“Yes, I Saw It –But Didn’t Read It …” A Cross-Country
Study, Exploring Relationships between Incidental News
Exposure and News Use across Platforms
, Brigitte Huber
and Homero Gil de Z
Smith School of Enterprise and the Environment, School of Geography and the Environment,
University of Oxford, Oxford, UK;
Department of Communication, University of Vienna, Vienna,
Political Science Department, University of Salamanca, Salamanca, Spain;
& Media Studies, Pennsylvania State University, State College, PA, USA;
on y Letras, Universidad Diego Portales, Santiago, Chile
Today, people are increasingly exposed to news on various
channels without actively seeking it. However, less is known
about the link between the so-called incidental news exposure
(INE) and actual news consumption. Using a two-wave panel data
set from 18 countries around the world, we study the so-far
under researched relation between INE and news consumption
across various platforms over time. In doing so, we control for
key micro-level variables such as news use, political interest and
trust in media as well as macro-level variables, including internet
connectivity, GDP, press freedom and literacy rate. The analyses
yield an optimistic picture, showing that INE plays a bridging
function across countries, leading to actual news consumption on
traditional, online and social media platforms. However, trust in
news and political interest do not seem to play key moderating
roles. Instead, individual analyses per country imply that the link
between INE and actual news use is more apparent for online
and social media news platforms, and particularly in countries
where general social media usage has been reported to be
considerably high (e.g. Brazil, Philippines, Taiwan, UK and USA).
Incidental news exposure;
news use; social media;
trust in news; political
“Yes, I saw it –but didn’t read it …” A cross-country study, exploring relationships
between incidental news exposure and news use across platforms in 19 countries
In today’s high-choice media environment (Aalberg, Blekesaune, and Elvestad 2013;
Prior 2007), people are constantly exposed to news—be it intentionally or not
(Hermida 2016). While checking one’s social media feeds, surfing on the net, watching
TV or simply listening to the radio; news exposure has become ubiquitous in people’s
digital lives (Hermida 2016; Sheller 2015). While in the past, consuming news has
involved a conscious decision, such as buying the daily newspaper, tuning in to the
radio or TV program, or subscribing to a news magazine; today, news is constantly
CONTACT Nadine Strauß email@example.com
ß2020 Informa UK Limited, trading as Taylor & Francis Group
2020, VOL. 8, NO. 9, 1181–1205
available for us and is an accompanying phenomenon in our everyday activities, both
online and offline.
More recently, data has shown that people around the world are increasingly exposed
to news first through social media (Newman et al. 2019). In America, for example, social
media has surpassed print news as people’s platform choice to get news for the first time
(Shearer 2018). In both the UK and the United States, the majority of people under the
age of 35 use social media to pick up news first (Newman et al. 2019). Even stronger
trends have been observed around the world. 66% of Brazilians use social media as a
source of news and the same is true for 45% of people in the UK; in comparison, 31% of
the respondents in Germany stated they seek news via social media (Newman et al. 2019).
However, the question that remains is: Are people who get incidentally exposed to snip-
pets of news (e.g. via social media) also more likely to actively read up on news? If so, to
what extent does incidental news exposure lead to active news consumption on various
platforms (either traditional, online, or social media)?
It can be argued that the plentitude of news snacking (Molyneux 2018) and topics that
people encounter in their everyday lives might lead people to consume more in-depth
information by actively seeking traditional, online or social media news. However, it can
also be contended that the sheer amount of news and information that is nowadays avail-
able across various platforms might lead people to feel “overloaded”(Nordensen 2008),
even resulting in a lack of active news consumption (e.g. Gil de Z
evol-Abreu 2017). Hence, whether serendipitous exposure actually motivates people to
read up on the news snippets might very much depends on the extent to which people
are interested in news about politics and public affairs and whether they trust the news
media. More importantly, this link might also be contingent on various cultural, political,
media and structural differences across countries.
Hence, the purpose of this study is to identify whether there is a relationship between
incidental news exposure (INE) and news consumption across various platforms; and, fur-
thermore, to what extent this relation depends on individuals’interest in politics and trust
in news. To find answers to these questions, we employ a cross-country approach to study
whether the relationship between INE and active news consumption varies across societies.
To do so, we use data from a two-wave survey among 18 countries, controlling for macro-
variables such as GDP, press freedom, internet connectivity and literacy rate. Whereas pre-
vious research has focused on democratic outcomes variables as a result of INE (e.g. polar-
ization: Heiss and Matthes 2019; participation: Valeriani and Vaccari 2016;attention/
knowledge: Feezell 2018;Oeldorf-Hirsch2018), our findings contribute to a better under-
standing of the effects of INE on active news use behavior—a decisive factor for overall
democratic behavior (Habermas 1989;Schudson1998).
Incidental News Exposure, Democratic Behavior and Its Missing Link
Societal and academic interest in incidental news exposure (INE) has vigorously picked
up since the beginning of the 21
century and the proliferation of the World Wide
Web (Tewksbury, Weaver, and Maddex 2001) and social media around the world (e.g.
Boczkowski, Mitchelstein, and Matassi 2018; Feezell 2018; Fletcher and Nielsen 2018;
Karnowski et al. 2017). With their seminal study, Tewksbury, Weaver, and Maddex
1182 N. STRAUß ET AL.
(2001) did not only define what scholars in communication science today broadly
understand under INE, namely “that people encounter current affairs information
when they had not been actively seeking it”(534), they have also shown that INE is
positively related with awareness to current affairs information (see also Feezell 2018).
However, awareness to particular issues merely means “having heard/seen”informa-
tion that comes by through various channels. But the more pressing question is
whether citizens are actively reading news due to INE, thereby becoming better
equipped to actively participate in democracies.
A range of studies has shown in the past that INE is, in fact, positively related with
political participation online and offline (Kim, Chen, and Gil de Z
niga 2013; Valeriani
and Vaccari 2016). Although these studies are indicative in explaining how today’s
fast-paced and pervasive news media environment might lead to better informed and
more politically engaged citizens, the implications of the results are limited. First,
some of the studies fail to account for causal effects (e.g. Valeriani and Vaccari 2016;
Kim, Chen, and Gil de Z
niga 2013), are only focused on the U.S. (Kim, Chen, and Gil
niga 2013) or student samples (e.g. Feezell 2018) and, more importantly, fail to
consider the assumed timely mechanism between INE and political behavior outcomes
and its crucial link, namely actual news use.
Despite the fact that some of the studies (Valeriani and Vaccari 2016; Kim, Chen,
and Gil de Z
niga 2013) control for news use in the regression analyses, the cross-sec-
tional datasets as well as the experiments (e.g. Feezell 2018) do not allow to follow
the logical argumentations made by the authors of the studies. According to the
scholars, INE leads to more information acquisition, and in turn, to increased levels of
political participation or political knowledge gain. However, most of the studies only
test the direct and indirect positive relationship between INE and diverse political
outcomes (e.g. participation or knowledge), but information consumption as a result
of INE has not been explicitly measured or tested so far (see for an exception: Fletcher
and Nielsen 2018). Hence, the purpose of this article is to shed light on whether the
assumed primary causal relationship between INE and news consumption exists.
Incidental News Exposure and Active News Use
With the emergence of possibilities to consume news online and via social media,
the way people perceive the availability of news and the need of staying informed
has changed. While Oeldorf-Hirsch (2018) warns that “individuals may be moving to
an increasingly passive exposure to information”(226), Gil de Z
niga and colleagues
(2017) speak of the so-called “news-finds-me”(NFM) perception, leading people to
believe that they do not have to actively seek the news anymore to stay informed.
Based on a U.S. panel-survey, Gil de Z
niga, Weeks, and Ard
conclude that people who score high on the NFM perception are less likely to use
traditional news sources and become less knowledgeable about politics over time.
Other researchers have identified a similar perception as “news fatigue”(e.g.
Nordensen 2008). It entails a negative attitude towards news that arises from a feeling
of incapability to deal with the ubiquity of news, eventually resulting in news
avoidance (Nordensen 2008). Similarly, Oeldorf-Hirsch (2018) infers from a structural
DIGITAL JOURNALISM 1183
model based on an online survey in the U.S. that INE might lead to engagement and
subsequently to elaboration about the content, but that elaboration does not lead to
any further knowledge gains.
Another stream of research is more hopeful and considers the proliferation of options
to retrieve information about current events through social media and online tools as
beneficial, for example, by enhancing democratic behavior (e.g. Gil De Z
and Rojas 2009;Shah,Kwak,andHolbert2001). Feezell (2018), for instance, found that
people who are exposed to political information on Facebook evince an agenda-setting
effect, meaning they report a higher perceived importance of policy issues that have been
shared on Facebook compared to people who have not been shown political information
on Facebook. More recently, a study provided evidence that mobile push notifications—a
common way of how individuals nowadays get incidentally exposed to news—does not
only increase the self-reported use of news apps, it can also lead to learning about the
news in some instances (Stroud, Peacock, and Curry 2019). Similarly, Lee and Kim (2017)
have shown by means of an experiment that INE has a positive and significant effect on
the recognition and short-term recall of information in news stories. In fact, research inves-
tigating “passive learning”proposes that learning can occur without being actually moti-
vatedtodoso(KrugmanandHartley1970) and that people who are accidentally exposed
to information might learn as a “side product.”
Hence, when converging these two streams of research it becomes clear that there is
still not yet a clear scholarly agreement on whether INE leads to active news use and/or to
knowledge gains. One theory that might better explain the seemingly causal link between
INE and knowledge gain might be Eveland’s(2001) cognitive mediation model that also
controls for actual information consumption (cf. news use). However, only recently,
Oeldorf-Hirsch (2018) has tested the cognitive mediation model for INE on social media
(Facebook and Twitter), showing that INE does neither lead to elaboration of news nor to
knowledge gains, but mostly to engagement with news on the respective social media
outlets (e.g. liking, commenting, sharing). By contrast, Lee and Kim (2017) have provided
evidence that the effect of INE on information recall of news is mediated by exposure to
information in the news (measured by the time spent for reading the news). Yet the study
by Lee and Kim (2017) is merely based on US data and only examines online news.
Thus, we aim at closing this research gap by studying the direct relationships
between INE and actual news consumption across various platforms and across 18
countries. By showing that the first causal relationship of the cognitive mediation
model persists across platforms (traditional, online, social media), future studies will be
in a better position to infer whether individuals actively engage with news due to INE,
and whether cognitive processing and thus knowledge gain could potentially take
place as a consequence. Accordingly, our first research question reads:(RQ1) How is
incidental news exposure related to a) traditional news media use, b) online news media
use, and c) social media news use?
The Moderating Role of Political Interest and Trust in News
Previous research shows that news use is tremendously influenced by, at least, two
key variables, namely political interest (Boulianne 2011; Str€
ack and Shehata 2010;
1184 N. STRAUß ET AL.
ack, Djerf-Pierre, and Shehata 2013) and trust in news (Ard
evol-Abreu and Gil
niga 2017; Tsfati and Cappella 2005; Tsfati 2010; Williams 2012). Particularly with
the emergence of online and social media, political interest has become a decisive fac-
tor in explaining why people seek and learn from news (Prior 2007). According to
Prior (2007), the transition from a low-choice to a high-choice media environment
offers an abundance of entertainment as well as information on politics and public
affairs. While less politically interested people are said to be more likely to seek enter-
tainment content in this environment, highly interested individuals might be more
likely to make use of the ubiquity of political information, eventually leading to a pol-
itical knowledge gap in society. A recent study by Heiss and Matthes (2019) has
indeed found support that INE can lead to the reinforcement of the political participa-
tion gap between politically interested and those who are less interested in politics.
Besides political interest, Karnowski et al. (2017) point to content-dependent factors
(e.g. topical interest) that explain how exposure to news on social media might lead
to engagement (i.e. reading the news). As psychological research has shown, prior
knowledge leads individuals to focus more on information that is relevant to their
already obtained knowledge than information that seems to appear irrelevant (Kim
and Rehder 2011). Especially when it comes to incidental news encounters, recom-
mendations through algorithms, as well as peer recommendations play a crucial role
for news selection and exposure (Van Damme et al. 2020). In fact, Mummolo (2016)
has demonstrated that interest in the news topic can even level off the negative effect
of source reputation. In other words, even if Democrats are shown an article that
comes from a Republican source (e.g. Fox); if the topic is relevant to them, they will
choose to read the article anyways.
Similarly, but focusing on INE on Facebook, Karnowski et al. (2017) employed a
mobile forced experience sampling study and found topical interest leading to both,
the intention to read the news article shown on Facebook and the intention to look
up further information. More recently, using self-confrontation interviews, K€
(2019) reports that engagement decisions while being exposed to news on social
media are mostly guided by interest in the issue covered in the respective article.
om and Belfrage (2018) identified interest in news, next to habit of
using online news services and age, as the main influencing factors that explain social
media news use. Hence, given that we focus in this study on general news use
defined as information on politics and public affairs, we want to replicate the already
established direct relationship between political interest and news use, showing that it
is also valid across 18 countries and for different platforms. Therefore, we presume:
(H1) There is a positive relationship between political interest and a) traditional news
media use, b) online news media use, and c) social media news use.
What is more, we know from previous research that political interest interacts with
INE with regard to political participation and the agenda-setting effect. More specific-
ally, the findings by Valeriani and Vaccari (2016) have shown that the correlation
between INE on social media and political participation online is moderated by polit-
ical interest: those scoring low in political interest elicited a stronger correlation than
those being already highly interested in politics. The agenda-setting effect as reported
by Feezell (2018) has shown similar differences depending on participants’level of
DIGITAL JOURNALISM 1185
political interest. For those participants with low political interest, the ranking of per-
ceived importance of political issues shown during INE on Facebook was higher than
for those that have reported a higher interest in politics. Hence, INE might actually
level off the so-far reported differences regarding political behavior due to political
interest. Following this logic, we predict the following moderation effect: (H2) Political
interest moderates the effect of INE on a) traditional news media use, b) online news
media use, and c) social media news use so that people scoring high in political interest
evince a weaker correlation between INE and news use (traditional, online and social
media) than people scoring low in political interest.
Besides political interest, the level of trust people ascribe to media might strongly
influences the extent to which they actively seek and consume news across various
platforms. For example, based on 112 interview transcripts from news consumers in
the US Midwestern region, Pentina and Tarafdar (2014) found that news users—inde-
pendent of their political views, motivations or interests in news topics—rank the reli-
ability and trustworthiness of the news sources as highly relevant in their news
consumption processes on social media. Respondents narrated that they would care-
fully select the news sources they consume on social media based on the reputation
of the source (e.g. brand). However, the reports by the news consumers interviewed
by Pentina and Tarafdar (2014) might be affected by social desirability. Indeed, empir-
ical support for the relationship between trust in news media and news exposure is
limited (Tsfati and Cappella 2005). Studying four large sample data sets, Tsfati and
Cappella (2003) reported only moderate correlations between trust in mainstream
news and mainstream news consumption. Furthermore, reviewing previous research,
bivariate correlations between respondents’level of trust in news media institutions
and the amount of new consumption lies below .20, according to Tsfati and
Despite these pessimistic findings, we argue that trust in news can still be viewed
as an important factor in explaining news media consumption. In fact, Tsafati and
Cappella (2005) have shown themselves that mainstream media skepticism is nega-
tively related with news exposure, even after controlling for a range of demographic
and political variables. Furthermore, Tsfati (2010) has replicated these findings showing
that both overall news media skepticism and skepticism toward online news is nega-
tively related with time spent surfing news websites. Similarly, Ard
evol-Abreu and Gil
niga (2017) have shown more recently that while trust in traditional media is
related to traditional news use, trust in social and citizen media is positively connected
with social media news use. Thus, following this line of research, we want to test
whether the positive relationship between trust in news and news use also holds
across 18 countries and across different news media platforms: (H3)There is a positive
relationship between trust in news and a) traditional news media use, b) online news
media use, and c) social media news use.
Next, we argue that trust in news is a moderating factor in explaining the relation-
ship between INE and news use across platforms. Previous research has repeatedly
theorized that trust is as an important factor to be related with INE (e.g. Pentina and
Tarafdar 2014; Tandoc and Johnson 2016; Yadamsuren and Heinstr€
om 2011). While
Fletcher and Nielsen (2018) and K€
umpel (2019) only controlled for trust/evaluation in/
1186 N. STRAUß ET AL.
of news in their analyses, Goyanes (2020) has empirically shown that trust in social
networking sites is positively related with INE. However, the interaction between the
frequency of using social media for news and trust in social networking sites did not
yield a positive effect on INE. Building up on these first empirical findings, it is
assumed that the combination of trust in news and INE will lead to more active news
consumption. Thus, we hypothesize: (H4)Trust in news moderates the effect of INE on a)
traditional news media use, b) online news media use, and c) social media news use so
that people scoring high in trust in news evince a stronger correlation between INE and
news use (traditional, online and social media) than people scoring low in trust in news.
Structural, Economic, Media, and Educational Differences across Countries
Cross-country research on INE is generally scarce. We could identify only one study by
Valeriani and Vaccari (2016) that studied INE and its effects on political participation
online across Germany, Italy and the UK. However, comparative research is becoming
increasingly important in an ever-more globalized world (Artz and Kamalipour 2003).
This study attempts to fill this research gap (cf. Yadamsuren and Erdelez 2017) by ana-
lyzing the relationship between INE and news use across 18 countries, and controlling
for cross-country differences, such as internet connectivity, GDP, press freedom and liter-
acy rate. All four macro-variables are of interest as they cover four different aspects by
which the 18 countries differ in terms of digital infrastructure, economy, media land-
scape and education.
High internet access or internet connectivity, for example, is related with a higher
rate of people using the Internet. Hence, more internet use might also be related with
more information and news consumption across the 18 countries. Furthermore, previ-
ous research in communication science has repeatedly argued that comparative stud-
ies across various countries should control for the specific media landscape in which
the research is conducted (Hallin and Mancini 2004). Following previous research, we
have chosen press freedom to account for political and economic differences across
countries (Himelboim and Limor 2008). Regarding the economic strength, a study by
the Pew Research Center has recently found that people in richer countries are more
likely to consume news online (Mitchell et al. 2018). Following that, the level of GDP
might also represent an important control variable to account for cross-country differ-
ences regarding news use across different platforms. Eventually, macro-level indicators
such as research and development spending, literacy and secondary education have
been found to be strongly and significantly related with the use of the Internet as
well as new and old media use (Norris 1996). Consequently, the level of literacy rate
might explain the different levels of news use across platforms and countries.
Sample and Data
The current study is part of a larger international project “Digital Influence”, a collabor-
ation between the University of Vienna and the University of New Zealand. It uses
two-wave panel data from 18 countries worldwide. The countries were selected to
DIGITAL JOURNALISM 1187
represent a variety of political, economic and cultural contexts as well as different con-
tinents (Americas, Asia, Europe, Africa). Originally, 22 countries were surveyed, but
only respondents from 18 countries gave valid answers to the questions of interest in
this study. Data for Wave one was collected in September 2015, and for Wave 2 in
March/April 2016. To perform the translation of the items, scholars from each country
were involved. Afterwards, the surveys were translated using either back-translation
with a team approach (Thato, Hanna, and Rodcumdee 2005), or the committee
approach (Brislin 1980). The survey was distributed by Nielsen that used stratified
quota sampling techniques in order to create samples in each country with demo-
graphics similar to those provided in reports of official census agencies. For more
details on the survey and a demographic breakdown by country, see Gil de Z
and Liu (2017); for more information on breakdowns for the variables of interest in
this study, see Table 1.
Incidental News Exposure
We told respondents in the survey that people sometimes come across news and
information on current events, public issues, or politics when they may have been
using media for a purpose other than to get the news. Afterwards, we asked them to
indicate (1 ¼never, 7 ¼always) how often that happens to them a) while watching TV,
listening to the radio, or reading the newspaper, and b) while on social media or the
Internet (Spearman-Brown ¼.47, M¼4.54, SD ¼1.29).
Traditional News Use
Traditional news use was measured by asking respondents how often they get news
(1 ¼never, 7 ¼always) from a) TV (cable or local network news), b) newspapers (printed
version), and c) radio (W1 a¼.59, M¼4.56, SD ¼1.30; W2 a¼.60, M¼4.63, SD ¼1.32).
Online News Use
To measure online news use, individuals were asked to indicate how often (1 ¼never,
7¼always) they get news from a) online news websites, and b) citizen journalism sites (W1
Spearman-Brown ¼.30, M¼4.25, SD ¼1.39; W2 Spearman-Brown ¼.40, M¼4.20, SD ¼1.42).
Social Media News Use
Building on prior research (Valenzuela, Arriagada, and Scherman 2012), social media news
use was measured by asking respondents a) how frequently they get news from social media
(1 ¼never, 7 ¼always),and b) how frequently they use social media to get news about cur-
rent events from mainstream media (1 ¼never, 7 ¼all the time; W1 Spearman-Brown ¼.49,
M¼4.30, SD ¼1.63; W2 Spearman-Brown ¼.58, M¼3.89, SD ¼1.70).
To gauge political interest, respondents were asked a) how closely they pay attention
to information about what is going on in politics and public affairs (1 ¼not at all,
7¼very closely), and b) how interested they are in information about what is going
1188 N. STRAUß ET AL.
Table 1. Descriptivesof key variables per country.
Country Cases INE
M SD M SD M SD M SD M SD M SD M SD M SD M SD M SD M SD
Argentina 1,146 360 4.78 1.33 4.70 1.25 4.66 1.26 4.43 1.47 4.33 1.48 4.93 1.44 4.66 1.50 4.52 1.45 4.84 1.40 3.63 0.97 3.61 0.94
Brazil 1,086 353 5.23 1.19 4.78 1.32 4.75 1.25 5.22 1.38 5.13 1.37 5.32 1.25 5.15 1.37 5.12 1.36 5.50 1.18 3.62 1.23 3.78 1.25
China 1,004 387 4.98 0.95 4.24 1.20 4.35 1.18 5.15 1.12 5.29 1.08 5.04 1.10 5.12 1.15 4.60 1.32 4.79 1.31 3.67 1.22 3.79 1.24
Estonia 1,168 733 4.38 1.19 4.99 1.33 5.10 1.28 4.24 1.28 4.31 1.21 3.98 1.52 3.92 1.53 4.46 1.26 4.56 1.29 3.61 0.87 3.68 0.85
Germany 1,054 645 4.44 1.29 4.97 1.32 5.07 1.37 3.81 1.42 3.70 1.40 3.56 1.72 3.11 1.73 5.23 1.47 5.29 1.53 3.45 1.12 3.44 1.11
Indonesia 1,080 305 5.05 1.13 5.31 1.12 4.73 1.19 4.14 1.31 5.25 1.05 5.36 1.07 5.31 1.04 4.35 1.32 4.41 1.36 3.84 0.98 3.93 0.97
Italy 1,041 579 4.36 1.39 4.27 1.05 4.89 1.15 4.13 1.20 4.59 1.44 4.17 1.32 4.38 1.55 4.81 1.45 4.98 1.41 3.68 1.09 3.74 1.13
Japan 975 574 4.19 1.22 4.11 1.36 4.25 1.48 4.03 1.35 4.12 1.36 3.23 1.54 3.16 1.56 4.63 1.29 4.67 1.21 3.33 0.95 3.34 0.92
Korea 944 573 4.48 1.24 3.95 1.34 4.05 1.36 4.40 1.26 4.32 1.27 3.98 1.48 3.77 1.51 4.21 1.27 4.31 1.28 3.58 1.09 3.62 1.03
New Zealand 1,157 605 4.04 1.21 4.63 1.27 4.65 1.32 3.52 1.25 3.41 1.19 3.69 1.62 3.31 1.63 4.07 1.53 4.15 1.53 3.21 1.01 3.10 0.92
Philippines 1,064 153 4.97 1.22 4.69 1.18 4.87 1.17 4.80 1.19 4.89 1.02 5.56 1.01 5.48 1.05 4.50 1.20 4.79 1.16 4.05 1.00 4.13 1.04
Poland 1,060 628 4.51 1.16 4.80 1.17 4.80 1.15 4.56 1.13 4.51 1.12 4.02 1.44 3.84 1.50 4.37 1.42 4.56 1.40 3.40 1.11 3.53 1.09
Russia 1,145 551 4.6 1.19 4.42 1.29 4.41 1.20 4.63 1.24 4.54 1.23 4.14 1.46 3.93 1.45 4.52 1.37 4.71 1.29 3.36 1.12 3.31 1.13
Spain 1,020 302 4.31 1.38 4.73 1.25 4.82 1.32 4.26 1.45 3.99 1.50 4.44 1.55 4.01 1.69 4.61 1.40 4.96 1.26 3.64 0.98 3.60 0.93
Taiwan 1,008 426 4.73 1.07 3.94 1.07 4.01 1.12 4.20 1.17 4.19 1.16 4.54 1.19 4.17 1.24 3.79 1.42 3.84 1.43 2.64 1.02 2.49 0.99
Turkey 961 331 4.99 1.23 4.44 1.15 4.80 1.12 4.45 1.16 4.65 1.19 5.19 1.27 5.06 1.22 5.19 1.38 5.38 1.22 3.25 1.16 3.30 1.14
UK 1,064 649 3.81 1.41 4.67 1.29 4.76 1.33 3.28 1.39 3.22 1.34 3.04 1.79 2.74 1.74 4.39 1.66 4.62 1.52 3.15 1.13 3.03 1.14
USA 1,161 489 4.04 1.31 4.25 1.38 4.25 1.39 3.37 1.36 3.08 1.32 3.39 1.77 2.82 1.67 4.40 1.62 4.70 1.64 3.04 1.09 2.89 1.08
ALL 19,138 8,643 4.55 1.29 4.56 1.30 4.63 1.32 4.25 1.39 4.20 1.42 4.30 1.63 3.89 1.70 4.54 1.45 4.69 1.44 3.45 1.11 3.42 1.11
DIGITAL JOURNALISM 1189
on in politics and public affairs (1 ¼not at all, 7 ¼a great deal; Spearman-Brown ¼.87,
M¼4.54, SD ¼1.45).
Trust in News
We asked individuals how much they trust news from various sources (1 ¼do no trust
at all, 7 ¼trust completely): a) news from mainstream news media (e.g. newspapers,
TV), b) news from alternative news media (e.g. blogs, citizen journalism), and c) news
from social media(a¼.77, M¼3.45, SD ¼1.11).
To account for individual differences across countries, we controlled for sociodemo-
graphic characteristics. We included age (M¼41.46, SD ¼14.80), gender (51.04%
female), education (1 ¼none, 7 ¼graduate school or higher; Median ¼4, some college),
income (range of scale: 1 ¼0-10 percentile; 5 ¼91-100 percentile; M¼2.94 (11-30 per-
centile), SD ¼1.09), and race (85.7% white) in our models.
To control for internet connectivity across the 18 countries investigated, we included
the measures percentage of internet users per country (M¼73.40, SD ¼19.14) and per-
centage of broadband access per country (M¼76.47, SD ¼23.12) and averaged the two
items for our analyses (Spearman-Brown ¼.75, M¼73.77, SD ¼19.95). Data about inter-
net connectivity were retrieved from www.webworldwide.io.
Press freedom was used from Freedom House where high levels of press freedom
means less freedom (M¼38.91, SD ¼21.11).
GPD per capita was retrieved from the World Bank, except for Taiwan which was only
available from the IMF (M¼22.73, SD ¼15.38).
Data to measure literacy rate were collected from the website www.ourworldindata.
org (M¼97.83, SD ¼1.90). However, given that the literacy rates in some countries are
very high, some countries only have the most recent figures (e.g. Germany, New
Zealand, Taiwan, UK, USA from 2003; South Korea from 2008; Japan from 2002).
First, we investigated the zero-order correlations to see whether there are reasonable
associations between our key variables (see Table 2 and 3). To test the hypotheses, we
pooled our data and applied multilevel modeling to account for country differences.
Before we estimated the full multilevel model, we first took a look at the null models
1190 N. STRAUß ET AL.
(no covariates included) to substantiate our decision to employ linear mixed models
for our analyses. The results indicated that the null hypotheses of no country differences
in the dependent variable (Traditional News Use
) were rejected. More specifically, the interclass correlation coefficient for
Traditional News Use is .08 (W1) and .06 (W2), for Online News Use .15 (W1) and .20 (W2),
for Social Media News Use .23 (W1) and .25 (W2) respectively. Hence, these results indicate
that a critical proportion of the variance of news use (traditional, online, social media) in
the population is explained by the grouping structure (cf. countries) which in turn substan-
tiates our choice of using multilevel analyses (Hox 2002).
Given the two-wave panel data, we did not only calculate the multilevel models for
the cross-sectional data, but also estimated the models with the lagged dependent
variables and the autoregressive term, the so-called static-score model (Finkel, 1995).
First, we estimated the cross-sectional model where the independent and dependent
variables stem from Wave 1. Second, for the lagged model, we again used the inde-
pendent variables from Wave 1 but regressed those on the dependent variable from
Wave 2. Third, the autoregressive model is equal to the lagged model but includes
another autoregressive term; meaning, we also included the dependent variable from
Wave 1 as a control in our model. Following Finkel (1995), the latter model can be
used to estimate the causal effect of Xon Yif the time between the two waves is not
too long and a “synchronous”or “cotemporal”effect is assumed (13). This is clearly
the case with our panel dataset in which only six months lie in between the two
waves. Furthermore, the variables under investigation (e.g. news use, INE, political
interest) are also rather static over time and thus justify the application of the static-
score model to derive causal claims.
While the multi-level models account for country-level differences by included
macro-level factors (e.g. GDP, literacy rate), we also estimated individual OLS models
for each country, controlling for the same variables as in the overall multi-level mod-
els. In addition, to rule out reversed relationships (i.e. news media use affecting INE),
Table 2. Zero-order correlations among all key variables wave 1.
2. Political Interest
3. Trust in News
4. Traditional News Use
.23 .31 .20
5. Online News Use
.37 .28 .24 .28
6. Social Media News Use
.50 .16 .32 .20 .55
Table 3. Zero-order correlations among all key variables wave 2 (except for INE).
2. Political Interest
3. Trust in News
4. Traditional News Use
.17 .37 .22
5. Online News Use
.37 .29 .30 .29
6. Social Media News Use
.44 .19 .37 .19 .61
DIGITAL JOURNALISM 1191
we ran a separate multi-level model that treated INE as the dependent variable.
However, given that INE was only measured in Wave 1, we could only run this model
for the cross-sectional data.
The first research question inquired how INE is related with a) traditional news media use,
b) online news media use, and c) social media news use. Table 4 shows that INE is signifi-
cantly related with news use across all news media platforms (traditional, online, social
media) as well as for the cross-sectional, lagged and autoregressive models. All direct
effects from INE to the respective news platforms are highly significant and positive (p
<.001). Checking for reversed effects (Table 5), the results indicate that social media news
is strongly correlated with INE (B¼0.34, p<0.001), whereas traditional news use (B¼0.13,
p<0.001) and online news to a lesser extent (B¼0.06, p<0.001).
Hypothesis one was formulated to replicate previous findings that showed a posi-
tive relationship between political interest and news use. In line with past studies, the
direct positive effects are highly significant across all three news media use platforms
(traditional, online, social media) and even persist over time (see Table 4). Thus, we
find support for H1. Hypothesis 2 presumed that there is an interaction effect between
INE and political interest on news use across platforms. The results reject H2, only
showing significant effects for traditional news use in the cross-sectional (B¼–0.02,
p<0.01) and the lagged model (B¼0.04, p<0.01) and for social media news use in
the cross-sectional model (B¼0.02, p<0.01). Moreover, in contrast to previous studies,
the interaction graphs (see Figures 1–3) show that the higher INE, the higher the
reported news use and that this relationship is stronger for individuals who score high
on political interest.
Similar to H1, Hypothesis three was formulated to substantiate the assumed direct
relationship between trust in news and news media use across platforms. The results
show that the positive direct effects not only persist across all platforms (traditional,
online, social media), but also remain highly significant for lagged and autoregressive
models (see Table 4). Thus, H3 is supported. Furthermore, in Hypothesis four we
wanted to test whether trust in news interacts with INE in explaining traditional,
online and social media news use. However, we only find limited support for this
hypothesis. The results merely point to a positive interaction effect in the cross-sec-
tional models for online news use (B¼0.02, p<0.001) and social media news use
(B¼0.03,p<0.001). As hypothesized, we find the relationship between INE and
social media news use to be stronger for those who trust the news more, compared
to those who are rather skeptical towards news (see Figures 4 and 5).
Regarding the macro-level control variables, the analyses show that none of the
indicators has a strong effect on the outcome variables. If at all, there are signs that a
higher GDP per capita leads to less online news use and higher internet connectivity
to lower social media news use (see Table 4). When exploring the relationships
between INE and news use across various platforms for each country individually, it
appears that particularly social media news use is affected by INE and strongest in
Brazil, Philippines, Taiwan, UK and the U.S (Table 6).
1192 N. STRAUß ET AL.
Table 4. Multilevel model predicting news use on various platforms across 18 countries.
Traditional News Use Online News Use Social Media News Use
B(SE) B(SE) B(SE) B(SE) B(SE) B(SE) B(SE) B(SE) B(SE)
Intercept –.00 (.06) –.07 (.05) –.11 (.04) .01 (.06) .09 (.04) .07 (.04) .00 (.04) .14 (.04) .17 (.02)
Trad. News Use
––.69 (.01) ––––––
Online News Use
–––––.54 (.01) –––
SM News Use
Age .30 (.01) .30 (.01) .10 (.01) –.10 (.01) –.08 (.01) –.03 (.01) –.13 (.01) –.11 (.01) –.05 (.01)
Gender (0 ¼m;1 ¼f) –.01 (.01) .01 (.01) .01 (.01) –.04 (.01) –.04 (.01) –.02 (.01) .05 (.01) .06 (.01) .03 (.01)
Education .02 (.01) .02 (.01) .02 (.01) .06 (.01) .07 (.01) .04 (.01) –.03 (.01) –.01 (.01) .01 (.01)
Income .07 (.01) .12 (.01) .05 (.01) .03 (.01) .05 (.01) .04 (.01) –.02(.01) .00 (.01) .01 (.01)
Race (0 ¼other;1
.03 (.01) .02 (.01) .00 (.01) –.01 (.01) .00 (.01) .01 (.01) –.01 (.01) .01 (.01) .01 (.01)
News Use Variables
Traditional News Use –––.02 (.01) .19 (.01) .07 (.01) –.01 (.01) .04 (.01) .03 (.01) –.00 (.01)
Online News Use .24 (.01) .11 (.01) .01 (.01) .37 (.01) .30 (.01) .06 (.01)
Social Media News Use .05 (.01) .06 (.01) .42 (.01) .32 (.01) .10 (.01)
News Use Antecedents .07 (.01) .02 (.01) .03 (.01) .03 (.01)
Political Interest .13 (.01) .18 (.01) .04 (.01) .12 (.01) .16 (.01) .09 (.01) .13 (.01) .14 (.01) .06 (.01)
Trust in News .09 (.01) .11 (.01) .03 (.01) .00 (.01) .04 (.01) .03 (.01) .23 (.01) .18 (.01) .05 (.01)
INE .13 (.01) .12 (.01) .05 (.01) .09 (.01) .05 (.01)
Internet Connectivity –.13 (.10) –.06 (.07) –.00 (.06) .16 (.09) –.02 (.07) –.06 (.06) –.13 (.06) –.14(.06) –.08(.03)
Freedom of Press –.15 (.09) –.12 (.06) –.02 (.05) .08 (.08) .01 (.06) –.02 (.05) –.03 (.06) –.02 (.05) –.02 (.02)
GDP –.08 (.10) –.07 (.07) .00 (.06) –.14 (.09) –.17(.07) –.08 (.06) –.04 (.07) –.09 (.06) –.05 (.03)
LiteracyRate .05 (.08) –.05 (.06) .03 (.05) –.01 (.08) –.01 (.06) –.01 (.05) –.09 (.05) –.08 (.05) –.03 (.02)
INEPolitical Interest –.02 (.01) –.04 (.01) –.01 (.01) –.01 (.01) .00 (.01) .01 (.01) .02 (.01) .02 (.01) .00 (.01)
INETrust in News .00 (.01) .00 (.01) .01 (.01) .02 (.01) .02 (.01) .01 (.01) –.03 (.01) –.01 (.01) .00 (.01)
DIGITAL JOURNALISM 1193
Table 4. Continued.
Traditional News Use Online News Use Social Media News Use
B(SE) B(SE) B(SE) B(SE) B(SE) B(SE) B(SE) B(SE) B(SE)
.68 .71 .38 .54 .59 .43 .47 .56 .38
.07 .03 .03 .06 .03 .02 .03 .02 .00
.10 .05 .06 .10 .05 .05 .06 .04 .01
Observations 14,139 6,496 6,379 14,139 6,509 6,442 14,139 6,317 6,237
.27 .24 .59 .41 .38 .54 .51 .43 .62
.34 .28 .62 .47 .41 .57 .54 .45 .63
Notes. Cell entries are unstandardized coefficients for linear mixed models for 18 countries; standard errors in parentheses;
1194 N. STRAUß ET AL.
Although news use has become ubiquitous within peoples’daily lives (Purcell et al.
2010), and citizens have more ways today to stay abreast of news than ever before
(e.g. smartphone, social media), it is still not clear whether incidental exposure to
news is followed by a more active and meaningful news consumption. Thus, the
purpose of this study was to investigate whether INE leads to actual news use across
various platforms (traditional, online, social media) and whether this holds true across
various societies around the world, controlling for individual factors such as news use,
political interest and trust in news as well as macro-level indicators such as internet
connectivity, GDP, press freedom and literacy rate.
Based on a two-wave panel survey among 18 countries, we employed multi-level
analyses to control for individual and country-level influences. Contrary to pessimistic
views (e.g. Prior 2007), our findings suggest a positive outlook regarding the level of
news consumption in today’s pervasive, multi-media news environment. First of all,
our results underpin previous findings that have shown that political interest is posi-
tively related with news consumption (Boulianne 2011; Str€
ack and Shehata 2010;
ack, Djerf-Pierre, and Shehata 2013). Results suggest that political interest leads
to more news use across all platforms (traditional, online, social media) and even over
time. Moreover, our study also lends support to the notion that trust in news is overall
Table 5. Multilevel model predicting incidental news exposure
across 18 countries.
Incidental News Exposure (INE) B(SE)
Intercept .01 (.03)
Age –.10 (.01)
Gender (0 ¼m;1 ¼f) .06 (.01)
Education –.02 (.01)
Income –.00 (.01)
Race (0 ¼other;1 ¼white) –.01 (.01)
News Use Antecedents .10 (0.01)
Political Interest .10 (0.01)
Trust in News
News Use Variables .06 (0.01)
Traditional News Use .34 (0.01)
Online News Use
Social Media News Use
Internet Connectivity .07 (0.04)
Freedom of Press .04 (0.03)
GDP –.03 (0.04)
Literacy Rate –.08(0.03)
Note. Cell entries are unstandardized coefficients for linear mixed models for 18 countries; standard errors in
DIGITAL JOURNALISM 1195
a strong positive factor for news consumption. While previous research has casted
doubt on this relationship (Tsfati and Cappella 2005; Tsfati 2010), this article shows
that the relationship between trust in news and news use seems to hold for all plat-
forms (traditional, online, social media) and, above all, that it seems to be causal for
different countries around the world.
More crucially, the analyses have provided compelling evidence that INE leads to
actual news consumption of traditional, online and social media. The significant find-
ings for the autoregressive models across all media platforms suggest a strong tem-
poral relationship. While previous research has mostly investigated the outcomes of
INE on democratic variables (e.g. polarization: Heiss and Matthes 2019; participation:
Valeriani and Vaccari 2016; attention/knowledge: Feezell 2018; Oeldorf-Hirsch 2018),
we have demonstrated in this study that INE plays a decisive role in explaining active
news use behavior in today’s ubiquitous news environment. It is in this vein that our
study showcases a first causal link within the cognitive mediation model regarding INE
(Oeldorf-Hirsch 2018; Eveland 2001), meaning that INE can be considered a driver for
Figure 1. Interaction effect of INEPolitical interest on TraditionalNewsUse
1196 N. STRAUß ET AL.
further news use. It is now up to future studies to demonstrate how subsequent consump-
tion of news, resulting from INE, leads to elaborative processing and, in turn, to more
active democratic behavior (e.g. political knowledge, participation, voting). Based on our
findings, follow-up research is thus advised to control for actual news use when investigat-
ing INE and any successive outcome, such as political behavioral variables.
However, contrary to the direct relationships, the interaction effects of political
interest and trust in news with INE on news use across platforms were less clear. The
interaction effects that we found for political interest stand in contrast to previous
findings that showed that INE can offset the effect of political interest on political par-
ticipation online or on the agenda-setting effect respectively (Valeriani and Vaccari
2016; Feezell 2018). Instead, our findings suggest that political interest is still a crucial
factor for explaining news consumption. In other words, for individuals who score
high on political interest the effect of INE on actual news consumption (e.g. traditional
Figure 2. Interaction effect of INEPolitical interest on TraditionalNewsUse
DIGITAL JOURNALISM 1197
news) is stronger than for individuals who are less interested in politics. Thus, in a
news and media environment that is highly characterized by snippets of news
(Molyneux 2018), it becomes increasingly relevant that individuals have developed a
certain level of political interest in order to actually seek news on traditional news
media (cf. Prior 2007). This finding also supports previous research that has shown
that political interest has become a determining factor in explaining informational TV
use in a high choice media environment (Hopmann et al. 2016; Prior 2007) or political
participation gaps (Heiss and Matthes 2019)
Regarding the interaction effects for trust in news, the findings were also contrary
to our expectations. Although we found that the relationship between INE and news
consumption across all platforms (traditional, online, social media) is stronger for those
who trust the news, the effects were not persistent and do not allow for any causal
inferences. Hence, we reason that trust in news might still play a role when it comes
to INE and subsequent news use, but that there might be further additional individual
cognitive, habitual and content-dependent factors at stake, such as interest in the
Figure 3. Interaction effect of INEPolitical interest on SocialMediaNewsUse
1198 N. STRAUß ET AL.
respective topics shown during INE (cf. K€
umpel 2019), that affect active news
In fact, although this study controlled for individual-level measures and macro-level
measures, it has become apparent that the country-level factors were hardly influential
in explaining news media consumption across the 18 countries investigated. Initially,
we expected that higher internet connectivity, GPD, freedom of press and literacy rate
would be related with higher news use. However, the results have shown that the
reverse is true for GDP and internet connectivity: The higher the GDP of countries and
the higher internet connectivity, the less people consume news online or on social
media. Yet given that these results were not consistent across news platforms, we
refrain from drawing any firm conclusions.
In contrast, the individual analyses for each country have shown that the strongest
relationship between INE and news use persists for social media news use, and par-
ticularly for Brazil, Philippines, Taiwan, UK and the United States. Recent numbers of
active social media users per country show that all countries have a considerably high
penetration (Brazil: 66%; Philippines: 71%; Taiwan: 89%; UK: 67%; USA: 70%), compared
Figure 4. Interaction effect of INETrust in News on Online News Use
DIGITAL JOURNALISM 1199
to the global average of 45% (Kemp 2019). Considering these numbers and given that
the autoregressive analyses only show limited evidence of significant relationships
over time for INE and social media news use, it appears that the reported level of
being incidentally exposed to news is strongly interrelated with reported social media
use across the 18 countries investigated.
Of course, this study does not come without limitations. First, the reliance on pure sur-
vey data regarding news use suggests the issue of social desirability in respondents’
answers (Schwarz and Oyserman 2001). Although news tracking data has shown that
there can be major differences between the amount of time people indicate in surveys
regarding news use and their actual news usage behavior, certain time references
have been found to be as reliable as actual tracking data (Haenschen 2020). However,
a news tracking study for 18 countries around the globe would have been resource-
intensive, let alone practically challenging in terms of privacy and data regulations.
Figure 5. Interaction effect of INETrust in News on SocialMediaNewsUse
1200 N. STRAUß ET AL.
Table 6. Incidental news exposure influencing news use variables across 18 countries.
Cross-Sectional Lagged Autoregressive
Argentina .16 (.04) –.04 (.04) .22 (.03) .12 (.07) .13 (.08) .24 (.07) .10 (.06) .18 (.07) .09 (.06)
Brazil .17 (.04) .06 (.04) .26 (.03) .17(.08) .21 (.08) .37 (.07) .07 (.07) .22 (.07) .22 (.07)
China .16 (.04) .08 (.03) .22 (.03) .03 (.08) .11 (.06) .20 (.05) –.05 (.06) .02 (.06) .12(.05)
Estonia .06 (.04) .13 (.04) .16 (.03) .05 (.04) .06 (.04) .07 (.05) –.00 (.03) –.04 (.04) .004 (.04)
Germany .17 (.04) .06 (.04) .18 (.04) .15 (.05) .13 (.04) .11(.05) .02 (.03) .09(.04) .01 (.04)
Indonesia .14 (.03) .05 (.04) .27 (.03) .07 (.07) –.03 (.06) .17(.05) .03 (.07) –.04 (.06) .01 (.05)
Italy .03 (.03) .01 (.03) .22 (.03) .07 (.04) .12(.04) .21 (.04) .07 (.04) .10(.04) .11(.04)
Japan .17 (.04) .09 (.03) .23 (.03) .18 (.06) .14 (.05) .14 (.04) .04 (.04) .08 (.04) .01 (.04)
Korea .04 (.04) .10 (.03) .13 (.03) .06 (.05) .09(.04) .09(.04) .02 (.03) .01 (.03) .03 (.03)
New Zealand .07(.04) .08(.03) .26 (.03) .11 (.05) .02 (.05) .26 (.05) .01 (.03) –.01 (.04) .04 (.04)
Philippines .20 (.04) –.07 (.03) .37 (.02) .14 (.12) .17 (.09) .48 (.06) .06 (.11) .11 (.07) .43 (.07)
Poland .11 (.04) .07 (.03) .17 (.04) .14 (.05) .07 (.04) .11(.05) .08(.04) .01 (.04) .04 (.04)
Russia .15 (.03) .04 (.03) .15 (.03) .19 (.05) .08 (.04) .11(.04) .04 (.03) .02 (.04) .07 (.04)
Spain .17 (.03) –.01 (.03) .27 (.03) .04 (.06) .001 (.06) .18 (.06) –.06 (.05) –.05 (.05) .005 (.05)
Taiwan .28 (.05) –.01 (.05) .33 (.04) .32 (.08) .16(.08) .24 (.07) .20 (.08) .17(.07) .16(.07)
Turkey .15 (.04) .002 (.03) .23 (.03) .29 (.07) .09 (.07) .17(.06) .22 (.06) .11 (.06) .10 (.05)
UK .07 (.04) .08(.03) .36 (.03) .02 (.05) .08 (.04) .31 (.04) –.02 (.03) .02 (.03) .08(.03)
USA .12 (.04) .08(.03) .34 (.03) .13(.06) .14(.05) .35 (.05) .03 (.03) .13 (.04) .07 (.04)
Notes. OLS Models controlling for all variables and interaction effects as shown in Table 4; standardized coefficients; standard errors in parentheses;
p<.001; color code for coefficient size:>.40; >.30; >.20; >.10; >.01
DIGITAL JOURNALISM 1201
Second, we did not measure cognitive processing or elaboration of the information
received from INE in this survey, but only reported active news consumption. Future
studies are thus invited to follow up on our study to test the cognitive mediation
model to its full extent (Eveland 2001).
Third, the distinction between our measurement of active news consumption and INE is
suboptimal. It could have been the case that individuals’responsestoactivenewsconsump-
tion questions might have captured some variance of INE, and thus might have adversely
affected our results. Future research should develop better measurements that allow a better
prediction of INE on active news consumption. Related to this—and the findings for individual
countries—the third issue arises with the respondents’ability to differentiate between INE and
actual news consumption online and on social media (cf. Broersma 2019). For some respond-
ents it could have been the case that INE means the same as scrolling through social media
platforms and actually reading headlines of news posts. While we did our best with specific
wordings in our questions (see Method section), future research is advised to find a better way
to disentangle the sequence of cognitive, elaborative and behavioral processes when encoun-
tering INE, using a “user-centered perspective”(Hasebrink 2016,373)andstudying“tactics to
navigate the digital news and information ecology”(Broersma 2019,516).Oneexampleof
such a study could be a non-intrusive observation, combined with think aloud protocols (cf.
umpel 2019). Yet studies of this kind will probably be limited to small samples and difficult
to be conducted across countries.
This is the first study in communication research to our knowledge that has investi-
gated INE across 18 countries and across various news platforms, thereby allowing to
make inferences about whether INE leads to active news consumption or not. The
results support the assumption that information encountered serendipitously leads to
actual news use in traditional, online and social news media. It is in this vein that we
ascribe online media and the pervasive flow of information in today’s news media
environment a bridging function for active news use across various platforms.
No potential conflict of interest was reported by the authors.
This research was supported by Grant FA2386-15-1-0003 from the Asian Office of Aerospace
Research and Development. Responsibility for the information and views set out in this study
lies entirely with the authors.
Nadine Strauß http://orcid.org/0000-0002-5050-7067
Homero Gil de Z
1202 N. STRAUß ET AL.
Aalberg, Toril, Arild Blekesaune, and Eiri Elvestad. 2013. “Media Choice and Informed Democracy:
Toward Increasing News Consumption Gaps in Europe?”The International Journal of Press/
Politics 18 (3): 281–303.
evol-Abreu, Alberto, and Homero Gil de Z
niga. 2017. “Effects of Editorial Media Bias
Perception and Media Trust on the Use of Traditional, Citizen, and Social Media News.”
Journalism & Mass Communication Quarterly 94 (3): 703–724.
Artz, L., and R. Kamalipour. 2003. The Globalization of Corporate Media Hegemony. Albany, NY:
om, A., and M. J. Belfrage. 2018. “News in Social Media.”Digital Journalism 6 (5): 583–598.
Boczkowski, Pablo J., Eugenia Mitchelstein, and Mora Matassi. 2018. “News Comes across When
I’m in a Moment of Leisure”: Understanding the Practices of Incidental News Consumption on
Social Media.”New Media & Society 20 (10): 3523–3539.
Boulianne, S. 2011. “Stimulating or Reinforcing Political Interest: Using Panel Data to Examine
Reciprocal Effects between News Media and Political Interest.”Political Communication 28 (2):
Brislin, R. W. 1980. “Translation and Content Analysis of Oral and Written Material.”In Handbook
of Cross-Cultural Psychology, edited by H. C. Triandis and J. W. Berry, 389–444. Boston, MA:
Allyn & Bacon.
Broersma, M. 2019. “Epilogue: Situating Journalism in the Digital. A Plea for Studying News
Flows, Users, and Materiality.”In The Routledge Handbook of Developments in Digital
Journalism Studies, edited by S. Aeldrige and B. Franklin, 515–526. New York: Routledge.
Eveland, W. 2001. “The Cognitive Mediation Model of Learning from the News: Evidence from
Nonelection, off-Year Election, and Presidential Election Contexts.”Communication Research
28 (5): 571–601.
Feezell, J. 2018. “Agenda Setting through Social Media: The Importance of Incidental News
Exposure and Social Filtering in the Digital Era.”Political Research Quarterly 71 (2): 482–494.
Finkel, S. E.(1995). Causal Analysis with Panel Data. Thousand Oaks, CA: SAGE.
Fletcher, R., and R. Nielsen. 2018. “Are People Incidentally Exposed to News on Social Media? a
Comparative Analysis.”New Media & Society 20 (7): 2450–2468.
Gil de Z
niga, H., and J. H. Liu. 2017. “Second Screening Politics in the Social Media Sphere:
Advancing Research on Dual Screen Use in Political Communication with Evidence from 20
Countries.”Journal of Broadcasting & Electronic Media 61 (2): 193–219.
Gil de Z
niga, Homero, Brian Weeks, and Alberto Ard
evol-Abreu. 2017. “Effects of the News-
Finds-Me Perception in Communication: Social Media Use Implications for News Seeking and
Learning about Politics.”Journal of Computer-Mediated Communication 22 (3): 105–123.
Gil De Z
niga, Homero, Eul
alia Puig-I-Abril, and Hernando Rojas. 2009. “Weblogs, Traditional
Sources Online and Political Participation: An Assessment of How the Internet Is Changing
the Political Environment.”New Media & Society 11 (4): 553–574.
Goyanes, M. 2020. “Antecedents of Incidental News Exposure: The Role of Media Preference, Use
and Trust.”Journalism Practice 14 (6): 714–729.
Habermas, J. 1989. The Structural Transformation of the Public Sphere: An Inquiry into a Category
of Bourgeois Society. Cambridge, MA: MIT Press.
Haenschen, K. 2020. “Self-Reported versus Digitally Recorded: Measuring Political Activity on
Facebook.”Social Science Computer Review 38 (5): 567–583.
Hallin, D. C., and P. Mancini. 2004. Comparing Media Systems: Three Models of Media and Politics.
Cambridge, UK: Cambridge University Press.
Hasebrink, U. 2016. “Audiences and Information Repertoires.”In The Routledge Companion to
Digital Journalism Studies, edited by B. Franklin and S. A. Eldrige, 364–374. New York:
Heiss, Raffael, and J€
org Matthes. 2019. “Does Incidental Exposure on Social Media Equalize or
Reinforce Participatory Gaps? Evidence from a Panel Study.”New Media & Society 21 (11–12):
DIGITAL JOURNALISM 1203
Hermida, A. 2016. “Social Media and the News.”In The Sage Handbook of Digital Journalism,
edited by T. Witschge, C. W. Anderson, D. Domingo, and A. Hermida, 81–94. Los Angeles, CA:
Himelboim, I., and Y. Limor. 2008. “Media Perception of Freedom of the Press: A Comparative
International Analysis of 242 Codes of Ethics.”Journalism: Theory, Practice & Criticism 9 (3):
Hopmann, David Nicolas, Anke Wonneberger, Adam Shehata, and Jonas H€
oijer. 2016. “Selective
Media Exposure and Increasing Knowledge Gaps in Swiss Referendum Campaigns.”
International Journal of Public Opinion Research 28 (1): 73–95.
Hox, J. 2002. Multilevel Analysis: Techniques and Applications. Mahwah, NJ: Lawrence Erlbaum
Karnowski, Veronika, Anna S. K€
umpel, Larissa Leonhard, and Dominik J. Leiner. 2017. “From
Incidental News Exposure to News Engagement. How Perceptions of the News Post and
News Usage Patterns Influence Engagement with News Articles Encountered on Facebook.”
Computers in Human Behavior 76: 42–50.
Kemp, S. 2019. “Digital around the World 2019.”Hootsuite. https://p.widencdn.net/kqy7ii/
Kim, Yonghwan, Hsuan-Ting Chen, and Homero Gil de Z
niga. 2013. “Stumbling upon News on
the Internet: Effects of Incidental News Exposure and Relative Entertainment Use on Political
Engagement.”Computers in Human Behavior 29 (6): 2607–2614.
Kim, Shinwoo, and Bob Rehder. 2011. “How Prior Knowledge Affects Selective Attention during
Category Learning: An Eyetracking Study.”Memory & Cognition 39 (4): 649–665.
Krugman, H., and E. Hartley. 1970. “Passive Learning from Television.”Public Opinion Quarterly
34 (2): 184–190.
umpel, A. S. 2019. “Getting Tagges, Getting Involved with News? A Mixed-Methods
Investigation of the Effects and Motives of News-Related Tagging Activities on Social Network
Sites.”Journal of Communication 69 (4): 373–395.
umpel, A.-S. 2019. “The Issue Takes It All?”Digital Journalism 7 (2): 165–186.
Lee, Jae Kook, and Eunyi Kim. 2017. “Incidental Exposure to News: Predictors in the Social Media
Setting and Effects on Information Gain Online.”Computers in Human Behavior 75:
Mitchell, A., K. Simmons, K. E. Matsa, and L. Silver. 2018. “Publics Globally Want Unbiased News
Coverage, But Are Divided on Whether Their News Media Deliver.”Per Research Center.http://
Molyneux, L. 2018. “Mobile News Consumption.”Digital Journalism 6 (5): 634–650.
Mummolo, J. 2016. “News from the Other Side: How Topic Relevance Limits the Prevalence of
Partisan Selective Exposure.”The Journal of Politics 78 (3): 763–773.
Newman, N., R. Fletcher, A. Kalogeropoulos, and Nielsen R. Kleis. 2019. Reuters Institute Digital
News Report 2019.https://reutersinstitute.politics.ox.ac.uk/sites/default/files/inline-files/DNR_
Nordensen, B. 2008. “Overload! Journalism’sBattleforRelevanceinanAgeofTooMuch
Information.”Columbia Journalism Review.http://www.uvm.edu/jbrown7/envjournalism/
Norris, P. 1996. “Does Television Erode Social Capital? A Reply to Putnam.”PS: Political Science &
Politics 29 (03): 474–480.
Oeldorf-Hirsch, A. 2018. “The Role of Engagement in Learning from Active and Incidental News
Exposure on Social Media.”Mass Communication and Society 21 (2): 225–247.
Pentina, Iryna, and Monideepa Tarafdar. 2014. “From “Information”to “Knowing”:ExploringtheRole
of Social Media in Contemporary News Consumption.”Computers in Human Behavior 35: 211–223.
Prior, M. 2007. Post-Broadcast Democracy: How Media Choice Increases Inequality in Political
Involvement and Polarizes Elections. Cambridge, UK: Cambridge University Press.
Purcell, K., L. Rainie, A. Mitchell, T. Rosenstiel, and K. Olmstead. 2010. “Understanding the
Participatory News Consumer. How Internet and Cell Phone Users Have Turned News into a
1204 N. STRAUß ET AL.
Social Experience.”Pew Research Center.http://assets.pewresearch.org/wp-content/uploads/
Schudson, M. 1998. The Good Citizen: A History of American Civic Life. New York: The Free Press.
Schwarz, Norbert, and Daphna Oyserman. 2001. “Asking Questions about Behavior: Cognition,
Communication, and Questionnaire Construction.”American Journal of Evaluation 22 (2): 127–160.
Shah, V., Nojin Kwak, and R. L. Holbert. 2001. “Connecting" and "Disconnecting" with Civic Life:
Patterns of Internet Use and the Production of Social Capital.”Political Communication 18 (2):
Shearer, E. 2018. “Social Media Outpaces Print Newspapers in the U.S. as a News Source.”Pew
Sheller, M. 2015. “News Now. Interface, Ambience, Flow, and the Disruptive Spatio-Temporalities
of Mobile News Media.”Journalism Studies 16 (1): 12–26.
ack, J., M. Djerf-Pierre, and A. Shehata. 2013. “The Dynamics of Political Interest and
News Media Consumption: A Longitudinal Perspective.”International Journal of Public Opinion
Research 25: 414–435.
ack, Jesper, and Adam Shehata. 2010. “Media Malaise or a Virtuous Circle? Exploring the
Causal Relationships between News Media Exposure, Political News Attention and Political
Interest.”European Journal of Political Research 49 (5): 575–597.
Stroud, N. J., C. Peacock, and A. L. Curry. 2019. “The Effects of Mobile Push Notifications on
News Consumption and Learning.”Digital Journalism 8: 32–48.
Tandoc, Edson C., and Erika Johnson. 2016. “Most Students Get Breaking News First from
Twitter.”Newspaper Research Journal 37 (2): 153–166.
Tewksbury, David, Andrew J. Weaver, and Brett D. Maddex. 2001. “Accidentally Informed:
Incidental News Exposure on the World Wide Web.”Journalism & Mass Communication
Quarterly 78 (3): 533–554.
Thato, S., K. J. Hanna, and B. Rodcumdee. 2005. “Translation and Validation of the Condom Self-
Efficacy Scale with Thai Adolescents and Young Adults.”Journal of Nursing Scholarship : An
Official Publication of Sigma Theta Tau International Honor Society of Nursing 37 (1): 36–40.
Tsfati, Y. 2010. “Online News Exposure and Trust in the Mainstream Media: Exploring Possible
Associations.”American Behavioral Scientist 54 (1): 22–42.
Tsfati, Y., and J. Cappella. 2005. “Why Do People Watch News They Do Not Trust? The Need for
Cognition as a Moderator in the Association between News Media Skepticism and Exposure.”
Media Psychology 7 (3): 251–271.
Tsfati, Yariv, and Joseph N. Cappella. 2003. “Do People Watch What They Do Not Trust?:
Exploring the Association between News Media Skepticism and Exposure.”Communication
Research 30 (5): 504–529.
an, Arturo Arriagada, and Andr
es Scherman. 2012. “The Social Media Basis of
Youth Protest Behavior: The Case of Chile.”Journal of Communication 62 (2): 299–314.
Valeriani, Augusto, and Cristian Vaccari. 2016. “Accidental Exposure to Politics on Social Media
as Online Participation Equalizer in Germany, Italy, and the United Kingdom.”New Media &
Society 18 (9): 1857–1874.
Van Damme, K., M. Martens, S. Van Leuven, M. V. Abeele, and L. De Merez. 2020. “Mapping the
Mobile DNA of News. Understanding Incidental and Serendipitous Mobile News
Consumption.”Digital Journalism 8 (1): 49–68.
Williams, A. E. 2012. “Trust or Bust? Questioning the Relationship between Media Trust and
News Attention.”Journal of Broadcasting & Electronic Media 56 (1): 116–131.
Yadamsuren, Borchuluun, and Sanda Erdelez. 2017. Incidental Exposure to Online News. Chapel
Hill, NC: Morgan & Claypool Publishers.
Yadamsuren, Borchuluun, and Jannica Heinstr€
om. 2011. “Emotional Reactions to Incidental
Exposure to Online News.”Information Research 16 (3). http://www.informationr.net/ir/16-3/
DIGITAL JOURNALISM 1205