrast, posters are often processed in
passing without people investing much cognitive energy.
2582 F. Marquart, J. Matthes, & E. Rapp International Journal of Communication 10(2016)
Our design allowed us to test selective exposure at a very basic level—the level of fixations in eye
movements. Using an experimental paradigm and based on selective exposure research, we assumed that
people who are left-leaning in their political ideology and who identify/sympathize with the Austrian Green
Party would attend selectively to its poster ads. Reversely, people high in left-wing political ideology would
expose themselves less to political posters by the right-wing Austrian Freedom Party. This led to our two
hypotheses:
H1: Left-wing political predisposition positively affects selective exposure to left-wing political
advertisements.
H2: Left-wing political predisposition negatively affects selective exposure to right-wing political
advertisements.
Method
We employed a within-subject experimental study encompassing a short survey as well as the
recording of eye movements while viewing political posters by the Austrian Green Party (left wing) and
Freedom Party (right wing). Participants for the study (N = 57; 56% female; Mage = 31.33 years, SD =
10.21, range = 19–66 years; 49% without a university degree) were asked to attend individual data
recording sessions in a laboratory at a large Austrian university; they did not receive any incentives for
participating. They were recruited via personal networks and invited on the basis of political
predispositions to balance the sample in terms of left- and right-wing political ideology. Because
recruitment of right-leaning participants was comparably difficult, potential individuals were additionally
addressed via party headquarters, student fraternities, and related community groups.
Stimuli and Procedure
The study used political posters from the Green Party and Freedom Party during the 2013
Austrian national election campaign (Election day September 29). Data collection lasted a week at the end
of November 2013. The stimuli were similar in terms of format, major slogan, depiction of the party leader
(Eva Glawischnig and Heinz-Christian Strache, respectively), and the party’s logo (see Figures 1 and 2).
Up to one week before the study took place, participants answered a short questionnaire
assessing their political ideology. A few subjects answered the survey at the laboratory shortly before
starting the session, as they had not done so in advance. After arriving at the lab, participants were
seated in front of the stationary eye tracker, where the investigator explained the procedure. Data
collection started after an initial calibration phase; participants were advised to look at the ads on the
monitor on the basis of their own interests: “Imagine you’re walking down a street and pass by these ads
on a billboard. . . .” No task was given to them for the course of the stimuli presentation. Individuals were
aware that their eye movements were recorded, but they were advised to act as they would normally in
the described situation. After an example ad for the purpose of making participants familiar with the
device, the target ads were shown on the screen. Participants were told that, at any time during
International Journal of Communication 10(2016) Selective Exposure in Political Advertising 2583
recording, they could revoke their consent for participation, for example, if they felt uncomfortable with
the procedure. However, this was not necessary.
Figure 1. The six poster ads by the Austrian Green Party used as stimuli in the experimental
design (from top to bottom, left to right: Posters 1–6). Copyright: Austrian Green Party.
Participants saw a total of 12 different target posters (six left wing, six right wing; see Figures 1
and 2) taken from the national election campaign in 2013. Two ads—one of each party—simultaneously
appeared on the screen, and participants were presented with six different dyads (within-subject factor) in
random order, with Poster 1 by the Green Party always shown next to Poster 1 by the Freedom Party and
so on. The two ads were set up on the left and right side of the screen and separated by a narrow bar of
blank space; their left or right position was randomized and later statistically controlled for. In addition,
the upper and lower portions of the screen were empty. Individual poster coverage thus varied between
25% and 26%, and an average of 48% of the screen remained empty (i.e., blank space).
Exposure time for each dyad was limited to 10 seconds. In their study on commercial poster ads,
Maughan and colleagues (2007) displayed photographs of street scenes in which one poster ad was
included for 5 seconds each. The average time awarded to political posters in passing is believed to range
between 6 and 8 seconds for pedestrians and as few as 4 seconds for drivers (Lessinger et al., 2003). Our
rather long time frame of 10 seconds was chosen because two ads were shown simultaneously, and
subjects should be given enough time to read all slogans without being interrupted in processing.
2584 F. Marquart, J. Matthes, & E. Rapp International Journal of Communication 10(2016)
Figure 2. The six poster ads by the Austrian Freedom Party used as stimuli in the experimental
design (from top to bottom, left to right: Posters 1–6). Copyright: Austrian Freedom Party.
Poster presentation (left/right position on the screen) was inverted in two experimental groups to
control for effects with regard to the ads’ arrangement, and was later included as a control. In between
the dyads, participants were shown a blank screen with a centered fixation cross for 3 seconds so as to
not confound the gaze direction with fixation of the last advertisement. Presentation order of the relevant
stimuli was randomized within groups, and distractor ads were randomly inserted in between the parties’
posters. The distractors (i.e., poster ads for products such as chocolate, bottled water, or detergent, a
charity organization, and service providers) served to veil the purpose of the study, and their eye-tracking
data were not analyzed. After the recording of eye movements, participants were thanked, debriefed, and
dismissed. For data analysis, survey and eye-tracking data were matched.
Measures
In addition to age and gender, participants were required to state their formal education (50.9%
university degree or equivalent). Identification with the relevant political parties was assessed by asking
individuals how strongly (1 = not at all, 6 = very strongly) they identified with the Green Party (M = 3.54,
SD = 1.63) and Freedom Party (M = 1.68, SD = 1.13). Because all of the ads shown during the
experiment pictured the respective party leader as well, we also asked participants for their liking of the
politicians (1 = do not like her/him at all, 6 = like her/him a lot); individuals clearly favored the Green
Party’s chairwoman (M = 3.36, SD = 1.41) over the Freedom Party’s leader (M = 1.77, SD = 1.17). For
the subsequent analyses, two steps were taken in computing the final predictor: First, party identification
and politicians’ evaluation were summarized into two indices for each party (mean average of party
identification and candidate liking), resulting in overall left-wing orientation (M = 3.44, SD = 1.43, α =
.85) and overall right-wing orientation (M = 1.73, SD = 1.09, α = .89). The two indices were moderately
International Journal of Communication 10(2016) Selective Exposure in Political Advertising 2585
negatively correlated (r = –.35, p = .01), but a test for multicollinearity was negative (tolerance = 1.0,
variance inflation factor = 1.0). However, to increase the variance for the predictor, we calculated a
difference score: We subtracted overall right-wing orientation from overall left-wing orientation (left –
right), with higher and positive values (maximum +6.5) indicating a more left-leaning as opposed to rightleaning
(maximum –6.5) political orientation. The resulting score of general political ideology (M = 1.67,
SD = 2.08, range = –4.0 to +4.5) shows that our sample was slightly more left-leaning in general. Yet, in
contrast to the singular predictors for both ideological orientations, the mean indicates enough variance to
estimate the effects of right- and left-wing political orientation. Participants were also asked for their
agreement on a number of issue-specific statements. Because these items are not relevant for the scope
of this article, their results are not addressed in the subsequent analysis.
The eye-tracking data were obtained with a stationary SMI iView X RED eye tracker, with each
full poster defined as one area of interest. Eye positions were sampled at 120 Hz. As our dependent
variable, participants’ exposure to the political poster advertisements was operationalized as the total
amount of time devoted to each individual ad in milliseconds (i.e., the fixation duration/time for each
party’s ads in the dyads; see Tables 1 and 2). Fixation time is an established and direct measure for
exposure in eye-tracking research (e.g., Jansen et al., 2005), and participants in our study looked at ads
by the Green Party for an average of 4.364 seconds (SD = 0.926), and spent an average 3.734 seconds
(SD = 0.977) fixating the Freedom Party ads. Fixation time for the Green Party and Freedom Party ads
was negatively correlated, ranging from r(57) = –.53, p < .01, to r(57) = –.69, p = .01, for the six dyads.
However, of 10 seconds viewing time in total, subjects in our study also spent an average of M = 1.91
seconds fixating neither poster, which is to say that the 48% of blank space on the screen accounted for
roughly 20% of fixations. Given that individuals actually did expose themselves to the neutral space,
analysis for the ads is executed separately, as fixations did not necessarily center on either political ad.
Results
In our first hypothesis, we assumed that overall left-wing political orientation would positively
affect fixation time for ads by the Green Party. To control for differences in the designs of the posters that
could have an impact on eye movements, we estimated a repeated-measure analysis of covariance.
Fixation time of the six Green Party posters was included as the within-subject factor, and ad positioning
(left side–right sight), age, gender, education, and political orientation were entered as covariates.
Regarding the within-subject factor, we did not find significant effects for the different levels, indicating
that the six designs of the Green Party posters did not affect fixation time in different ways, F(5, 240) =
0.57, p = .72, partial η2
= .01 (see Table 1). In contrast, and supporting our first hypothesis, general
political ideology was a significant and strong predictor for fixation time on the Green Party ads, F(1, 48)
= 9.35, p = .004, partial η2
= .16. Importantly, we did not find an interaction effect of the within-subject
factor and political identification, F(1, 48) = 1.32, p = .257, partial η2
= .03, which is to say that the
impact of political orientation did not vary for the individual posters. None of the other covariates exerted
a relevant impact (see Table 1).
2586 F. Marquart, J. Matthes, & E. Rapp International Journal of Communication 10(2016)
Table 1. Repeated-Measure One-Way Analysis of Covariance of
Fixation Time (milliseconds) for Ads by the Green Party.
Variable df F Partial η2
p
Green Party ads (RM) 5 0.57 .01 .72
Ad positioning 1 0.79 .02 .38
Gender 1 0.25 .01 .62
Age 1 0.25 .01 .62
Education 1 0.03 .00 .86
Political orientation 1 9.35 .16 .00
Note. RM = repeated-measure factor (party posters).
Hypothesis 2 stated that overall right-wing political orientation would positively influence
participants’ fixation time of political ads by the Freedom Party (given that our relevant predictor ranged
from right to left wing with increasing values). Again, we did not see an effect of the six different levels of
the within-subject factor, F(5, 240) = 0.34, p = .89, partial η2
= .01 (see Table 2), whereas overall
political orientation had a significant impact on fixation time, F(1, 48) = 5.25, p = .03, partial η2
= .10. No
interaction effect of posters and political orientation could be found, F(1, 48) = 0.338, p = .563, partial η2
= .01. Age positively affected exposure to right-wing poster ads, albeit not on a significant level, F(1, 48)
= 3.94, p = .053, partial η2
= .08. Neither gender nor education or ad positioning affected viewing time
(see Table 2). Thus, H2 was supported as well.1
1
Following a reviewer’s suggestion, we also computed the same repeated-measure ANCOVAs including
not one general political orientation score, but the two separate variables for left- and right-wing ideology,
which allows us to draw (cautious) conclusions about the possible selective avoidance of opinionincongruent
advertisement content as well. For posters by the Green Party, in line with the reported
results, a positive evaluation of the party and her leader strongly and positively affects fixation time (F(1,
47) = 10.99 p = .002, partial η2
= .19), but supporters of the Freedom Party do not selectively avoid the
green content (F(1, 47) = .08, p = .78, partial η2
= .002). Subjects in favor of the Freedom Party spend
more time looking at their ‘own’ ads (F(1, 47) = .1.32, p = .26, partial η2
= .03), but individuals
supporting the Green Party do not avoid fixation the Freedom Party’s ads F(1, 47) = 2.53 p = .118, partial
η2
= .05). However, note that the effects for the Freedom Party’s posters are not significant, most likely
due to the small sample size of right-wing-oriented individuals in our sample and accompanying small
variance. We therefore refrain from making explicit claims about actual selective avoidance processes
here.
International Journal of Communication 10(2016) Selective Exposure in Political Advertising 2587
Table 2. Repeated-Measure One-Way Analysis of Covariance of
Fixation Time (milliseconds) for Ads by the Freedom Party.
Variable df F Partial η2
p
Freedom Party ads (RM) 5 0.34 .01 .89
Ad positioning 1 1.41 .03 .24
Gender 1 0.96 .02 .33
Age 1 3.94 .08 .05
Education 1 0.00 .00 .95
Political orientation 1 5.25 .10 .03
Note. RM = repeated-measure factor (party posters).
The overall effects as tested in the repeated-measure analyses of covariance show that political
orientation was the most relevant predictor of fixation time for both left- and right-wing political poster
ads, but this procedure in analysis did not allow us to interpret the magnitude of effects in terms of
changes in milliseconds. Therefore, we also estimated linear regression models, analyzing fixation time for
the six different dyads as dependent on the aforementioned variables, with general political orientation
remaining the relevant predictor (and higher values indicating a more left-leaning identification; data not
shown). For the Green Party, the impact of political orientation on exposure to the poster ads ranged from
b = 82.725 (SE = 104.593, ns, Dyad 1) to b = 327.193 (SE = 91.136, p ≤ .001, Dyad 6; unstandardized
coefficients represent milliseconds). For the Freedom Party, political orientation (right wing) affected
fixation time for the party’s ads in a range from b = –74.865 (SE = 102.932, ns, Dyad 5) to b = –235.493
(SE = 103.675, p < .05, Dyad 2). The negative sign for right-wing ads corresponds to the predictor’s
opposite poling, as lower values represent higher right-wing orientation. Across all six dyads, political
orientation (left wing) positively and significantly affected selective exposure to ads by the Green Party (b
= 201.331, SE = 65.856, p < .01), and, reversely, negatively affected selective exposure to ads by the
right-wing Freedom Party (b = –157.475, SE = 68.753, p < .05).
Discussion
This study has documented effects of political predispositions on individuals’ selective exposure to
political poster advertisements. Using eye-tracking methodology to measure actual viewing behavior
rather than (potentially biased) self-reports, we found participants’ left-wing orientation to significantly
predict exposure (i.e., fixation time) to political ads by the Green Party. In a reverse relationship,
individuals’ left-wing orientation negatively affected the amount of time spent looking at ads by the
Freedom Party. Neither age, nor education, nor gender (significantly) predicted fixation duration on any of
the posters. Whether a poster was displayed on the left or right side of the screen was also unrelated to
fixation duration.
Our measurement of overall political orientation allows us to conclude that an individual’s
evaluation of the advertising party as well as its respective chairman or chairwoman influences selective
exposure at the level of eye movements: People spend more time looking at political ads that are in line
with their own political view. This study is the first to detect such selective exposure effects at a
2588 F. Marquart, J. Matthes, & E. Rapp International Journal of Communication 10(2016)
comparably early level (i.e., with the regard to eye movements). Moreover, the difference in exposure
time we detected here (i.e., 3.2 seconds in between the two extremes on the ideological scale) indicates
that political orientation plays a large role in the process of selective exposure, accounting for about half
of the time individuals usually dedicate to posters.
However, recent research has argued that people do not necessarily avoid other, inconsistent
messages (Garrett, 2009) while attending to like-minded information. That is to say that although
selective exposure does take place (and encompasses an important strategy to address the modern
world’s huge amount of information; e.g., Bennett & Iyengar, 2008), the occurrence of simultaneous
selective avoidance should, at the very least, be questioned. Indeed, a study by Garrett and colleagues
(2013; see also Garrett, 2009; Garrett & Stroud, 2014; Jang, 2014; Knobloch-Westerwick & Meng, 2009)
showed that American voters who visited like-minded political websites did not by all means avoid crosscutting
online information. Rather, both behaviors were positively correlated in terms of frequency.
Accordingly, selective exposure and selective avoidance should be treated as distinct phenomena (Garrett,
2009; Jang, 2014). We were not able to directly test selective avoidance processes in this study, and we
are cautious in making any inferences in this regard. However, future research is strongly recommended
to distinguish between exposure and avoidance effects. To do this, a more diverse sample in terms of
political orientation would be necessary: As outlined above, we experienced difficulties in the recruiting of
(admitted) right-leaning individuals, although our sample offered a large enough range to distinguish leftfrom
right-leaning participants by use of the difference score. We argue, however, that open support for
the Austrian Freedom Party’s positions (e.g., in terms of migration) still may be susceptible to a socialdesirability
effect, which made it difficult to recruit individuals for this study. For this reason, we strongly
recommend that future studies on the selective exposure (and, possibly, avoidance) of political content
use measurements of implicit attitudes in addition to explicitly asking individuals for their liking of the
parties and leaders as well because they may add predictive value in the context of political behaviors
such as voting, especially for undecided voters (e.g., Arcuri, Castelli, Galdi, Zogmaister, & Amadori,
2008). Although we were not able to implement such measurements here, they could prove useful when
studying evaluations of political entities as a determinant of selective exposure to said content. To
accurately address selective avoidance, it also would be necessary to present subjects with political stimuli
next to unrelated (i.e., nonpolitical) ads to determine whether exposure and avoidance occur
simultaneously.
In this study, we operationalized selective exposure to a political content by measuring
participants’ fixation times with regard to the poster ads, arguing that the time spent looking at poster ads
is closely intertwined with content processing and precedes any possible effects. Yet, it should be noted
that—in line with the processes encompassing selective exposure—eye movements are also at least in
part guided by comparably pre-attentive and less conscious perceptions: During a very short period of
time, people are able to gain a first impression as well as a general overview of a stimulus presented to
them, without (yet) actively attending to the content. It is possible that individuals in our study switched
back and forth between the two ads in each dyad simply to scan the content, identify or compare the
respective positions, and make themselves familiar with the ads. However, research suggests that as little
as 100 milliseconds are enough for individuals to get a general impression of a novel stimulus (Higgins et
al., 2014). Moreover, work by Bar, Neta, and Linz (2006) showed that individuals are able to make
International Journal of Communication 10(2016) Selective Exposure in Political Advertising 2589
judgments about peoples’ (threatening) facial expressions after only a brief exposure time of less than 40
milliseconds. Along similar lines, Ballew and Todorov (2007) proved that exposure to politicians’ faces for
100 milliseconds is sufficient for individuals to form a judgment that would later predict election outcomes.
Importantly, individuals need to at least briefly identify the context of a stimulus for a relevant selfconcept
to be activated (Knobloch-Westerwick, 2015). To account for such an orientation phase, we reran
our analysis for the regressions, but deducted the first 100 milliseconds of exposure time in each dyad. No
significant changes were detected with regard to our predicting variables: Left-wing political ideology still
positively and significantly (b = 199.126, p < .01) affected overall fixation time on the Green Party ads,
and negatively influenced exposure to the Freedom Party ads (b = –107.697, p = .078). These results
give strong support to our claim that, after allowing participants to gain a first short overview, they were
able to distinguish the presented parties and, accordingly, engage in active selective exposure behavior
based on the activation of their political self-concept. Given the long exposure time of 10 seconds, we
deem it unlikely that all of this time was devoted to gaining an overview of the presented positions only.
Because exposure time was held constant across dyads and conditions, effects with regard to any preattentive
orientation phase should also be invariable for all participants and posters. However, in our point
of view, the aforementioned distinction between (selective) attention and exposure deserves closer
scholarly attention. Future studies should therefore aim to measure study participants’ depth of processing
of a given message dependent on viewing time to better understand which phase of the selective
exposure process is adjudged to orientation or initial attention before message processing begins.
Limitations and Future Research
As has been stated before, future investigations should aim to recruit a larger share of right-wing
supporters. Alternatively, this research should be extended to parties from the moderate political
spectrum. To test for processes of avoidance as well, future studies should not only contrast political
information by two opposing parties but should also include neutral content. Such a design would allow for
investigators to distinguish between the active exposure to like-minded and avoidance of incongruent
information at the same time. Furthermore, the findings reported here need to be replicated for other
countries with different party systems and in the context of other types of political advertisements. We
used classic political poster ads with a party head and a simple slogan; however, political ads can differ in
many aspects, such as negativity, emotional content, use of visuals, and so on. We aimed to present
equal posters from two different but opposing parties that were similar in their visual design. As shown in
the repeated-measure analysis, differences between the posters displayed in each dyad did not affect
participants’ viewing behavior. Yet, we encourage future studies to further examine the possibility that
content-related differences in the ads affect fixation time, for example, with regard to the party leaders’
genders, the size of slogans, or coloring. Of course, it also would be worthwhile to extend this research to
other media, such as websites or newspaper advertisements: We argued that the specific characteristics
of political poster advertisements provide an excellent opportunity to study selective exposure at the level
of eye movements, as individuals only briefly encounter them in passing and hardly avoid them. However,
these features can also be found in other content, not least commercial advertising or public
communication campaigns (e.g., Pease, Brannon, & Pilling, 2006). Whether or not people actively expose
themselves to such content as well, depending on the relevant self-concept that gets activated, warrants
further research.
2590 F. Marquart, J. Matthes, & E. Rapp International Journal of Communication 10(2016)
We operationalized eye movements as an indicator of selective exposure as the outcome variable
in our design, which is to say that we did not aim to identify possible effects of fixation duration.
Nevertheless, the outcomes of selective exposure processes may be versatile and deserve closer scholarly
attention, and we strongly encourage future studies to address such potential effects in connection with
eye-movement data and political ads. Within other areas of research, fixation duration has been found to
significantly affect individuals’ evaluation of a presented target stimulus as well as subsequent choices and
behaviors (e.g., Schotter et al., 2012), and similar links should be investigated for the field of political
communication. Moreover, additional variables might influence selective exposure to political content, both
mediating and moderating the effect of political orientation on the time spent looking at such ads: For
example, although we controlled for subjects’ age, gender, and educational level (and did, indeed, not find
any significant effects with regard to these factors), research indicates that conservative and liberal voters
differ in their cognitive styles and brain activities (e.g., Garrett & Stroud, 2014; Schreiber et al., 2013). It
was beyond the scope of this study to further address such variables, yet they may provide important
explanatory power in the context under question.
This being said, another important limitation warrants attention: Unarguably, presenting political
poster ads on a computer screen does not fully compare with a real-world situation on the street. We tried
to partly account for this problem by limiting exposure time and simultaneously presenting two ads at a
time to simulate an actual election campaign. Eye-tracking data could be obtained from mobile devices as
well (and participants thus be tested, e.g., while walking down a street), but this was beyond our
capacity. Yet, in our point of view, collecting behavioral field data in the context of real campaigns opens a
promising field for future studies.
Finally, we strongly encourage researchers to replicate the findings reported here, be it with
similar or advanced designs, different types of media, or in other countries. Although we argue that
selective exposure to like-minded (or incongruent) content can be successfully measured with the help of
eye-tracking, our results are limited in their ability to be generalized to other contexts, and this claim
should thus be further tested, refined, and extended. Such extensions also should aim at validating the
measurement: In relying on research from other fields, we claim eye-tracking to be a valid method for
studying selective exposure. However, we did not directly ask participants whether they (consciously)
sought out the opposing parties’ poster ads; neither did we employ additional scales or other, established
measures that directly connect to selective exposure as addressed in communication science. More work is
definitely required here.
Conclusion
Besides our substantive findings on selective exposure in the context of political advertising, we
have made a methodological contribution to the literature by introducing eye-tracking methodology as a
straightforward measure of selective exposure processes. We believe that eye-tracking may overcome the
perceptual biases of self-reports frequently employed in this literature, and that—depending on the
stimulus in question—it may capture selective exposure more accurately than other forms of behavioral
measures. This is especially relevant when it comes to media content that does not require much time or
resources, and that is therefore hard to address with conventional behavioral measures that have been
International Journal of Communication 10(2016) Selective Exposure in Political Advertising 2591
applied in selective exposure research before. We do not wish to suggest that the recording of eye
movements may replace such measurements, but we do believe that it may complement our
methodological array in this area of research. Eye-tracking can be applied for studying selective exposure
for all kinds of media, such as political poster ads, newspaper articles, online banners, TV spots, or social
media posts. Given the comparable costly and complex data collection process, it may not be suitable for
all areas of research and for any underlying questions. Yet, selective exposure is not a static yes/no
decision, but is a complex process that unfolds over time. People read, stop reading, allocate their
attention to something else, and may return again. Ignoring this complexity may lead to wrong or
misguided conclusions about processes that may affect a variety of outcomes in terms of attitudes and
behavior. We therefore believe that selective exposure scholars should carefully consider the expediency
of eye-tracking measurement. We hope that our study provides a preliminary effort on this very promising
new avenue of selective exposure research.
References
Arcuri, L., Castelli, L., Galdi, S., Zogmaister, C., & Amadori, A. (2008). Predicting the vote: Implicit
attitudes as predictors of the future behavior of decided and undecided voters. Political
Psychology, 29(3), 369–387. doi:10.1111/j.1467-9221.2008.00635.x
Ballew, C. C., & Todorov, A. (2007). Predicting political elections from rapid and unreflective face
judgments. Proceedings of the National Academy of Sciences, 104(46), 17948–17953.
doi:10.1073/pnas.0705435104
Bar, M., Neta, M., & Linz, H. (2006). Very first impressions. Emotion, 6(2), 269–278. doi:10.1037/1528-
3542.6.2.269
Bebko, C., Sciulli, L. M., & Bhagat, P. (2014). Using eye tracking to assess the impact of advertising
appeals on donor behavior. Journal of Nonprofit & Public Sector Marketing, 26(4), 354–371.
doi:10.1080/10495142.2014.965073
Bennett, W. L., & Iyengar, S. (2008). A new era of minimal effects? The changing foundations of political
communication. Journal of Communication, 58(4), 707–731. doi:10.1111/j.1460-
2466.2008.00410.x
Chaffee, S. H., Saphir, M. S., Graf, J., Sandvig, C., & Hahn, K. S. (2001). Attention to counter-attitudinal
messages in a state election campaign. Political Communication, 18(3), 247–272.
doi:10.1080/10584600152400338
Clay, R., Barber, J. M., & Shook, N. J. (2013). Techniques for measuring selective exposure: A critical
review. Communication Methods and Measures, 7(3–4), 147–171.
doi:10.1080/19312458.2013.813925
2592 F. Marquart, J. Matthes, & E. Rapp International Journal of Communication 10(2016)
Donsbach, W. (2009). Cognitive dissonance theory—Roller coaster career: How communication research
adapted the theory of cognitive dissonance. In T. Hartmann (Ed.), Media choice: A theoretical and
empirical overview (pp. 128–149). London, UK: Routledge.
Dumitrescu, D. (2010). Know me, love me, fear me: The anatomy of candidate poster designs in the 2007
French legislative elections. Political Communication, 27(1), 20–43.
doi:10.1080/10584600903297117
Feldman, L., Stroud, N. J., Bimber, B., & Wojcieszak, M. (2013). Assessing selective exposure in
experiments: The implications of different methodological choices. Communication Methods and
Measures, 7(3–4), 172–194. doi:10.1080/19312458.2013.813923
Festinger, L. (1957). A theory of cognitive dissonance. Stanford, CA: Stanford University Press.
Früh, W., & Schönbach, K. (1982). Der dynamisch-transaktionale Ansatz. Ein neues Paradigma der
Medienwirkungen [The dynamic-transactional approach: A new paradigm of media effects].
Publizistik, 27(1–2), 74–88.
Garrett, R. K. (2009). Politically motivated reinforcement seeking: Reframing the selective exposure
debate. Journal of Communication, 59, 676–699. doi:10.1111/j.1460-2466.2009.01452.x
Garrett, R. K. (2013). Selective exposure: New methods and new directions. Communication Methods and
Measures, 7(3–4), 247–256. doi:10.1080/19312458.2013.835796
Garrett, R. K., Carnahan, D., & Lynch, E. (2013). A turn toward avoidance? Selective exposure to online
political information, 2004–2008. Political Behavior, 35(1), 113–134. doi:10.1007/s11109-011-
9185-6
Garrett, R. K., & Stroud, N. J. (2014). Partisan paths to exposure diversity: Differences in pro- and
counterattitudinal news consumption. Journal of Communication, 64(4), 680–701.
doi:10.1111/jcom.12105
Gidlöf, K., Wallin, A., Dewhurst, R., & Holmqvist, K. (2013). Using eye tracking to trace a cognitive
process: Gaze behavior during decision making in a natural environment. Journal of Eye
Movement Research, 6(1), 1–14.
Higgins, E., Leinenger, M., & Rayner, K. (2014). Eye movements when viewing advertisements. Frontiers
in Psychology, 5, 1–15. doi:10.3389/fpsyg.2014.00210
Iyengar, S., Hahn, K. S., Krosnick, J. A., & Walker, J. (2008). Selective exposure to campaign
communication: The role of anticipated agreement and issue public membership. The Journal of
Politics, 70(1), 186–200. doi:10.1017/S0022381607080139
International Journal of Communication 10(2016) Selective Exposure in Political Advertising 2593
Jamieson, K. H., & Cappella, J. N. (2008). Echo chamber: Rush Limbaugh and the conservative media
establishment. Oxford, UK: Oxford University Press.
Janiszewski, C., Kuo, A., & Tavassoli, N. T. (2013). The influence of selective attention and inattention to
products on subsequent choice. Journal of Consumer Research, 39(6), 1258–1274.
doi:10.1086/668234
Jang, S. M. (2014). Challenges to selective exposure: Selective seeking and avoidance in a multitasking
media environment. Mass Communication and Society, 17(5), 665–688.
doi:10.1080/15205436.2013.835425
Jansen, A., Nederkoorn, C., & Mulkens, S. (2005). Selective visual attention for ugly and beautiful body
parts in eating disorders. Behaviour Research and Therapy, 43(2), 183–196.
doi:10.1016/j.brat.2004.01.003
Johnston, R. J., & Pattie, C. J. (1998). Campaigning and advertising: An evaluation of the components of
constituency activism at recent British general elections. British Journal of Political Science,
28(4), 677–685. doi:10.1017/S0007123498000301
Kaid, L. L. (2012). Political advertising as political marketing: A retro-forward perspective. Journal of
Political Marketing, 11(1–2), 29–53. doi:10.1080/15377857.2012.642731
Knobloch-Westerwick, S. (2015). The selective exposure self- and affect-management (SESAM) model:
Applications in the realms of race, politics, and health. Communication Research, 42(7), 959–
985. doi:10.1177/0093650214539173
Knobloch-Westerwick, S., & Meng, J. (2009). Looking the other way: Selective exposure to attitudeconsistent
and counterattitudinal political information. Communication Research, 36(3), 426–448.
doi:10.1177/0093650209333030
Knobloch-Westerwick, S., & Meng, J. (2011). Reinforcement of the political self through selective exposure
to political messages. Journal of Communication, 61(2), 349–368. doi:10.1111/j.1460-
2466.2011.01543.x
Lessinger, E.-M., Moke, M., & Holtz-Bacha, C. (2003). “Edmund, Essen ist fertig!” Plakatwahlkampf 2002—
Motive und Strategien [“Edmund, supper is on!” Poster campaigning 2002—Motives and
strategies]. In C. Holtz-Bacha (Ed.), Die Massenmedien im Wahlkampf. Die Bundestagswahl 2002
(pp. 216–242). Wiesbaden, Germany: Westdeutscher Verlag.
Matthes, J. (2012). Exposure to counterattitudinal news and the timing of voting decisions.
Communication Research, 39(2), 147–169. doi:10.1177/0093650211402322
2594 F. Marquart, J. Matthes, & E. Rapp International Journal of Communication 10(2016)
Matthes, J., & Marquart, F. (2015). A new look at campaign advertising and political engagement:
Exploring the effects of opinion-congruent and -incongruent political advertisements.
Communication Research, 42(1), 134–155. doi:10.1177/0093650213514600
Maughan, L., Gutnikov, S., & Stevens, R. (2007). Like more, look more. Look more, like more: The
evidence from eye-tracking. Journal of Brand Management, 14(4), 335–342.
doi:10.1057/palgrave.bm.2550074
McGuire, W. J. (2013). McGuire’s classic input-output framework for constructing persuasive messages. In
R. E. Rice & C. K. Atkin (Eds.), Public communication campaigns (pp. 133–145). Los Angeles, CA:
SAGE Publications.
Orquin, J. L., & Mueller Loose, S. (2013). Attention and choice: A review on eye movements in decision
making. Acta Psychologica, 144(1), 190–206. doi:10.1016/j.actpsy.2013.06.003
Pease, M. E., Brannon, L. A., & Pilling, V. K. (2006). Increasing selective exposure to health messages by
targeting person versus behavior schemas. Health Communication, 19(3), 231–240.
doi:10.1207/s15327027hc1903_5
Pieters, R., & Warlop, L. (1999). Visual attention during brand choice: The impact of time pressure and
task motivation. International Journal of Research in Marketing, 16(1), 1–16.
doi:10.1016/S0167-8116(98)00022-6
Plasser, F., & Plasser, G. (2002). Global political campaigning: A worldwide analysis of campaign
professionals and their practices. Westport, CT: Praeger.
Prior, M. (2013). Media and political polarization. Annual Review of Political Science, 16(1), 101–127.
doi:10.1146/annurev-polisci-100711-135242
Ryan, T. J., & Brader, T. (2015). Gaffe appeal: A field experiment on partisan selective exposure to
election messages. Political Science Research and Methods. Advance online publication.
doi:10.1017/psrm.2015.62
Sandberg, H., Gidlöf, K., & Holmberg, N. (2011). Children’s exposure to and perceptions of online
advertising. International Journal of Communication, 5, 21–50. doi:1932–8036/20110021
Schmitt-Beck, R., & Wolsing, A. (2010). Der Wähler begegnet den Parteien [The voter meets the parties].
In K.-R. Korte (Ed.), Die Bundestagswahl 2009 (pp. 48–68). Wiesbaden, Germany: VS.
Schotter, E. R., Gerety, C., & Rayner, K. (2012). Heuristics and criterion setting during selective encoding
in visual decision-making: Evidence from eye movements. Visual Cognition, 20(9), 1110–1129.
doi:10.1080/13506285.2012.735719
International Journal of Communication 10(2016) Selective Exposure in Political Advertising 2595
Schreiber, D., Fonzo, G., Simmons, A. N., Dawes, C. T., Flagan, T., Fowler, J. H., & Paulus, M. P. (2013).
Red brain, blue brain: Evaluative processes differ in Democrats and Republicans. PLoS One, 8(2),
e52970. doi:10.1371/journal.pone.0052970
Sears, D. O., & Freedman, J. L. (1967). Selective exposure to information: A critical review. Public Opinion
Quarterly, 31(2), 194–213. doi:10.1086/267513
Seidman, S. A. (2008). Posters, propaganda, and persuasion in election campaigns around the world and
through history. New York, NY: Peter Lang.
Sunstein, C. R. (2001). Republic.com. Princeton, NJ: Princeton University Press.
Taber, C. S., & Lodge, M. (2006). Motivated skepticism in the evaluation of political beliefs. American
Journal of Political Science, 50(3), 755–769. doi:10.1111/j.1540-5907.2006.00214.x
Valentino, N. A., Banks, A. J., Hutchings, V. L., & Davis, A. K. (2009). Selective exposure in the Internet
age: The interaction between anxiety and information utility. Political Psychology, 30(4), 591–
613. doi:10.1111/j.1467-9221.2009.00716.x
van den Bulck, J. (1993). Estimating the success of political communication strategies: The case of
political poster impact in a Belgian election. European Journal of Communication, 8(4), 471–489.
doi:10.1177/0267323193008004004
Vraga, E., Bode, L., & Troller-Renfree, S. (2016). Beyond self-reports: Using eye tracking to measure topic
and style differences in attention to social media content. Communication Methods and Measures,
10(2–3), 149–164. doi:10.1080/19312458.2016.1150443