ArticlePDF Available

The Citizen Communication Mediation Model Across Countries: A Multilevel Mediation Model of News Use and Discussion on Political Participation

Authors:

Abstract and Figures

Since introduced by Professor McLeod and the Wisconsin School at the turn of the century, a large body of research has employed the communication mediation model. Yet, most of these studies rely on cross-sectional and individual-level survey data collected in the United States. This paper seeks to address these shortcomings by testing a specification of the model—the citizen communication mediation model—across cultures. Relying on panel survey data from 19 countries, this study advances a multilevel citizen communication mediation model. Our findings indicate that discussion remains a strong mediating predictor of political participation across countries and political contexts, though the relationship is moderated by macro-level repressive democratic expressive norms.
Content may be subject to copyright.
Journal of Communication ISSN 0021-9916
ORIGINAL ARTICLE
The Citizen Communication Mediation Model
Across Countries: A Multilevel Mediation
Model of News Use and Discussion on
Political Participation
Homero Gil de Zúñiga
1,2
, Trevor Diehl
3
, Brigitte Huber
4
, & James H. Liu
5
1 University of Vienna
2 Universidad Diego Portales - Chile
3 Central Michigan University
4 University of Vienna
5 Massey University
Since introduced by Professor McLeod and the Wisconsin School at the turn of the cen-
tury, a large body of research has employed the communication mediation model. Yet,
most of these studies rely on cross-sectional and individual-level survey data collected in
the United States. This paper seeks to address these shortcomings by testing a specica-
tion of the modelthe citizen communication mediation modelacross cultures.
Relying on panel survey data from 19 countries, this study advances a multilevel citizen
communication mediation model. Our ndings indicate that discussion remains a
strong mediating predictor of political participation across countries and political con-
texts, though the relationship is moderated by macro-level repressive democratic expres-
sive norms.
Keywords: Communication Mediation Model, Multilevel Communication Mediation
Model, Multilevel Mediation, News Use, Political Discussion, Political Participation.
doi:10.1093/joc/jqz002
Over the past three decades, few theoretical concepts in political communication
have spurred more scholarly attention than the communication mediation model
(for an overview, see Lee, 2017;Shah et al., 2017). The basic idea behind the com-
munication mediation model is that the relationship between individual orienta-
tions and political participation is mediated by communication patterns, such as
media use and political discussions (McLeod, Kosicki, & McLeod, 1994;McLeod
et al., 2001;Sotirovic & McLeod, 2001). This study focuses on a specic case of the
model that speaks to the role of interpersonal discussion in explaining why
Corresponding author: Homero Gil de Zúñiga; e-mail: homero.gil.de.zuniga@univie.ac.at
1Journal of Communication 00 (2019) 124 © The Author(s) 2019. Published by Oxford University Press on behalf of
International Communication Association. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com
Downloaded from https://academic.oup.com/joc/advance-article-abstract/doi/10.1093/joc/jqz002/5371208 by University Library of Vienna user on 11 March 2019
informational uses of media often lead to political participation. The so-called citi-
zen communication model (Shah, Cho, Eveland, & Kwak, 2005) is widely applied in
the study of discursive and participatory democracy. The model has become
ingrained in the eld of political communication, but it is not without serious
limitations.
First, most research on the model relies solely on U.S. data, making it dicult to
establish robust ndings on a global scale and rendering concurrent validity depen-
dent on the cultural context. Second, macro-level data, highlighting how social value
systems elevate or depress democratic norms, is lacking in the literature (Lee, 2017).
For instance, the citizen communication mediation model may work dierently in
distinct political contexts, such as advanced democratic societies, as opposed to
more limited democratic or non-democratic societies. Important macro-level fac-
tors, such as freedom of the press and freedom of expression, may also reveal how
democratic social norms drive the relationship between communication and
participation.
With these research goals in mind, this study (a) examines the applicability of
the citizen communication model across 19 countries (two waves, N=8,708), and
(b) explores the country-level factors of freedom of expression and freedom of the
press as potential moderators. We do so by applying a broad concept of political
participation (i.e., donating money to political campaigns, voting, and practicing
boycotting) and various forms of news use (traditional news use, as well as social
media news use). The results indicate that discussion remains a strong indicator of
political participation, though the eect may be stymied by repressive democratic
expressive norms.
News use and political participation
The relationship between media use and political and civic participation have
increasingly attracted scholarly attention, since Putnam (2000) primarily blamed
TV for destroying civic engagement in the United States. According to the time dis-
placement hypothesis, TV was thought to circumvent the productive use of leisure
time by disrupting opportunities for connecting with others. Moreover, based on
the so-called mean world eect(Gerbner, Gross, Morgan, & Signorielli, 1980),
Putnam proposed that heavy TV watchers develop an ingrained skepticism of other
people and a general pessimism about human nature.
However, empirical research has shown that time spent with television does not
inuence civic engagement through perceptions of time pressure eects (Moy,
Scheufele, & Holbert, 2009). Media eects are determined by exposure to specic
content and the way people use media, rather than by media use per se (Shah,
McLeod, & Yoon, 2001). More specically, research shows that informational uses
of media are positively related to civic and political participation, whereas using
media for entertainment is inversely related to participation (McLeod, Scheufele, &
Moy, 1999;Zhang & Chia, 2006). The positive relationship between news use and
2Journal of Communication 00 (2019) 124
The Citizen Communication Mediation Model Across Cultures H. Gil de Zúñiga et al.
Downloaded from https://academic.oup.com/joc/advance-article-abstract/doi/10.1093/joc/jqz002/5371208 by University Library of Vienna user on 11 March 2019
political participation has been established in several studies, for dierent types of
news use, including newspaper reading (Choi, 2016), Internet news use (Bakker &
de Vreese, 2011), social media news use (Gil de Zúñiga, Garci ́
a, & McGregor, 2015),
and multi-platform news use, such as mobile phones (Kim, Chen, & Wang, 2016)
and dual-screening (McGregor & Mourão, 2017). The current manuscript includes
traditional as well as emerging forms of news consumption in both online and social
media environments. But how does news use foster peoples political participatory
habits?
Reasons for the positive link between news use and political participation are
manifold. Being up-to-date with news and public aairs provides mobilizing infor-
mation (Lemert, 1992); news oers important informational resources that are key
for people to engage in political activities. For example, information about where
and when to attend a political rally, vote in an election, or get involved in a local
campaign is often transmitted through the news media. Consuming news also
allows citizens to orient themselves in the political world (Sotirovic & McLeod,
2001) and to understand how the political process works (Barabas & Jerit, 2009).
News consumption strengthens peoples feelings of condence in their ability to
understand politics, reduces levels of political cynicism, and, likewise, fosters politi-
cal trust in government institutions (Chan, 2014;Halpern, Valenzuela, & Katz,
2017;Idid, Ahmed, & Souket, 2017,Gil de Zúñiga et al., 2019). These characteristics
are ideal elements in the development of a more participatorythough not always
healthierdemocracy.
Based on these ndings, we expected a positive relationship between news use
and political participation.
H1: News use is positively related to political participation.
Political discussion and political participation
As far back as Converse (1964), scholars have speculated that social mechanisms,
like inter-personal discussions, help articulate belief systems that operate to increase
the expediency of political information. Thus, political discussion plays an essential
role in spurring democratic processes. Studies from the deliberative paradigm con-
sistently show that political discussion is positively associated with political engage-
ment across a variety of discursive and participatory contexts (Kim, Wyatt, & Katz,
1999;Lu, Heatherly, & Lee, 2016;McClurg, 2003;Pattie & Johnston, 2009;Shah
et al., 2005). One reason why political discussion fosters political participation is
that talking about politics helps individuals understand complex issues, concepts,
and arguments (Kim et al., 2016). For example, an increased frequency of political
discussion is positively related to political learning (Meirick & Wackman, 2004).
Discussing with others may introduce new political facts, which leads to further
elaboration. In other words, the act of discussion helps people make connections
3Journal of Communication 00 (2019) 124
H. Gil de Zúñiga et al. The Citizen Communication Mediation Model Across Cultures
Downloaded from https://academic.oup.com/joc/advance-article-abstract/doi/10.1093/joc/jqz002/5371208 by University Library of Vienna user on 11 March 2019
between the information obtained while talking to others and previously held politi-
cal knowledge (Hively & Eveland, 2009).
Prior research also indicates that one motive for citizens to engage in political
discussion is the willingness to learn (Conover, Searing, & Crewe, 2002). When
engaging in discussion, individuals organize their thoughts and weigh the pros and
cons of an argument (Benhabib, 1996). Talking enables individuals to develop and
express their views, learn the positions of others, identify shared concerns and pre-
ferences, and come to understand and reach judgments about matters of public con-
cern(Delli Carpini, Cook, & Jacobs, 2004, p. 319). Finally, discussing politics also
creates an opportunity to persuade others (Lindström, 1997;Zúñiga, Barnidge &
Diehl, 2018). People who attempt to persuade others prot from engaging in discus-
sion with those who have dierent views, and become more informed citizens them-
selves (Thorson, 2012).
The underlying conceptual thread in this research is that the cognitive benets
of discussion lead to political behavior, because people are better able to identify
opportunities to address public problems or personal concerns. There is a counter-
argument in the literature that suggests that when people are exposed to discussion
disagreement, they are actually more likely to detach from politics altogether (Mutz,
2006), and that real-life political discussion does not align with the deliberative
norm. However, empirical studies rather reliably nd that talking about politics has
the potential to foster both online and oine political participation (Pattie &
Johnston, 2009). Hence, the current manuscript includes both forms of political
participation: talking about politics and public aairs face-to-face and online.
Considering the ndings discussed above, we posed the following hypothesis:
H2: Political discussion is positively related to political participation.
The citizen communication mediation model
A prominent theoretical framework that integrates the relationships discussed
abovethe direct relationship between news use and political participation on one
hand, and the relationship between political discussion and participation on the
otheris the citizen communication mediation model. The model, proposed by
Shah et al. (2005), is drawn from concepts that explicated the more general commu-
nication mediation model (McLeod et al. 2001;Sotirovic & McLeod, 2001). The
general communication mediation model is an approach that is widely used to theo-
rize the potential, causal mechanisms of media eects on various outcome variables
(Lee, 2017;Shah et al., 2017). These concepts rely heavily on social scientists
renewed interest in psychological path modeling in the late 1980s.
The general communication mediation theory states that communication pat-
terns, such as media use and inter-personal discussion, mediate (or partially
explain) the direct relationships between individual orientations (including demo-
graphics and previous attitudinal dispositions) and outcome behaviors (McLeod
4Journal of Communication 00 (2019) 124
The Citizen Communication Mediation Model Across Cultures H. Gil de Zúñiga et al.
Downloaded from https://academic.oup.com/joc/advance-article-abstract/doi/10.1093/joc/jqz002/5371208 by University Library of Vienna user on 11 March 2019
et al., 2001;Sotirovic & McLeod, 2001). This line of thinking can be traced back to
the psychological O-S-O-R model (Markus & Zajonc, 1985), which was introduced
to political communication by McLeod et al. (1994). The assumption of the model
is that a previously held set of orientations (O; e.g., demographics) inuences how
people are stimulated (S) through media use, which in turn inuences their subse-
quent orientations to politics (second O; e.g., political knowledge). Both the stimu-
lus and subsequent orientations may act as mediators between the stimulus (S) and
an outcome response variable (R) of interest (e.g., civic or political participation; see
McLeod et al., 1999).
Scholars have made eorts to advance the theoretical framework: The O-S-O-R
model has been extended to an O-S-R-O-R model (Cho et al., 2009;Jung,
Yonghwan, & Gil de Zúñiga, 2011) by adding a second R for reasoning variables
(e.g., political discussion, motivated reasoning).
First, one specic application of the general communication mediation model is
the citizen communication mediation model. The citizen communication model
emphasizes the centrality of expression and discussion for the distillation of ideas
encountered in the news(Shah et al., 2007, p. 678). In our study, we focused on
this specic case of the communication mediation model, which centers on political
discussion as the mediator of the direct connection between news use and political
participation (Shah et al., 2005). Prior research has shown that media use for infor-
mational purposes positively predicts political discussion (Jung et al., 2011;Kim
et al., 1999;McLeod et al., 1999). This is because news can provide people with new
ideas, facts, and opinions that stimulate political discussion (Beaudoin & Thorson,
2006). Second, people encounter dissimilar political views in the news, which they
would not be exposed to otherwise, and this can stimulate political discussion
(Mutz & Martin, 2001).
The indirect path (mediation) from news use to participation via political dis-
cussion has been well supported by empirical ndings as well. Shah et al. (2005)
showed that both reading newspapers and seeking information online are positively
related to interpersonal political discussion, which in turn leads to higher levels of
civic participation. Similarly, people who watch TV news or read news online are
more likely to talk about politics, which in turn increases their chances to become
politically active (Shah et al., 2007). Finally, this mediating mechanism is not only
observable in face-to-face-situations, but also in the form of online political discus-
sions (Kim et al., 2016;Lee, Shah, & McLeod, 2012). Hence, the authors also include
online and oine participation in the empirical models.
Following this line of research, we assume that individuals who consume more
news also tend to discuss politics and public aairs more often, which in turn leads
them to participate in politics. Accordingly, we formulate the following hypothesis:
H3: Political discussion mediates the relationship between news use and political
participation.
5Journal of Communication 00 (2019) 124
H. Gil de Zúñiga et al. The Citizen Communication Mediation Model Across Cultures
Downloaded from https://academic.oup.com/joc/advance-article-abstract/doi/10.1093/joc/jqz002/5371208 by University Library of Vienna user on 11 March 2019
Macro-level variables and the communication mediation model: Freedom of
the press and democratic expression
The communication mediation model has rarely been tested outside the United
States and, with some exceptions (McLeod & Lee, 2012), few comparative studies
exist. Therefore, validating the assumptions of the citizen communication mediation
model across countries and cultures continues to be a pressing research goal. As Lee
(2017, p. 134) argued, there is a general need to test whether the communication
mediation model holds at the international level.The current study addresses this
research gap by testing the citizen communication mediation model in 19 countries.
Individual participatory behaviors are likely shaped by immediate social and politi-
cal contexts. Scholars have long theorized about the role of political and economic
structures in organizing the types of societies people live in (Dahl, 1989). Yet, the
absence of comparative empirical ndings in political communication places the cit-
izen communication mediation model on tenuous ground.
First, individual political behavior is highly inuenced by the social norms culti-
vated around politics (Verba, Schlozman, & Brady, 1995). That is, institutions
such as the home, schools, labor unions, and civic associationsplay a major role in
fostering democratic behaviors (Putnam, 1995;Schlozman, Verba, & Brady, 2012).
Without these institutions, free association may not be cultivated and, in some
cases, may be legally barred (Levitsky & Way, 2002). In contexts where politically
expressive behaviors are limited, political participation is either forced or non-
existent. Accordingly, institutional freedom of expression (Coppedge et al., 2016)is
a likely indicator of participatory behaviors in general. It may be, for example, that
processes derived from the citizen communication mediation model buck country-
level trends to support pro-democratic behaviors, regardless of the context. On the
other hand, the relative prevalence of free association in society may stymie this
mechanism. In other words, it remains an open research question as to whether
group-level norms determine the theorys viability.
Second, prior research suggests that low freedom of the press is associated with
lower levels of political learning and political participation, and the reverse is true
for countries with higher freedom of the press (Leeson, 2008;Schoonvelde, 2014).
Rather intuitively, without the free ow of public information, people will lack the
general resources and knowledge to participate in politics (Leeson, 2008).
Accordingly, freedom of the press (Freedom House, 2017) may be a contextual
macro factor for the citizen communication mediation model to take place. This
index considers the legal environment for the media, political pressures that inu-
ence journalism, and economic factors that inuence access to news and
information.
Country-level and institutional contexts help forward our understanding of the
eects of news media use in society on at least two accounts. First, if the theoretical
assumptions made by the citizen communication mediation model hold across
countries, scholars have strong empirical support for the role of news in fostering
6Journal of Communication 00 (2019) 124
The Citizen Communication Mediation Model Across Cultures H. Gil de Zúñiga et al.
Downloaded from https://academic.oup.com/joc/advance-article-abstract/doi/10.1093/joc/jqz002/5371208 by University Library of Vienna user on 11 March 2019
communicative action. In other words, cross-cultural ndings oer some scientic
validity of the cognitive underpinnings of the model. Second, and perhaps more
importantly, there are normative implications for cross-cultural research. Where
democratic values are at risk, or not present at all, results from the citizen commu-
nication mediation model may identify opportunities or bottlenecks for political
activity.
By including these macro variables in our analysis, the study addresses a long-
standing gap in the literature. There is a general need for multilevel studies. Studies
that make a connection between the eects of individual-level variables and macro-
level variables are scant (Lee, 2017). Given the potential for institutional press indi-
cators to stand in for the free ow of information, as well as the role of free demo-
cratic expression in developing participatory norms, these two macro-level
indicators are likely to inuence outcomes in the citizen communication mediation
model. In order to assess how macro-level variables might inuence the ways in
which people partake of political behaviors, the following hypothesis was
formulated:
H4: Country-level contexts of (a) freedom of expression and (b) press freedom
indexes are positively related to individual-level political participation.
Moreover, it is expected that when citizens are socialized into contexts where
political and social institutions curate a public sphere conducive to a free press and
open individual expression, the association between news, political discussion, and
political participation will be stronger. This is particularly the case for political dis-
cussion, since that behavior is intrinsically tied to news (press freedom) and open
institutions (democratic expression index) to exist. Simply put, we expected to nd
the following relationship:
H5: Political discussion is positively related to political participation in countries
with greater (a) freedom of expression and (b) freedom of the press.
Last, given the condition that the citizen communication mediation model was
originally theorized as an individual-level process, no other study, at least to the
authorsknowledge, has tested the process in a more nuanced context. That is, it is
still of interest to test whether the importance of individual-level mechanisms
remain as relevant when introducing more complex models that integrate distinct
micro- and macro-level factors within the process of the citizen communication
mediation model. Since the quality and quantity of news and discussion may be
inuenced by system-level indicators of democratic freedom, it stands to reason
that the mediation process may not operate the same in all socio-political contexts.
For example, recent research has suggested that freedom of expression moderates
the relationship between network heterogeneity and political expression on social
media, so that the relationship is stronger in countries with low freedom of expres-
sion (Barnidge, Huber, Gil de Zúñiga, & Liu, 2018). Wei, Lo, Xu, Chen, and Zhang
(2014) showed that alternative ways of getting news are of higher importance in
7Journal of Communication 00 (2019) 124
H. Gil de Zúñiga et al. The Citizen Communication Mediation Model Across Cultures
Downloaded from https://academic.oup.com/joc/advance-article-abstract/doi/10.1093/joc/jqz002/5371208 by University Library of Vienna user on 11 March 2019
countries with lower levels of press freedom than in countries with high levels of
press freedom (e.g., alternative to state-controlled media in China via mobile
phones). Similarly, the role of political discussion might dier in countries with dif-
ferent levels of press freedom. Hence, while the citizen communication mediation
model might not work the exact same way in all countries, there is little evidence
that the counter-factual to the citizen communication mediation model exists; the
informational and cognitive benets implied in the model may happen less often,
but one struggles to imagine a context where people do not respond to information
cues in the environment. Given the lack of empirical ndings in this area, this study
also tests a multilevel mediation model for the communication mediation process
(Figure 1;Rockwood & Hayes, 2017).
RQ1: How does the citizen communication mediation model operate in a multi-
country, multilevel setting?
Methods
Sample and data
The study relies on an international, two-wave, online panel survey conducted in
late September of 2015 (Wave 1, N=20,361) and March of 2016 (Wave 2, N=
8,708). Valid responses were collected from 19 participating countries in the second
wave of data collection (Table 1; see also Supporting Information) and the within-
wave cooperation rate was acceptable (77%; Cooperation rate (CR3); American
Association of Public Opinion Research, 2016). Nielsen administered non-
probability sampling procedures based on stratied quota sampling, in cooperation
with local national partners. Due to the complexity and scale of the collection pro-
cess, as well as the nature of online paneling sampling in general, typical response
Macro Level
c
News Use
(group mean
centered)
Political
Discussion (group
mean centered)
Political participation
Use (group mean
centered)
ab
Individual Level
Press
Freedom
Democratic
Expression
Figure 1 Multilevel mediation model theoretical representation. The inward and outward
rectangles represent the rst- and second-level measurements, respectively. c=direct eect
of News use (group mean centered) on Politcal participation (group mean centered). ab =
Indirect eect of ews use (group mean centered) on Politcal participation (group mean cen-
tered) through Politcal Discussion (group mean centered).
8Journal of Communication 00 (2019) 124
The Citizen Communication Mediation Model Across Cultures H. Gil de Zúñiga et al.
Downloaded from https://academic.oup.com/joc/advance-article-abstract/doi/10.1093/joc/jqz002/5371208 by University Library of Vienna user on 11 March 2019
Table 1 Descriptive and Reliability Statistics for Variables of Interest and Country-Level Indicators (19 countries)
Political
Participation
W1
Political
Participation
W2
Political
Discussion
W1
News Use
W1
Country
Indicators
Country N
W2
M/SD rM/SD rM/SD rM/SD rPFDE
Argentina 293 3.10/.94 .75 3.15/.92 .73 3.32/1.35 .88 4.73/1.04 .65 50 .91
Brazil 308 3.20/1.14 .81 3.28/1.27 .85 3.65/1.40 .88 5.10/1.10 .74 46 .96
China 373 2.80/1.30 .92 2.74/1.29 .93 3.16/1.22 .91 4.72/.98 .76 87 .25
Estonia 694 2.88/.65 .66 2.66/.65 .65 2.7/.87 .79 4.60/.98 .57 16 .97
Germany 627 2.79/1.04 .78 2.77/1.04 .78 2.96/1.20 .85 4.34/1.06 .64 20 .97
Indonesia 282 2.95/1.19 .86 3.12/1.26 .86 3.30/1.36 .91 4.97/.99 .74 49 .85
Italy 562 3.04/1.13 .82 3.05/1.08 .81 3.12/1.29 .88 4.20/.87 .56 31 .95
Japan 552 2.25/.85 .75 2.23/.87 .75 2.02/1.00 .86 3.94/1.10 .66 26 .86
South Korea 536 2.67/1.01 .79 2.73/.96 .77 2.68/1.18 .88 4.14/1.12 .74 33 .80
New Zealand 569 2.68/.80 .70 2.64/.83 .74 2.24/.92 .82 4.14/.91 .41 20 .96
Philippines 142 3.08/1.16 .85 3.48/1.20 .86 3.69/1.14 .82 4.95/.95 .72 44 .88
Poland 607 2.89/1.05 .81 2.94/1.03 .82 2.96/1.18 .87 4.68/.93 .65 28 .95
Russia 520 2.50/1.01 .81 2.51/.97 .81 2.72/1.19 .89 4.44/1.08 .70 83 .35
Spain 284 2.98/1.02 .81 2.96/.98 .76 3.04/1.18 .86 4.59/1.09 .70 28 .96
Taiwan 395 2.88/1.06 .82 2.81/.94 .79 2.44/1.05 .88 4.21/.90 .66 26 .84
Turkey 327 3.51/1.08 .80 3.44/.99 .75 3.65/1.37 .89 4.65/.97 .70 71 .49
United Kingdom 644 2.69/.95 .76 2.65/.94 .77 2.34/1.06 .85 3.96/1.02 .56 25 .99
Ukraine 96 2.34/1.01 .69 2.48/1.02 .70 2.67/1.12 .87 4.15/.97 .56 53 .61
United States 483 2.75/.09 .76 2.71/.94 .75 2.30/1.02 .83 3.82/1.05 .56 21 .94
Total 8,295 2.80/1.04 .79 2.79/1.03 .78 2.79/1.23 .88 4.44/1.07 .64 ––
Note. Statistics are based on complete observations for political participation W2. Reliability (r) is Cronbachs alpha. Higher scores on PF and DE
equate to less freedom and more openness, respectively. DE =democratic expression; PF =press freedom; W1 =Wave 1; W2 =Wave 2.
9Journal of Communication 00 (2019) 124
H. Gil de Zúñiga et al. The Citizen Communication Mediation Model Across Cultures
Downloaded from https://academic.oup.com/joc/advance-article-abstract/doi/10.1093/joc/jqz002/5371208 by University Library of Vienna user on 11 March 2019
rates were not calculated (see American Association of Public Opinion Research,
2016,p.4950). Though designs of this nature have limitations, they are increas-
ingly becoming the industry standard (Callegaro et al., 2014) as Internet penetration
rates increase. The survey was designed by an international consortium of scholars,
headed by the University of Vienna and Massey University. The surveys were trans-
lated into native languages, in coordination with project partners at local universi-
ties, and employed back-translation methods when possible (Behling & Law, 2000).
For more information on the sample and a demographic breakdown by country, see
Supporting Information Appendix.
Measures
Outcome and mediation model variables
Political participation. The outcome variable of interest in the study is an omnibus
measure of political participation, designed to capture an array of institutional and
extra-institutional participatory behaviors (Shah et al., 2007;Sotirovic & McLeod,
2001;Verba et al. 1995). The survey asked respondents how often (1 =never/7 =all
the time) in the past three months they had engaged in the following activities: (a)
contacted an elected public ocial, (b) donated money to a campaign or political
cause, (c) participated in groups that took any local action for social or political
reform, (d) attended a political rally or participated in any demonstrations, protests,
or marches, (e) bought a certain product or service because of the social or political
values of the company, (f) boycotted a certain product or service because of the
social or political values of the company, (g) voted in local or statewide elections,
and (h) voted in national or presidential elections (8-item averaged construct). This
item was measured at both Wave 1 (political participation;
W1
) and Wave 2 (politi-
cal participation;
W2
; see Table 1for descriptive and reliability statistics in the
pooled samples and sub-samples).
Political discussion. The frequency of inter-personal conversations about political
issues and topics was operationalized based on previous literature (Eveland &
Hively, 2009;Gil de Zúñiga, Valenzuela, & Weeks, 2016;Kwak et al., 2005).
Respondents were asked how often they talk about politics or public aairs face-to-
face with (a) family, relatives, or friends, (b) a spouse or partner, (c) acquaintances,
and (d) strangers. Questions were asked about these interactions both online
(4 items) and oine (4 items). The nal measure was an average item index of online
and oine political discussions (Wave 1, 7-point scale, 8-item index; Table 1).
News use. News use was conceptualized as a respondents attention to online and
oine media channels for news and public aairs information. The following sur-
vey questions were based on previously employed measures (Prior, 2005;Yuan,
2011). Respondents were reminded that people get news from various sources and
were asked how often they get news from the following media sources (1 =never/7
=all the time): (a) television news (cable or local network news), (b) printed
10 Journal of Communication 00 (2019) 124
The Citizen Communication Mediation Model Across Cultures H. Gil de Zúñiga et al.
Downloaded from https://academic.oup.com/joc/advance-article-abstract/doi/10.1093/joc/jqz002/5371208 by University Library of Vienna user on 11 March 2019
newspapers, (c) online news websites, (d) radio, (e) social media, and (f) citizen
journalism websites (6-item scale, Wave 1; Table 1).
Predictor variables
Demographics. The survey also included questions related to perceived income rela-
tive to others, as is commonly used in comparative designs (Keller, Medgyesi, &
Tóth, 2010), as well as level of education, gender, race, and age. Social demographics
often explain political behavior, and these factors are therefore included in the mod-
els (Schlozman et al., 2012). Based on the pooled sample estimates for those who
participated in both waves of the survey, the median age was 45.07 years (SD =
14.58); 50% of participants were female; race (coded as 1 =majority); the median
level of education was 4.39, as measured on an eight-point scale (1 =none and 7 =
post-graduate degree; SD =1.32); and the median level of income was 5.42 (with par-
ticipants rating their perceived income on a scale with 10 being people who are the
most well oin society and 1 being the people who are the least well o;SD =1.81).
Political antecedents. Political participation has been extensively studied, relative to
a range of social, structural, and cognitive factors, at the individual level (Schlozman
et al., 2012; Verba et al., 1995). To avoid spurious relationships, the statistical mod-
els also include three antecedents to political participation and discussion: political
ecacy (Wave 1, 2-item average scale: [a] people like me can inuence government
and [b] I consider myself well qualied to participate in politics; M=3.47, SD =
1.47, Spearman-Brown r =.69), strength of political ideology (Wave 1, measured on
a 10-point scale as how progressive/conservative respondents consider themselves
on social, political, and economic issues, folded index: 0 =moderate/ambivalent,
6=strong ideologue; M=2.8, SD =1.51, Cronbachsα=.91), and network size
(logged; Wave 1, open-ended questions: during the past month, about how many
total people have you talked to: [a] face-to-face or over the phone about politics or
public aairs, [b] via the Internet, including email, chat rooms, social networking
sites, and micro-blogging sites). Due to the skewed nature of this variable, the natu-
ral log was used on both items before creating the scale (M=.46, SD =.42,
Spearman-Brown r =.71).
Country-level indicators. Due to the nested structure of the sample (country-level),
the study was able to take advantage of adding additional macro-level indicators.
These indicators are directly related to the variables of interest in the current study.
First, a widely used measure of freedom of democratic expression approximates the
health of open democratic expression in terms of freedom of discussion, govern-
ment censorship of the media, press censorship, and academic expression (www.v-
dem.net). The freedom of expression index is based on nine indicators: government
censorship eorts regarding media, harassment of journalists, media self-
censorship, freedom of discussion for men, freedom of discussion for women, free-
dom of academic and cultural expression, media bias, print/broadcast media critical,
and print/broadcast media perspectives (Coppedge et al., 2016). The V-dem index
11Journal of Communication 00 (2019) 124
H. Gil de Zúñiga et al. The Citizen Communication Mediation Model Across Cultures
Downloaded from https://academic.oup.com/joc/advance-article-abstract/doi/10.1093/joc/jqz002/5371208 by University Library of Vienna user on 11 March 2019
scores for the countries included in our study range from .25 (China) to .99 (United
Kingdom; Table 1). In countries with higher scores, people have the institutional
support and inter-personal freedom to express themselves.
Second, a fruitful approach to measure the degree of media freedom is provided
by the Freedom House (2017). Their annual report delivers an overview of media
independence worldwide by assessing the degree of freedom for print, broadcast,
and digital media. Overall, press freedom was used to estimate group levels of the
political, legal, and economic environment for the commercial press (freedom-
house.org). The press freedom index scores for the countries included in our study
range from 16 (Estonia) to 87 (China), where the lowest score indicates the highest
level of press freedom and the highest score indicates the lowest level of press free-
dom (Table 1).
Statistical approach
This study relies on a three-part analysis strategy. First, in order to test the direct
relationship between news and political participation (H1) and political discussion
and political participation (H2), a series of ordinary least squares (OLS) regression
models were employed. Given the panel nature of the data, these relationships were
tested with the pooled dataset in the cross-sectional (Wave 1), lagged (Wave 1 pre-
dictors on Wave 2 outcome variable), and auto-regressive frameworks (Wave 1 pre-
dictors on Wave 2 outcome variable, controlling for the outcome variable in Wave
1). The mediation hypothesis (H3) was tested in the auto-regressive framework,
employing the software developed by Tingley, Yamamoto, Hirose, Keele, and Imai
(2014) and double checked with a test developed by Rockwood and Hayes (2017).
Second, a series of tests were conducted using hierarchical linear modeling (multi-
level modeling [MLM]) procedures to account for country dierences and to test
H4, H5, and RQ1. Cross-level interaction eects (H5) were probed at the 10
th
,50
th
,
and 90
th
percentiles for visualization purposes. All regression models included con-
trols for demographics and political antecedents.
This study controlled for several variables at the individual and country levels.
That is, covariates in the MLM models were centered to the grand mean. An alter-
native plausible approach would be to residualize these variables by regressing the
endogenous variables of interest over these controls and then testing for the mediat-
ing mechanism. Perhaps a more precise course of action would be to covariate these
distinct level variables, allowing for and estimating xed and random eects in the
mediation (Hayes & Scharkow, 2013). This paper follows the latter approach.
Results
H1 predicted that news use is positively related to political participation. In the
pooled sample OLS model (Table 2), news use is a positive, statistically signicant
predictor of political participation in the cross-sectional (Model 1, b=.019,
SE =.007, p<.01), lagged (Model 2, b=.087, SE =.011, p<.001), and auto-
12 Journal of Communication 00 (2019) 124
The Citizen Communication Mediation Model Across Cultures H. Gil de Zúñiga et al.
Downloaded from https://academic.oup.com/joc/advance-article-abstract/doi/10.1093/joc/jqz002/5371208 by University Library of Vienna user on 11 March 2019
regressive frameworks (Model 3, b=.077, SE =.09, p<.001). H2 predicted that
political discussion is positively related to political participation. Political discussion
showed the strongest association with political participation in the cross-sectional
model (Model 1, b=.395, SE =.008, p<.001), and the association remained posi-
tive and statistically signicant for both the lagged (Model 2, b=.202, SE =.012,
p<.001) and the more stringent auto-regressive frameworks (Model 3, b=.028,
SE =.011, p<.01). H1 and H2 are conrmed in the pooled time-series models.
Together, news and discussion account for about 8.5% of the explained variance in
Model 1, with a total variance explained by the model of R
2
at 38%. In Model 2
(lagged), the total variance explained is R
2
at 29%; our independent variables of
Table 2 Cross-Sectional, Lagged, and Autoregressive OLS Regression Models Predicting
Political Participation (pooled sample)
Political Participation
Cross-Sectional Lagged Autoregressive
Models (1) (2) (3)
Block 1: Autoregressive term
Political participation
W1
––.503 (.010)***
ΔR
2
––17.6%
Block 2: Demographics
Age .001 (.001) .001 (.001) .001 (.001)
Gender, female .047 (.014)*** .021 (.021) .003 (.018)
Education .005 (.005) .015 (.008) .013 (.007)
Income .008 (.007) .052 (.010)*** .048 (.008)***
Race .032 (.020)* .020 (.030) .005 (.026)
ΔR
2
2.7% 3.5% 3.5%
Block 3: Political antecedents
Political interest
W1
.038 (.006)*** .034 (.009)*** .015 (.008)
Political ecacy
W1
.156 (.006)*** .107 (.008)*** .029 (.007)***
Strength of ideology
W1
.035 (.005)*** .058 (.007)*** .041 (.006)***
Network size, large
W1
.144 (.021)*** .278 (.031)*** .206 (.027)***
ΔR
2
27.2% 21.8% 21.8%
Block 4: Variables of interest
News use
W1
.019 (.007)** .087 (.011)*** .077 (.009)***
Political discussion
W1
.395 (.008)*** .202 (.012)*** .028 (.011)**
ΔR
2
8.5% 4.5% 4.5%
Total R
2
38% 29% 47%
Observations 15,009 6,391 6,396
Note: Cell entries are nal-entry ordinary least squares, unstandardized beta coecients,
with standard errors in parentheses. Model 1 DV are from Wave 1; Models 2 & 3
Dependent Variable (DV) are from Wave 2. W1 =Wave 1; W2 =Wave 2. *p<.05; **p
<.01; ***p<.001.
13Journal of Communication 00 (2019) 124
H. Gil de Zúñiga et al. The Citizen Communication Mediation Model Across Cultures
Downloaded from https://academic.oup.com/joc/advance-article-abstract/doi/10.1093/joc/jqz002/5371208 by University Library of Vienna user on 11 March 2019
interest predicted R
2
at 4.5%. Finally, Model 3 explained R
2
at 47%, where news and
discussion account for 4.5% of the variance.
When looking at the relationship between news use and political participation,
one has to consider that this relationship is expected to be mediated through politi-
cal discussion (H3). The result from a mediation test (output model =Model 3;
Table 2) conrms that about 17% of the direct relationship between news and politi-
cal participation (H1) is explained by political discussion. The point estimate of the
indirect eect is .012 (CI .008.016; p=000) and of the total eect is .071
(CI .05.08; p=000). H3 was conrmed.
The remaining research questions and hypotheses were concerned with the
country-level factors that inuence political participation. Table 3reports results
Table 3 Null, Multilevel, and Cross-Level Interaction Autoregressive Models Predicting
Political Participation
Political Participation
W2
Models (1) (2) (3)
Intercept 2.859 (.073)*** 2.865 (.036)*** 2.855 (.037)***
Individual-level indicators
Age .001 (.001)* .001 (.001)*
Gender, female .016 (.019) .016 (.019)
Education .023 (.008)** .023 (.008)**
Income .021 (.009)* .021 (.009)*
Race .009 (.034) .008 (.034)
Political interest
W1
.016 (.009) .017 (.009)*
Political ecacy
W1
.034 (.008)*** .034 (.008)***
Strength ideology
W1
.038 (.007)*** .039 (.007)***
Network size, large
W1
.089 (.031)** .084 (.031)**
News
W1
.057 (.011)*** .058 (.011)***
Political discussion
W1
.053 (.012)*** .055 (.012)***
Political participation
W1
.541 (.012)*** .540 (.012)***
Country-level indicators
Democratic expression .760 (.211)** .701 (.215)**
Press freedom .007 (.002)** .006 (.002)*
Interaction terms
Democratic expression x discussion
W1
.231 (.092)*
Press freedom x discussion
W1
.003 (.001)**
Variance of random components (SD)
Within countries
σ
i
2
.979 (.989) .551 (.742) .549 (.741)
Between countries
σ
u
2
.098 (.314) .005 (.067) .004 (.068)
Log likelihood 11,719.340 7,173.696*** 7,169.155*
Note. These data are based on 19 groups. Coecients are unstandardized: b(SE). Log likeli-
hood tests are model t tests based on previous model. N=8,295. W1 =Wave 1; W2 =
Wave 2. *p<.1; **p<.05; ***p<.01.
14 Journal of Communication 00 (2019) 124
The Citizen Communication Mediation Model Across Cultures H. Gil de Zúñiga et al.
Downloaded from https://academic.oup.com/joc/advance-article-abstract/doi/10.1093/joc/jqz002/5371208 by University Library of Vienna user on 11 March 2019
from the MLM models. Model 1 represents the null model (random intercept, Log
Likelihood [LL] =11,719.340). That model was tested against an OLS model,
without random intercepts (LL =12,006.10). Analyses of variance tests of the dif-
ferences between model t statistics Likelihood Ratio Test (LL) reveal that there are
dierences between countries in their levels of political participation (Wave 2; L.
Ratio =573.37; p<.001). The average level of participation is 2.86 (intercept,
Table 3; Model 1) and the between-country standard deviation (SD) is .31. The
country means for the dependent variables are reported in Table 1.
H4 proposed that country-level contexts of freedom of expression (H4a) and
press freedom (H4b) indexes would positively predict individual-level political par-
ticipation. Model 2 (Table 3) includes all individual and macro-level covariates.
Both freedom of expression (b=.760, SE =.211, p<.01) and press freedom
(b=.007, SE =.002, p<.01) were positive, statistically signicant predictors in the
MLM framework. H5 proposed that political discussion would be positively related
to participation in countries with greater freedom of expression (H5a) and freedom
of the press (H5b). Both interaction terms were statistically signicant predictors
(Table 3, Model 3). The democratic expression index moderates the relationship
between political discussion and participation (b=.231, SE =.092, p<.05).
Similarly, freedom of the press moderates the relationship between political discus-
sion and political participation (b=.003, SE =.001, p<.01).
Political Discussion W1
Participation W2
−2
0
2
4
6
Dem.Expression:Low
−2 −1 0 1 2 3 4
Dem.Expression:Med
−2 −1 0 1 2 3 4 −2 −1 0 1 2 3 4
Dem.Expression:High
Figure 2 The plot shows the cross-level interactions on political participation (y-axis)
between political discussion (x-axis) and democratic expression index for the 10
th
,50
th
, and
90
th
percentiles of the democratic expression index (where a higher score indicates more
openness). W1 =Wave 1; W2 =Wave 2.
15Journal of Communication 00 (2019) 124
H. Gil de Zúñiga et al. The Citizen Communication Mediation Model Across Cultures
Downloaded from https://academic.oup.com/joc/advance-article-abstract/doi/10.1093/joc/jqz002/5371208 by University Library of Vienna user on 11 March 2019
These relationships are plotted in Figures 2and 3. Figure 2shows that, in coun-
tries with low freedom of expression, political discussion is negatively related to
political participation (left panel): that is, discussing politics with others diminishes
peopleschances to participate in politics. Conversely, in countries with medium
and high levels of democratic expression freedom, the more people discuss politics
with others, the more they participate in politics. Hence, hypothesis H5a, which
posited a positive relationship between political discussion and participation in
countries with greater freedom of expression, was conrmed. Figure 3shows inter-
action results for the press freedom index (H5b): in less open societies (right panel;
high press freedom index, reverse coded), talking with others about politics makes
people more likely to participate in politics, while this is not the case in countries
with more open press systems (left panel; low press freedom index; reverse coded).
Hence, H5b was rejected.
Finally, the research question asked how the citizen communication mediation
model operates in a multi-country, multilevel setting. Model 2 (Table 3) showed that
the relationships between news and political participation (H1, b=.057, SE =.011,
p<.001), as well as discussion and participation (H2, b=.053, SE =.012, p<.001),
were conrmed. These relationships remain after controlling for country-level
Political Discussion W1
Participation W2
0
2
4
6
Press Freedom Index:Low
−2 −1 0 1 2 3 4
Press Freedom Index:Med
−2 −1 0 1 2 3 4 −2 −1 0 1 2 3 4
Press Freedom Index:High
Figure 3 The plot shows the cross-level interactions on political participation (y-axis)
between political discussion (x-axis) and a press freedom index for the 10
th
,50
th
, and 90
th
percentiles of the press freedom index (where a higher score indicates less freedom). W1 =
Wave 1; W2 =Wave 2.
16 Journal of Communication 00 (2019) 124
The Citizen Communication Mediation Model Across Cultures H. Gil de Zúñiga et al.
Downloaded from https://academic.oup.com/joc/advance-article-abstract/doi/10.1093/joc/jqz002/5371208 by University Library of Vienna user on 11 March 2019
dierences, individual-level demographics and political antecedents, macro-
indicators, and political participation in Wave 1. Figure 2reports results from the
multilevel mediation analysis. Before running the analysis, the authors tested com-
peting models (versus Model 2, Table 3) for random slope eects within the coun-
tries for discussion (L. Ratio =1.89, p=.381) and news (L. Ratio =1.88, p=.390).
Since the variance of these random parameters failed to diverge from zeroand
failed to improve the model tthe output mediation model (total eects) was spec-
ied as Model 2 (Table 3). The average mediation eect (indirect eect) was positive
and statistically signicant after accounting for country dierences, and discussion
mediates about 14% of the direct eect (Figure 4).
Discussion
This study advances our empirical and theoretical accounting of the citizen commu-
nication mediation model, using an international panel survey with multilevel
modeling. Results indicate that general exposure to the news and political discussion
are consistently related to political participation across countries. This positive rela-
tionship was found in cross-sectional, lagged, and auto-regressive pooled sample
models (Table 2), and the positive relationship remained after accounting for coun-
try dierences in levels of participation (Table 3). In addition, political discussion
mediates the relationship between news and participation, accounting for about
14% of the direct eect in both the pooled sample and the multilevel framework
(Figure 4). System-level factors also help explain this process. Both democratic
expression norms and indicators of press freedom are directly and positively associ-
ated with political participation. Finally, these two macro-indicators interact with
political discussions in divergent ways. While in countries where citizens are less
free to express themselves, political discussion is negatively related to political par-
ticipation, in more democratically open societies, discussion is positively associated
with political participation. Moreover, our results show that the level of press
Political
ParticipationW2
Political
DiscussionW1
(b =.057, se = .011, p < .001)
NewsW1
(b =.053, se = .012, p < .001) (b =.18, se = .011, p < .001)
Figure 4 The model represents a auto-regressive, multilevel mediation analysis for 19
groups (N=6,391). The path entries are unstandardized coecients. This is based on quasi-
Bayesian CIs to 500 simulations. The model includes the same controls and predictors as
Model 3, in Table 2. The point estimate of the indirect eect across groups is .009
(CI .005.013; p=000) and of the total eect is .067 (CI .047.087; p=000). CI =con-
dence interval; W1 =Wave 1; W2 =Wave 2.
17Journal of Communication 00 (2019) 124
H. Gil de Zúñiga et al. The Citizen Communication Mediation Model Across Cultures
Downloaded from https://academic.oup.com/joc/advance-article-abstract/doi/10.1093/joc/jqz002/5371208 by University Library of Vienna user on 11 March 2019
freedom of a country also matters: the relationship between political discussion and
political participation is strongest in less open societies.
In the context of the current study, the results oer a few important conclusions.
First, the majority of the variance in the MLM models is attributed to the individual
level (Table 3, Model 1). In other words, individual-level communication processes
drive the relationships between information consumption, discussion, and political
participation. Second, news facilitates informational and social needs for public life
in a variety of country contexts, and over time. Information ows from the news
media are essential to identifying relevant issues of the day, and to connecting those
issues with relevant opportunities to participate in politics. Third, political discus-
sion is the most powerful indicator of whether an individual will participate in a
broad array of political activities. Inter-personal discussion in a variety of social set-
tings provides additional and important social information, as it helps people con-
textualize the news in a way that becomes deeply benecial to participatory
processes. Finally, the results provide evidence that the benets of discussion help
explain how news use leads to participation across 19 countries, lending a consider-
able level of concurrent validity to the citizen communication mediation model, as
tested here.
For decades, social scientists have employed some version of the communication
mediation model to connect theoretical assumptions to empirical path models.
Though not without limitation, these path models provide a useful tool for scientists
interested in understanding how news and information shape a variety of outcomes,
from political knowledge (Eveland, 2001) to network attributes and both civic and
political participation (Shah et al., 2007;Shahetal.,2001). The current study adds to
this literature in several meaningful ways. First, the study veries the underlying mech-
anisms of inter-personal social relations in shaping political behavior. Based on the
auto-regressive MLM estimates, the mediation process was observed, even after con-
trolling for variation across the 19-country sample. The model may be tapping under-
lying communicative processes that exist across cultural and system-level contexts.
The paper also has important implications for comparative work on political
communications processes. As expected, macro indicators of democratic expression
seemed to support long-standing normative assumptions about free and open politi-
cal discussion as an ideal driver of participatory behaviors. That is, in less open soci-
eties, conversation matters very little for political activity (H5a). This may be due to
spiral-of-silence eects, or perhaps conversation leads to increased cynicism and
retraction from politics. Either way, future work should explore the inverse relation-
ship between discussion and participation in closed societies. Results also indicated
an inverse relationship between press freedom and political discussion (H5b,
rejected). When journalism is suppressed, political conversation is vital in determin-
ing political behaviors. When news media is less available, conversation becomes
much more important for learning about and responding to political opportunities.
Yet, we expected these macro variables to operate in a similar way. More work
18 Journal of Communication 00 (2019) 124
The Citizen Communication Mediation Model Across Cultures H. Gil de Zúñiga et al.
Downloaded from https://academic.oup.com/joc/advance-article-abstract/doi/10.1093/joc/jqz002/5371208 by University Library of Vienna user on 11 March 2019
needs to be done to validate macro-level indexes, since they oer inconclusive
results that countered our conceptual expectations.
While the paper provides powerful evidence of citizen communication media-
tion processes in a cross-cultural framework, the breadth of the study also presents
several limitations. First, the panel design of online sampling techniques means that
several parameters in the data collection are unknown. This problem is com-
pounded in larger international projects, since the reliability of the local partners
may be inconsistent (American Association of Public Opinion Research, 2016).
Note that the Wave 2 panel sample sizes varied (Table 1) and, therefore, we can
make no strong argument for generalizability in a number of countries (especially
large ones like China). This is not a fatal aw, since we did not model separate
countries, and instead relied on the MLM framework (N=8,295). Thus, we feel
that the sample size is sucient enough to make modest claims of cross-country
eects. Also, the research design includes an auto-regressive term. As much as it
helps dissipate causal order quandaries, this is not the same thing as causality.
Perhaps using more waves could oer further causal evidence. In any case, the cur-
rent design is robust for standards in the eld. There is also a persistent problem
with studies of political participation in general, since most people do not partici-
pate in politics, let alone read the news or discuss important issues. Since this study
is interested in these processes explicitly, the authors were less concerned with those
individuals who are not actively engaged in public life. For instance, the political
participation measurement employed in this paper was broad and inclusive. It also
included voting at local and national elections. These behaviors should be studied
by separate indexes, rather than in a unied index. This is particularly the case for
voting, since no elections were held between waves in most countries. However, in
the context of this study, the results were practically identical when either including
or excluding voting from the measurements.
Finally, the citizen communication mediation model is also limited. As with
path models in general, there are several causal assumptions made about individual
motivations and reactions to news content along each pathin the model. More
work needs to be done to carefully test these assumptions, especially in experimental
designs that better isolate individual cognitive eects. Despite these limitations, no
other paper has attempted to simultaneously test the eects of both macro- and
micro-level contextual indicators of the citizen communication mediation model.
This paper addresses a signicant research gap in terms of validating a long-
standing theoretical model outside a few Western countries. The results establish a
stepping stone for future comparative research.
Supplementary Material
Supplementary material are available at Journal of Communication online.
19Journal of Communication 00 (2019) 124
H. Gil de Zúñiga et al. The Citizen Communication Mediation Model Across Cultures
Downloaded from https://academic.oup.com/joc/advance-article-abstract/doi/10.1093/joc/jqz002/5371208 by University Library of Vienna user on 11 March 2019
Acknowledgments
This research was supported by Grant FA2386-15-1-0003 from the Asian Oce of
Aerospace Research and Development.
References
American Association of Public Opinion Research. (2016). Standard denitions: Final
dispositions of case codes and outcome rates for surveys. Retrieved from https://www.
aapor.org
Bakker, T. P., & de Vreese, C. H. (2011). Good news for the future? Young people, Internet
use, and political participation. Communication Research, 38(4), 451470. doi:10.1177/
0093650210381738
Barabas, J., & Jerit, J. (2009). Estimating the causal eects of media coverage on policy-
specic knowledge. American Journal of Political Science, 53,7389. doi:10.1111/j.1540-
5907.2008.0035.x. https://www.jstor.org/stable/25193868
Barnidge, M., Huber, B., Gil de Zúñiga, H., & Liu, J. H. (2018). Social media as a sphere for
riskypolitical expression: A twenty-country multilevel comparative analysis. The
International Journal of Press/Politics, 23(2), 161182. doi:10.1177/1940161218773838
Beaudoin, C. E., & Thorson, E. (2006). The social capital of Blacks and Whites: Diering
eects of the mass media in the United States. Human Communication Research, 32(2),
157177. doi:10.1111/j.1468-2958.2006.00007.x
Behling, O., & Law, S. K. (2000). Translating questionnaires and other research instruments:
Problems and solutions. Thousand Oaks, CA: Sage.
Benhabib, S. (1996). Toward a deliberative model of democratic legitimacy. In S. Benhabib
(Ed.), Democracy and dierence (pp. 6794). Princeton, NJ: Princeton University Press.
Callegaro, M., Baker, R. P., Bethlehem, J., Göritz, A. S., Krosnick, J. A., & Lavrakas, P. J.
(2014). Online panel research. Chichester: John Wiley & Sons West Sussex. United
Kingdom.
Chan, M. (2014). Exploring the contingent eects of political ecacy and partisan strength
on the relationship between online news use and democratic engagement. International
Journal of Communication, 8,11951215. https://ijoc.org/index.php/ijoc/article/view/2559
Cho, J., Shah, D. V., McLeod, J. M., McLeod, D. M., Scholl, R. M., & Gotlieb, M. R. (2009).
Campaigns, reection, and deliberation: Advancing an O-S-R-O-R model of
communication eects. Communication Theory, 19(1), 6688. doi:10.1111/j.1468-2885.
2008.01333.x
Choi, J. (2016). Dierential use, dierential eects: Investigating the roles of dierent modes
of news use in promoting political participation. Journal of Computer-Mediated
Communication, 21(6), 436450. doi:10.1111/jcc4.12176
Conover, P. J., Searing, D. D., & Crewe, I. M. (2002). The deliberative potential of political
discussion. British Journal of Political Science, 32(1), 2162. doi:10.1017/
S0007123402000029
Converse, P. E. (2006). The nature of belief systems in mass publics (1964). Critical review,
18(13), 174. doi:10.1080/08913810608443650
Coppedge, M., Gerring, J., Lindberg, S. I., Skaaning, S.-E., Teorell, J., Altman, D., et al.
(2016). V-Dem codebook v6. Retrieved from www.v-dem.org.
Dahl, R. A. (1989). Democracy and its critics. New Haven, CT: Yale University Press.
20 Journal of Communication 00 (2019) 124
The Citizen Communication Mediation Model Across Cultures H. Gil de Zúñiga et al.
Downloaded from https://academic.oup.com/joc/advance-article-abstract/doi/10.1093/joc/jqz002/5371208 by University Library of Vienna user on 11 March 2019
Delli Carpini, M. X., Cook, F. L., & Jacobs, L. R. (2004). Public deliberation, discursive
participation, and citizen engagement: A review of the empirical literature. Annual
Review of Political Science, 7, 315344. doi:10.1146/annurev.polisci.7.121003.091630
Eveland, W. P., Jr. (2001). The cognitive mediation model of learning from the news:
Evidence from nonelection, o-year election, and presidential election contexts.
Communication Research, 28(5), 571601. doi:10.1177/009365001028005001
Eveland Jr, W. P., & Hively, M. H. (2009). Political discussion frequency, network size, and
heterogeneityof discussion as predictors of political knowledge and participation.
Journal of Communication,59(2), 205224. doi:10.1111/j.1460-2466.2009.01412.x
Freedom House. (2017). About freedom of the press. Retrieved from https://freedomhouse.
org/report-types/freedom-press
Gerbner, G., Gross, L., Morgan, M., & Signorielli, N. (1980). The mainstreamingof
America: Violence prole no. 11. Journal of Communication, 30(3), 1029. doi:10.1111/j.
1460-2466.1980.tb01987.x
Gil de Zúñiga, H., Garci ́
a, V., & McGregor, S. (2015). What is second screening? Exploring
motivations of second screen use and its eects on online political participation. Journal
of Communication, 65(5), 793815. doi:10.1111/jcom.12174
Gil de Zúñiga, H., Barnidge, M., & Diehl, T. (2018). Political Persuasion on Social Media: A
Moderated Moderation Model of Political Disagreement and Civil Reasoning. The
Information Society, 34(5), 302315. doi:1080/01972243.2018.1497743
Gil de Zúñiga, H., Ardèvol-Abreu, A., Diehl, T., Gómez-Patiño, M., & Liu, H. J. (2019).
Trust in Institutional Actors across 22 Societies: Examining Political, Media, and Science
Trust around the World. Revista Latina de Comunicación Social,74, 237262. doi:10.
4185/RLCS-2019-1329
Gil de Zúñiga, H., Valenzuela, S., & Weeks, B. (2016). Motivations for political discussion:
Antecedents and consequences on civic engagement. Human Communication Research,
42(4), 533552. doi:10.1111/hcre.12086
Halpern, D., Valenzuela, S., & Katz, J. E. (2017). We face, I tweet: How dierent social media
inuence political participation through collective and internal ecacy. Journal of
Computer-Mediated Communication, 22, 320336. doi:10.1111/jcc4.12198
Hayes, A. F. & Scharkow, M. (2013). The relative trustworthiness of inferential tests of the
indirect eect in statistical mediation analysis: Does method really matter? Psychological
Science, 24(10), 19181927. doi:10.1177/0956797613480187
Idid, S. A., Ahmed, I. S. Y., & Souket, R. (2017). Examining political ecacy among
Malaysian voters. The role of traditional and new media. Journal of Asian Pacic
Communication, 27(1), 4364. doi10.1075/japc.27.1.03idi
Jung, N., Yonghwan, K., & Gil de Zúñiga, H. (2011). The mediating role of knowledge and
ecacy in the eects of communication on political participation. Mass Communication
and Society, 14(4), 407430. doi:10.1080/15205436.2010.496135
Keller, T., Medgyesi, M., & Tóth, I. G. (2010). Analyzing the link between measured and
perceived income inequality in European countries [Research note no. 8]. Brussels,
Belgium: European Commission.
Kim, J., Wyatt, R. O., & Katz, E. (1999). News, talk, opinion, participation: The part played
by conversation in deliberative democracy. Political Communication, 16(4), 361385.
doi:10.1080/105846099198541
21Journal of Communication 00 (2019) 124
H. Gil de Zúñiga et al. The Citizen Communication Mediation Model Across Cultures
Downloaded from https://academic.oup.com/joc/advance-article-abstract/doi/10.1093/joc/jqz002/5371208 by University Library of Vienna user on 11 March 2019
Kim, Y., Chen, H.-T., & Wang, Y. (2016). Living in the smartphone age: Examining the
conditional indirect eects of mobile phone use on political participation. Journal
of Broadcasting & Electronic Media, 60(4), 694713. doi:10.1080/08838151.2016.1203318
Lee, N. J. (2017). Communication mediation model. In P. Rössler, C. A. Honer, & A. L. van
Zoonen (Eds.), The International Encyclopedia of Media Eects (pp. 128137).
Chichester, England: Wiley Blackwell.
Lee, N. J., Shah, D. V., & McLeod, J. M. (2012). Processes of political socialization: A
communication mediation approach to youth civic engagement. Communication
Research, 40(5), 669697. doi:10.1177/0093650212436712
Leeson, P. T. (2008). Media freedom, political knowledge, and participation. Journal of
Economic Perspectives, 22(2), 155169. doi:10.1257/jep.22.2.155
Lemert, J. B. (1992). Eective public opinion. In J. D. Kennmer (Ed.), Public opinion, the
press, and public policy (pp. 4161). West Port, CT: Praeger Publishers.
Levitsky, S., & Way, L. (2002). The rise of competitive authoritarianism. Journal of
Democracy, 13(2), 5165. doi:10.1353/jod.2002.0026
Lindström, P. (1997). Persuasion via facts in political discussion. European Journal of Social
Psychology, 27(2), 145163. doi:10.1002
Lu, Y., Heatherly, K. A., & Lee, J. K. (2016). Cross-cutting exposure on social networking
sites: The eects of SNS discussion disagreement on political participation. Computers in
Human Behavior, 59,7481. doi:10.1016/j.chb.2016.01.030
Markus, H., & Zajonc, R. B. (1985). The cognitive perspective in social psychology. In G.
Lindzey & E. Aronson (Eds.), The Handbook of Social Psychology (pp. 137229). New
York, NY: Random House.
McClurg, S. D. (2003). Social networks and political participation: The role of social
interaction in explaining political participation. Political Research Quarterly, 56(4),
448464. doi:10.1177/106591290305600407
McGregor, S. C., & Mourão, R. R. (2017) Second screening Donald Trump: Conditional
indirect eects on political participation. Journal of Broadcasting & Electronic Media, 61
(2), 264290. doi:10.1080/08838151.2017.1309418
McLeod, J. M., Kosicki, G. M., & McLeod, D. M. (1994). The expanding boundaries of
political communication eects. In J. Bryant & D. Zillman (Eds.), Media eects, advances
in theory and research (pp. 123162). Hillsdale, NJ: Lawrence Erlbaum.
McLeod, J. M., & Lee, N. J. (2012). Multilevel designs for comparative communication
research. In F. Esser & T. Hanitzsch (Eds.), Handbook of comparative communication
research (pp. 430444). New York, NY: Routledge.
McLeod, J. M., Scheufele, D. A., & Moy, P. (1999). Community, communication, and
participation: the role of mass media and interpersonal discussion in local political
participation. Political Communication, 16(3), 315336. doi:10.1177/
009365085012001001
McLeod, J. M., Zubric, J., Keum, H., Deshpande, S., Cho, J., Stein, S., & Heather, M. (2001,
August). Reecting and connecting: Testing a communication mediation model of civic
participation. Paper presented at the Association for Education in Journalism and Mass
Communication (AEJMC) conference, Washington, DC.
Meirick, P. C., & Wackman, D. B. (2004). Kids voting and political knowledge: Narrowing
gaps, informing votes. Social Science Quarterly, 85(5), 11611177. doi:10.1111/j.0038-
4941.2004.00269.x
22 Journal of Communication 00 (2019) 124
The Citizen Communication Mediation Model Across Cultures H. Gil de Zúñiga et al.
Downloaded from https://academic.oup.com/joc/advance-article-abstract/doi/10.1093/joc/jqz002/5371208 by University Library of Vienna user on 11 March 2019
Moy, P., Scheufele, D. A., & Holbert, R. L. (2009). Television use and social capital:
Testing Putnams time displacement hypothesis. Mass Communication and Society, 2,
2745. doi:10.1080/15205436.1999.9677860
Mutz, D. C. (2006). Hearing the other side: Deliberative versus participatory democracy.
Cambridge, England: Cambridge University Press.
Mutz, D. C., & Martin, P. S. (2001). Facilitating communication across lines of political
dierence: The role of mass media. American Political Science Review, 95,97114.
doi:10.1017/S000305540707044X
Pattie, C., & Johnston, R. (2009). Conversation, disagreement and political participation.
Political Behavior, 31(2), 261285. doi:10.1007/s11109-008-9071-z
Prior, M. (2005). News vs. entertainment: How increasing media choice widens gaps in
political knowledge and turnout. American Journal of Political Science, 49, 577592.
doi:10.1017/S000305540707044X
Putnam, R. D. (1995). Tuning in, tuning out: The strange disappearance of social capital in
America. PS: Political science & politics,28(4), 664683. doi:10.2307/420517
Putnam, R. D. (2000). Bowling alone: The collapse and revival of American community. New
York, NY: Simon & Schuster.
Rockwood, N. J. & Hayes, A. F. (2017, May). MLmed: An SPSS macro for multilevel
mediation and conditional process analysis. Poster presented at the annual meeting of the
Association of Psychological Science, Boston, MA.
Schlozman, K. L., Verba, S., & Brady, H. E. (2012). The unheavenly chorus: Unequal political
voice and the broken promise of American democracy. Princeton, NJ: Princeton
University Press.
Schoonvelde, M. (2014). Media freedom and the institutional underpinnings of political
knowledge. Political Science Research and Methods, 2(2), 163178. doi:10.1017/psrm.
2013.18
Shah, D. V., Cho, J., Eveland, W. P., Jr., & Kwak, N. (2005). Information and expression in a
digital age. Modeling Internet eects on civic participation. Communication Research, 32
(5), 531565. doi:10.1177/0093650205279209
Shah, D. V., Cho, J., Nah, S., Gotlieb, M. R., Hwang, H., Lee, N.-J., ... McLeod, D. M. (2007).
Campaign ads, online messaging, and participation: Extending the communication
mediation model. Journal of Communication, 57(4), 676703.
Shah, D. V., McLeod, D. M., Rojas, H., Cho, J., Wagner, M., & Friedland, L. A. (2017).
Revising the communication mediation model for a new political communication
ecology. Human Communication Research, 43(4), 491504. doi:10.1111/hcre.12115
Shah, D. V., McLeod, J. M., & Yoon, S. H. (2001). Communication, context and community:
An exploration of print, broadcast and Internet inuences. Communication Research, 28
(4), 464506. doi:10.1177/009365001028004005
Shah, D. V., Cho, J., Eveland, W. P., Jr., & Kwak, N. (2005). Information and expression in a
digital age: Modeling Internet eects on civic participation. Communication research, 32
(5), 531565. doi:10.1177/0093650205279209
Sotirovic, M., & McLeod, J. M. (2001). Values, communication behavior, and political
participation. Political Communication, 18(3), 273300. doi:10.1080/
10584600152400347
23Journal of Communication 00 (2019) 124
H. Gil de Zúñiga et al. The Citizen Communication Mediation Model Across Cultures
Downloaded from https://academic.oup.com/joc/advance-article-abstract/doi/10.1093/joc/jqz002/5371208 by University Library of Vienna user on 11 March 2019
Thorson, E. (2014) Beyond opinion leaders: How attempts to persuade foster political
awareness and campaign learning. Communication Research,41(3), 353374. doi:10.
1177/0093650212443824
Tingley, D., Yamamoto, T., Hirose, K., Keele, L., & Imai, K. (2014). Mediation: R package for
causal mediation analysis. UCLA Statistics/American Statistical Association. Retrieved
from http://dspace.mit.edu/handle/1721.1/91154
Verba, D. S., Schlozman, K. L., & Brady, P. H. E. (1995). Voice and equality: Civic
voluntarism in American politics. Cambridge, Massachusetts, USA: Harvard University
Press.
Wei, R., Lo, V.-H., Xu, X., Chen, Y.-N. K., & Zhang, G. (2014). Predicting mobile news use
among college students: The role of press freedom in four Asian cities. New Media &
Society, 16(4), 637654. doi:10.1177/1461444813487963
Yuan, E. (2011). News consumption across multiple media platforms: A repertoire
approach. Information, Communication & Society, 14(7), 9981016. doi:10.1080/
1369118x.2010.549235
Zhang, W., & Chia, S. C. (2006) The eects of mass media use and social capital on civic and
political participation. Communication Studies, 57(3), 277297. doi:10.1080/
10510970600666974
24 Journal of Communication 00 (2019) 124
The Citizen Communication Mediation Model Across Cultures H. Gil de Zúñiga et al.
Downloaded from https://academic.oup.com/joc/advance-article-abstract/doi/10.1093/joc/jqz002/5371208 by University Library of Vienna user on 11 March 2019
... The theoretical model was originally proposed in the 1990s by McLeod and colleagues (1994) and has since been widely used by other researchers during the last 30 years (Borah, Barnidge, & Rojas, 2021). From this theoretical approach, it is assumed which initial orientations of the audience (O1) determine what media or content is consumed (Gil de Zúñiga et al., 2019). These initial orientations are established by a set of cultural, cognitive, motivational, and structural traits that lead people to consume media content (Lee, 2017). ...
... Since its incorporation into the field of political communication, the model has captured the attention of many researchers and has become a widely used theoretical model (Borah et al., 2021;Gil de Zúñiga et al., 2019). Despite this, the empirical application of the model has a few limitations. ...
... A second limitation is determined by the lack of consideration of contextual-level factors, somewhat because the model was originally theorized as a process that only considered individual-level factors (Borah et al., 2021;Gil de Zúñiga et al., 2019). However, recent studies have begun to determine the possible "effects of psychological and social structural factors toward various outcome variables" (Lee, 2017, p. 128). ...
Article
Full-text available
Election campaigns provide a crucial moment in the relationship between political actors and the citizenry, with the media playing a key role in the transmission of relevant political information to the public—an activity that has a bearing on shifts in political attitudes and behaviors. The communication mediation model explains this media influence process as an indirect effect of media consumption on participation through the development of postconsumption orientations and reasoning on the message. Following the model, this article seeks to analyze whether the same process took place during the 2021 Mexican federal election campaign by conducting a two-wave panel survey, with a representative sample of 1,750 participants in the first wave and 596 in the second wave. The findings made it possible to determine how the model manages to explain the effect of political interest on electoral engagement through attention to campaign news and political conversation, both jointly and separately.
... A common finding is that news consumption through traditional news media and digital media leads to interpersonal political discussion and online opinion expression, which then lead to political participation indirectly through political knowledge and efficacy (Chan, 2016;. In a related vein, research on a communication mediation model, a variant grounded in the OSROR framework, focuses on the role of communicative reasoning and has revealed that interpersonal political discussion and online opinion expression can mediate the effects of news consumption on political participation without the second orientation factor involved in the process (Chan, 2016;Cho et al., 2009;Gil de Zúñiga et al., 2019;Shah et al., 2005Shah et al., , 2007. ...
... Our study draws from the OSROR and communication mediation models to assess the processes by which journalistic credibility and UGC consumption and production are related to political participation. In doing so, we integrate traditional news use and interpersonal political discussion, two traditional communication variables found to be central in this process (e.g., Cho et al., 2009;Gil de Zúñiga et al., 2019;Shah et al., 2007). Also, as citizen journalism is practiced solely online or in collaboration with traditional news outlets (e.g., Nah et al., 2015), we examine UGC consumption and production on professional and citizen news websites separately. ...
... The first centers on traditional media and interpersonal communication. The indirect link between professional journalism credibility and political participation through traditional news use and interpersonal political discussion is consistent with previous studies that stresses the role of citizen-to-citizen conversations in the political process (e.g., Cho et al., 2009;Gil de Zúñiga et al., 2019). Although confirmatory in nature, this finding is significant given the current political communication environment where news exposure tends to occur via social media and people's political discussion networks are becoming smaller and more homogeneous (Lee & Bearman, 2020;Shearer & Mitchell, 2021). ...
Article
This study examines how journalistic credibility and consumption of user-generated content (UGC) are related to political participation. Data from a national online survey shows two main pathways. One concerns traditional communicative pathways involving professional journalism credibility, traditional news use, and interpersonal political discussion. The other concerns citizen journalism credibility and UGC consumption and production on citizen news websites. Thus, while traditional political communication continues to play a role in political participation, citizen journalism practice presents unique opportunities for citizens to engage in politics.
... Recent research has begun to test the models in other political contexts. In particular, Gil de Zúñiga et al., (2019) found that the CCMM holds true across 19 countries, although the relationship is moderated by country characteristics such as the level of press freedom. The authors emphasize the importance of considering country variances in applying the model, recommending that more research needs to be done to explicate the nuances in closed societies. ...
... This study follows previous scholars' work (Gil de Zúñiga et al., 2019;Shah et al., 2005) and extends a specific line of the CMM research-the CCMM-to Mainland China. The model includes three key variables in the CMM framework (i.e., S-R-R) and highlights the significance of online political talk in mediating the media effects in this digital world. ...
... Incorporating the examination of two types of citizenship norms clarifies an important assumption rooted in political communication research about democratic countries, thus revealing the uniqueness and complexities of China's changing political and media environment. In particular, our theorization and empirical findings of the mediated effect moderated by the democratic citizenship norm systematically explicates how CCMM operates in a closed society, adding more nuances to the existing literature (Gil de Zúñiga et al., 2019). Future researchers could consider using our proposed model (see Figure 1) to explore the variations of citizenship norms and their effects on the CCMM and the CMM framework in general in countries with different political systems. ...
Article
This study incorporates the examination of citizenship norms in testing the Citizen Communication Mediation Model (CCMM) in China, exploring to what extent online political expression mediates the impact of informational use of social media on offline civic engagement and how beliefs in citizenship norms moderate the CCMM. Results based on a two-wave panel survey among a national sample of 1,199 Chinese adults provide strong support for the CCMM in the Chinese context. In addition, embracing the democratic citizenship norm significantly enhances the CCMM effect, whereas embracing the pro-government citizenship norm that encourages pro-government speech does not show the same effect.
... In the family, political socialization takes place in two main ways: resulting from family activities that convey political attitudes or beliefs or through the transmission of socioeconomic advantages (Brady et al., 2015;Janoski & Wilson, 1995;Mustillo et al., 2004;Son & Wilson, 2012). On the one hand, a series of practices or characteristics within the family, such as levels of conversation on social or political topics (Gil de Zúñiga et al., 2019;Shehata & Amnå, 2017), operate as sources transmitting political preferences or identification. Likewise, it has been proven in university students that their parents' level of participation in collective actions predicts the youths' involvement in radical activities. ...
... En el marco de la familia, la socialización política se produce por dos vías principales, producto de actividades familiares que transmiten actitudes o creencias políticas y/o por medio de la transmisión de las ventajas socioeconómicas (Brady et al., 2015;Janoski & Wilson, 1995;Mustillo et al., 2004;Son & Wilson, 2012). Por un lado, una serie de prácticas o características que se dan al interior de la familia, como los niveles de conversación sobre temas sociales y/o políticos (Gil de Zúñiga et al., 2019;Shehata & Amnå, 2017), operan como fuentes de transmisión de las preferencias o la identificación política. Así mismo, se ha mostrado en jóvenes universitarios que los niveles de participación de los padres en acciones colectivas, predice el involucramiento de jóvenes en acciones radicales. ...
Article
Full-text available
Can exposure to mass events with the presence of radical actions influence the legitimation of this type of protest in adolescents? How are changes in legitimation levels conditioned by socioeconomic status or by political socialization at home and school? To answer these questions, we used a sample of 1,203 students in their second year of secondary school in Chile who were surveyed as part of the School Citizenship Panel (Panel de Ciudadanía Escolar, PACES). In the middle of fieldwork in this study, the event known as the estallido social (‘social outburst’) stood out for its mass and presence of violence. We compare the students surveyed before and after this event. Contrary to our expectations, the students had similar levels of legitimation of radical actions before and after the estallido. We also found a trend towards moderation in students with a higher socioeconomic status, who had more frequent political conversations with their parents, and in students from schools that conduct civic education activities. In short, experiences of family and school political socialization can help lower an adolescent’s support of radical actions in radicalized contexts.
... Samples from nine other countries (Czech Republic, Denmark, Finland, Hong Kong, Israel, Kuwait, Italy, Taiwan, and the United Kingdom) were used only in one study. Two studies compared data from more than one country Saldaña et al., 2015), 8 and two studies analyzed pooled data from more than one country Huber et al., 2019). 9 ...
... 10. As noted in the description of sample characteristics, two studies aggregate data from several countries Huber et al., 2019) and therefore cannot be included in analyses that test for country-level moderators. 11. ...
Article
Full-text available
[In press; article accepted on May 20, 2022] The rapid rise of digital media use for political participation has coincided with an increase in concerns about citizens' sense of their capacity to impact political processes. These dual trends raise the important question of how people's online political participation is connected to perceptions of their own capacity to participate in and influence politics. The current study overcomes the limitation of scarce high-quality cross-national and overtime data on these topics by conducting a meta-analysis of all extant studies that analyze how political efficacy relates to both online and offline political participation using data sources in which all variables were measured simultaneously. We identified and coded 48 relevant studies (with 184 effects) representing 51,860 respondents from 28 countries based on surveys conducted between 2000 and 2016. We conducted a multilevel random effects meta-analysis to test the main hypothesis of whether political efficacy has a weaker relationship with online political participation than offline political participation. The findings show positive relationships between efficacy and both forms of participation, with no distinction in the magnitude of the two associations. In addition, we tested hypotheses about the expected variation across time and democratic contexts, and the results suggest contextual variation for offline participation but cross-national stability for online participation. The findings provide the most comprehensive evidence to date that online participation is as highly associated with political efficacy as offline participation, and that the strength of this association for online political participation is stable over time and across diverse country contexts.
... Sementara itu kalangan muda di Swedia kurang antusias untuk berdiskusi tentang politik di media sosial, karena ada resiko konflik, kesalahpahaman, dan berita yang tidak benar (Sveningsson, 2014). Di pihak lain banyak hasil studi yang menyatakan peran positif diskusi politik dalam memediasi penggunaan media dan partisipasi politik (Vaccari & Valeriani, 2018;Zúñiga, Diehl, Huber, & Liu, 2019). Belum adanya kesepakatan peran diskusi politik dalam hubungan penggunaan media dan partisipasi politik dari studi terdahulu juga menjadi dasar perhatian dilakukannya penelitian dalam buku ini. ...
... Banyak penelitian sebelumnya melihat diskusi politik sebagai komunikasi interpersonal tentang politik yang mempengaruhi partisipasi (Morey & Eveland, 2016;Zúñiga et al., 2019). Sebuah studi terhadap kalangan muda Indonesia menunjukkan bahwa media sosial banyak digunakan untuk memperkuat interaksi dengan teman dan keluarga sebelum mengambil keputusan politik, berdiskusi secara langsung dengan keluarga dan kolega terkait informasi politik dari media sosial menjadi rujukan untuk menentukan pilihan politik (Loina Lalolo Krina Perangin-angin & Zainal, 2018). ...
Preprint
Full-text available
Dalam buku ini membahas mengenai pengaruh media khususnya media sosial dan televisi serta radio dan surat kabar terhadap partisipasi politik. Secara konseptual penelitian dalam buku ini menggunakan teori Media System Dependency yang menyatakan bahwa semakin besar kebutuhan seseorang terhadap informasi, semakin seseorang bergantung kepada media, dan semakin penting peran media dalam kehidupannya, sehingga media dan pesan-pesannya akan lebih memberikan pengaruh kepada kognisi, sikap, dan perilaku; pengaruh media tersebut dapat terjadi secara langsung atau tidak langsung. Dengan melihat realitas politik pada masa pemilihan presiden Indonesia 2019, dilihat juga peran diskusi politik, pengetahuan politik, dan efikasi politik dalam memediasi hubungan pengaruh penggunaan media terhadap partisipasi politik.
Article
This study examines whether and how exposure to information about health disparities from mediated and interpersonal sources and political trust are jointly associated with policy support and political advocacy. Using a two-wave panel survey design with a national U.S. sample, we found that political trust moderated the association between exposure to health-disparities information from interpersonal sources and policy support. Specifically, a negative relationship between exposure from interpersonal sources and policy support was observed among respondents with a lower than average level of political trust. In contrast, respondents with a higher than average level of political trust showed a positive association between exposure from interpersonal sources and policy support. Among respondents with average level of political trust, the relationship between exposure from interpersonal sources and policy support was generally flat. In addition, exposure to health-disparities information from media sources was positively related to political advocacy after controlling for possible confounders.
Article
Using a representative Taiwan sample from the 2019 Asia Barometer Survey, this study investigated the effect of political interest on two types of political participation (i.e., participation in electoral campaigns and political advocacy) based on the communication mediation model. Findings showed that three communication channels (i.e., news involvement, on-line political expression, and off-line political discussion) mediate the positive relationship between political interest and participation through different paths. When people are more interested in politics, they will be more likely to read political news, leading to more on-line political expression, which then motivates higher level of political participation. In parallel, political interest could also increase political participation first through news involvement and then through off-line political discussion. More importantly, compared with the total indirect effects on participation in electoral campaigns, political interest could exert stronger indirect effect on participation in political advocacy through the three communication channels under study. In addition, the mediating effect of on-line political expression is more powerful than off-line political discussion in mobilizing participation in political advocacy. However, in terms of participation in electoral campaigns, indirect effect through on-line and off-line communication showed no significant difference. Implications of the findings are discussed.
Article
Integrating the extant research on partisan selective exposure, discussion networks and communication mediation model, this survey study (N = 1,203) tests the relationship between proattitudinal partisan media use (Fox News, MSNBC) and discussion networks (like-minded, cross-cutting) in the context of political participation. My results find evidence of asymmetrical relationships, such that Republicans watching Fox News are more likely to participate in politics through like-minded discussion, while MSNBC use is more likely to mobilize viewers directly. This study advances our understanding of the effects of partisan media use and discussion networks on political participation in the era of partisan polarization.
Article
Full-text available
In this article, we explore the relationship between social media use and participation in protests in Chile. In October 2019, Chile faced the most massive protests since the country’s return to democracy. Due to its magnitude, the media and analysts refer to this process as the “social outburst.” Although these protests engaged broad sectors of the population, most of the protesters were young people. Using a probabilistic and face-to-face survey applied to young people aged 18-29 years, we find that the only social media platform associated with participation in protests was Facebook. Our analysis also shows the importance of the specific activities that people engage in social media. Taking part in political activities on social media is strongly associated with attending protests but using social media platforms to get information or share common interests with other users is not. Furthermore, we examine whether social media has an indirect impact on participation through interpersonal conversation. The results show that Instagram—one of the most popular social media platforms among young Chileans—spurs interpersonal conversation, which in turn increases the likelihood of participating in protests. Our findings suggest that social media still plays a role in shaping people’s political behavior despite changes in the social media environment and in social media consumption patterns.
Article
Full-text available
Social trust has long attracted the interest of researchers across different disciplines. Most of previous studies rely on single-country data and consider only one dimension of social trust at a time (e.g., trust in science, the media or political institutions). This research extends a framework developed by the Global Trust Inventory (GTI) by discussing several dimensions of social trust, while simultaneously analyzing how trust in institutions varies across societies. Drawing on an online panel survey collected in 22 countries (N = 22,033), we examine cross-country differences in social trust—including government trust, trust in governing bodies, security, and knowledge producers. Additionally, this paper fills a gap in current literature by including a measure of trust in the media. Findings are discussed in the context of comparing emerging and developed countries based on the Human Development Index.
Article
Full-text available
Social media and news use arguably contribute to the prevalence of contentious politics because individuals may express dissent through their social networks as they consume news. This study seeks to test whether individuals might be more open to political persuasion in this context, especially if they are exposed to political disagreement or discuss politics in a civil manner. Relying on survey data from the UK, results based on a moderated moderation model show that (a) social media news use predicts political persuasion on social media (direct effects) and, (b) discussion disagreement and civil reasoning moderate this relationship in two-way and three-way interactions.
Article
Full-text available
In the context of the United States, research shows a positive relationship between network heterogeneity and political expression on social media at the individual level. This study builds on that research, relying on multilevel analysis that (1) leverages a twenty-country comparative survey and (2) includes country-level data on freedom of expression. Results show a positive relationship between network heterogeneity and political expression on social media across countries, but that relationship is stronger where freedom of expression is more limited. Keywords social media, political expression, network heterogeneity, freedom of expression index, democracy, political communication This paper examines how a country's freedom of expression shapes "risky" political expression on social media. Political expression has flourished on social media (Halpern and Gibbs 2013; Vaccari et al. 2015), and individuals who express their political views in these environments enjoy an increasingly substantial role, alongside political and media elites, in shaping public narratives about political issues in Western societies (Chadwick 2017; K. Thorson and Wells 2015). Recent research in the
Article
Full-text available
A long tradition of research focuses on conversation as a key catalyst for community integration and a focal mediator of media influence on participation. Changes in media systems, political environments, and electoral campaigning demand that these influences, and the communication mediation model, be revised to account for the growing convergence of media and conversation, heightened partisan polarization, and deepening social contentiousness in media politics. We propose a revised communication mediation model that continues to emphasize the centrality of face-to-face and online talk in democratic life, while considering how mediational and self-reflective processes that encourage civic engagement and campaign participation might also erode institutional legitimacy, foster distrust and partisan divergence, disrupting democratic functioning as a consequence of a new communication ecology.
Book
Politically active individuals and organizations make huge investments of time, energy, and money to influence everything from election outcomes to congressional subcommittee hearings to local school politics, while other groups and individual citizens seem woefully underrepresented in our political system. This book is a comprehensive and systematic examination of political voice in America, and its findings are sobering. The book looks at the political participation of individual citizens alongside the political advocacy of thousands of organized interests—membership associations such as unions, professional associations, trade associations, and citizens groups, as well as organizations like corporations, hospitals, and universities. Drawing on numerous in-depth surveys of members of the public as well as the largest database of interest organizations ever created—representing more than 35,000 organizations over a 25-year period—this book conclusively demonstrates that American democracy is marred by deeply ingrained and persistent class-based political inequality. The well-educated and affluent are active in many ways to make their voices heard, while the less advantaged are not. This book reveals how the political voices of organized interests are even less representative than those of individuals, how political advantage is handed down across generations, how recruitment to political activity perpetuates and exaggerates existing biases, how political voice on the Internet replicates these inequalities—and more. In a true democracy, the preferences and needs of all citizens deserve equal consideration. Yet equal consideration is only possible with equal citizen voice. This book reveals how far we really are from the democratic ideal and how hard it would be to attain it.
Article
La confianza social ha atraído desde hace mucho tiempo el interés de los investigadores de diferentes disciplinas. La mayoría de los estudios sobre esta materia se han basado en datos de un solo país, y han considerado una sola dimensión de la confianza social cada vez (por ejemplo, la confianza en la ciencia, en los medios de comunicación o en las instituciones políticas). Esta investigación desarrolla el modelo propuesto por el Inventario Global de la Confianza (Global Trust Inventory, GTI) al analizar simultáneamente varias dimensiones de la confianza social, además de examinar cómo la confianza en las instituciones varía en las distintas sociedades. A partir de una encuesta online de panel distribuida en 22 países (N = 22.033), estudiamos las diferencias en la confianza social entre sociedades, incluyendo en nuestro análisis la confianza en el Gobierno, en los organismos gubernamentales, en las instituciones nacionales de seguridad y en los productores de conocimiento científico. Además, el estudio contribuye a una operacionalizacíon más amplia del concepto de confianza social al introducir una medida de confianza en los medios de comunicación. Los resultados se discuten a partir de la comparación entre países emergentes y desarrollados, de acuerdo con su Índice de Desarrollo Humano (IDH). [EN] Social trust has long attracted the interest of researchers across different disciplines. Most of previous studies rely on single-country data and consider only one dimension of social trust at a time (e.g., trust in science, the media or political institutions). This research extends a framework developed by the Global Trust Inventory (GTI) by discussing several dimensions of social trust, while simultaneously analyzing how trust in institutions varies across societies. Drawing on an online panel survey collected in 22 countries (N = 22,033), we examine cross-country differences in social trust—including government trust, trust in governing bodies, security, and knowledge producers. Additionally, this paper fills a gap in current literature by including a measure of trust in the media. Findings are discussed in the context of comparing emerging and developed countries based on the Human Development Index.
Article
This study advances a theoretical model centered on collective and internal efficacy to explain the separate pathways through which political sharing on Facebook and Twitter may influence individuals to engage in political activities. We test the model with data from a 2-wave panel survey conducted with an adult population in 2013 in Chile. We found that frequent usage of Facebook and Twitter for sharing political information is conducive to higher levels of participation through different efficacy measures. Facebook has a significant effect on collective—not internal—efficacy, whereas Twitter's effect is on internal—not collective—efficacy. Results are discussed in light of the diverse affordances and strengths of network ties of Facebook and Twitter.